Climate Change Impacts on Alpine Lakes

Summary

Introduction...... 7

1. Climate change and alpine lakes...... 8 Global Climate Change...... 9 Definitions...... 9 Climate change in the past...... 9 Human caused climate change...... 10 Consequences of present climate change...... 11 Future climate change...... 11 Climate change and lakes...... 11 Impacts on physical, chemical and biological lake parameters...... 12 Conclusions...... 13 References...... 16 List of figures...... 20 List of tables...... 20

2. Survey on climate change...... 22

Introduction...... 23 The past trend...... 24 Water temperatures...... 24 Air Temperature...... 29 Precipitation...... 32 Transparency...... 35 Oxygen...... 36 Phosphorus...... 40 Chlorophyll a...... 42 Phytoplankton – Chlorophyceae...... 43 Climate Driven Scenarios...... 44 Regional Climatic Scenarios in the Alpine Space...... 44 Multimodel Super Ensemble Technique...... 45 Climate Driven Alpine Scenarios...... 45 References...... 48 List of figures...... 48 List of tables...... 49

2

3. A model ecosystem for small mesotrophic and eutrophic sub-alpine lakes ...... 50 . Model formulation – single layer and spatially homogeneous conditions ...... 51 Simulation results for Lake Viverone – homogeneous model ...... 54 Model formulation – two vertical layers ...... 59 Simulations for Lake Viverone – Two layer model ...... 60 Effects of climate and environmental change ...... 62 A comment on the use of one- or two-layer models ...... 63 Application of the lake ecosystem models to other lakes ...... 63 References ...... 74 List of tables ...... 74 List of figures ...... 74

4. Hydrodynamic model...... 79 . Preface ...... 79 The hydrodynamic model ...... 79 ..Adaption.to.alpine.lakes...... 80 ..The.models.parameterization.and.meteorological.forcing...... 80 ..Calibrating.the.model.with.an.evolutionary.algorithm...... 81 . The three pilot sites ...... 82 ..Lake.Constance:.Data,.parameterization,.and.calibration.results...... 82 ..Lago.di.Viverone:.Data,.parameterization,.and.calibration.results...... 84 ..Wörthersee:.Data,.parameterization,.and.calibration.results...... 85 Scenario development ...... 86 ..Lake.Constance...... 86 ..Lago.di.Viverone...... 89 ..Lake.Wörthersee...... 90...... 90 Tracer calculations ...... 91...... Results and Discussion ...... 92 . Results.for.Lake.Constance...... 92...... Results.at.Lago.di.Viverone...... 93...... 93 ..Results.at.Lake.Wörthersee...... 94...... 94 Conclusion ...... 96...... 96 Appendix to the results of HCILS: Figures ...... 97...... References ...... 109...... List of figures ...... 109...... List of tables ...... 115

3

5. Hydrological water balance modelling, isotope investigations of lake circulation and residence time, meromixis and climate change ...... 116 Hydrological water balance modelling...... 117 Introduction...... 117 Methodology...... 117 Ossiacher See...... 120 Wörthersee...... 125 Klopeiner See...... 131 Conclusions of lake water balance determination by hydrological modelling...... 135 APPENDIX: Field survey to assist in modelling runoff . from the direct catchments into the lakes...... 137 Investigations of lake circulation and residence time of deep lake water using environmental isotopes...... 141 Introduction...... 141 Literature review...... 141 Methodology...... 142 Ossiacher See...... 144 Wörthersee...... 144 Klopeiner See...... 146 Summary and conclusions on hydrological water balance modelling and isotope investigations of lake circulation and residence time...... 148 References...... 150 List of tables...... 151 List of figures...... 151

6. Climate induced changes in water temperature and mixing behaviour of Carinthian lakes ...... 156 Introduction...... 157 Meromixis...... 158 Limnological characterisation of investigated lakes...... 158 Methodology...... 160 Long-term development of surface water temperature...... 161 Annual lake stratification – based on recorded water temperature...... 161 Limnological data...... 161 Climate - Air temperature...... 162 Result...... 162 Long term development of surface water temperature...... 162 Temperature stratification – based on recorded water temperature...... 164 Climate change input on mixing behaviour of meromictic lakes...... 165 Conslusion...... 167 References...... 167 List of figures...... 167 List of tables...... 168

4

7. Preparing for climate change in alpine lakes ...... 170

. Introduction ...... 171 . Climate Change and the Water Framework Directive ...... 171 ..The.Climate.Change.and.the.classification.of.the.lakes...... 171 . Climate change indicators...... 173 . Climate change adaptation and mitigation ...... 174 ..Adaptation.and.mitigation.options.for.alpine.lakes...... 175 . A support system for assessing Climate Change on alpine lakes ...... 179 ..The.Alpine.Lakes.Database...... 179 . Conslusions ...... 180 References ...... 181 List of figures ...... 182 List of tables ...... 182.

5

Introduction

Natural and artificial lakes are a main Lake. Research. State. Institute. for. Environment. for. Germany,. characteristic of the Alpine Space and belong, the.National.Institute.of.Biology.per.la.Slovenia,.Rhône-Alpes. regional. authority. for. French,. the. Regional. Government. of. with their catchment areas, to the European Carinthia. and. the. Joanneum. Research. for. Austria. and. the. heritage. They have much in common in Regional. Agency. for. the. environment. Protection. in. Piedmon,. the. Lombardy. Region. and. the. Agenzia. per. la. Protezione. their physical, ecological and even socio- Ambientale.di.Trento.for.Italy),.shared.methods.and.frames.of. economic features. Despite the administrative references,.and.tested.models.in.main.types.of.lakes,.to.identify. boundaries, they all convey the same identity. likely.scenarios.in.which.lakes.could.be.involved..The.network. was.designed.as.a.virtual.laboratory.producing.a.dynamic.vision. These lakes are important from the economical of.each.situation,.positioned.into.identified.general.trends.and. point of view, and they are used in different related.to.environmental.requirements. ways (touristic, industrial, agricultural, In order to reach the goals, WP4 has been subdivided in drinking water). 4 actions: • Action 1: Coordination.and.animation In. 2004. was. born. the. Alpine. lake. network. through. Alplakes. • Action 2: Specifying. lakes. bodies. and. shores. ecological. Project,. co-funded. by. European. Regional. Development. Fund. functioning. and. trends. by. collecting. and. integration. of. (ERDF).within.the.Alpine.Space.Programme.Interreg.IIIB,.with. ecosystem..indices,.related.to.climate.variability the.aim.of.sharing.information.and.management.tools.between. • Action 3: Characterizing. hydrological. impacts. of. climate. the.lakes.managers,.for.the.strategic.issues.within.the.principle. change. on. lakes. and. catchments. by. applying. hydro. and. of.sustainable.development. thermodynamic.models.and.isotopic.analysis For.a.3.year.period.(2004.-.2007).French,.Italian,.Austrian,.Swiss. • Action 4: Integrating. biological. hydrological. and. data. into. and.Slovenian.institution..worked.on.a.range.of.issues,.including. scenario.in.relation.with.main.alpine.lakes.types ecotourism,. sustainable. development. and. the. environment. of. the. Alpine. lakes,. in. order. to. preserve. and. improve. the. lake. Action. 1. is. provided. by. Arpa. Piemonte. to. plan. meetings,. fix. environment.. The. discussions. and. shared. experience. were. methodologies.and.make.transverse.approach.with.other.WPs. recorded.in.a.large.body.of.documents,.which.are.available.on. possible,.in.order.to.define.the.timetable.of.activities internet.(www.alpine-space.org). Since.lakes.are.complex.dynamic.systems,.interacting.with.local. Since.the.importance.of.this.subject,.the.institutions.involved.in. environment.and.connected.to.the.water.cycle,.WP4.followed. Alplakes.decide.to.continue.the.activity.started.in.2004,.starting. first.two.parallel.ways.to.throw.light.on.lakes.evolution.factors. up.in.2009.with.the.3-years.project.“Sustainable.Instrument.for. due. to. human. activities. and. climatic. variability:. the. biological. lakes.management.tools.in.Alpine.Space”.(SILMAS),.co-funded. approach.and.the.physical/chemical.approach by. European. Regional. Development. Fund. (ERDF). within. the. Action. 2. concerns. the. biological. approach:. the. collection. of. Alpine.Space.Programme.co-funded.by.FESR,.as.completion. chemical,. physical. and. biological. historical. data. is. used. to. of.the.Alplakes.project. analysed. the. trends. and. the. significant. ecological. events.... In. SILMAS. worked. 14. partners. of. 5. countries. (Italy,. France,. Action. 4.3. concerns. the. determination. of. climate. change. Austria,.Slovenia.and.Germany),.leaded.by.Region.Rhône-Alpes. on. lakes. and. in. the. catchment. areas..(hydrological. and. (France).(Fig. 1). thermodynamic. aspects,. mixing. conditions,. residence. times).. Applying.models.developed.by.LUBW.PP.and.isotopic.analysis. SILMAS. involved. scientists,. academics. and. technicians. from. will.deliver.data.of.past.and.scenarios.of.future.hydrological.and. the. public. authorities. in. charge. of. managing. the. lakes. that. mixing.conditions. pooled. their. knowledge,. and. identified. methods. and. tools. in. order.to.preserve.and.protect.the.lakes.environment. In. a. second. stage. this. knowledge. has. been. valorised. by. integrating. findings. of. 4.2. and. 4.3. investigations. into. helpful. SILMAS.focused.on.three.main.areas:.the.effects.on.of.climate. decision-making. tool. (action. 4.4). They. have. been. integrated. change.on.alpine.lakes.(Work package 4),.Resolving.conflicts. in. a. temporal. perspective,. to. describe. likely. scenarios. to. be. between.the.different.uses.of.the.lakes.(Work package 5).and. expected.from.climate.change,.linked.to.the.main.types.of.lakes. Educating.the.public.in.sustainable.development.as.it.relates.to. encountered. in. the. alpine. space.. Adjustment. strategies. have. the.Alpine.lakes.(Work package 6). been.outlined.for.stakeholders,.according.to.challenges.every. Within.SILMAS.project,.Work.Package.4.“Alpine.Lakes.running. networking..lake.will.have.to.meet,.in.terms.of.water.resource. changes”. worked. on. analysis. of. effects. on. climate. change. management.and.ecosystems.preservation.or.restoration. on. lakes. eco. system:. the. partners. involved,. (the. Institute. for.

7 1. Climate Change and Alpine Lakes

8 Global Climate Change

According. to. the. IPCC. (2007a). definition,. atmospheric. in.diurnal.cycles.that.create.atmospheric.motions.so.that.finally. greenhouse. gases. (GHG). absorb. more. than. other. gases. the. move. water. rather. than. ice. through. the. landscape,. forming. thermal. infrared. radiation. emitted. by. the. Earth’s. surface,. the. catchment.areas.as.well.as.lakes..Lakes.are.the.consequence. atmosphere.itself,.and.by.clouds..These.GHG.block.the.Earth. of. sensitive. interactions. among. processes. in. the. earth’s. back. radiation. to. space. causing. a. warmer. atmosphere. and. a. atmosphere..As.illustrated.in.Paragraph.2.1,.there.exists.a.strong. warmer.earth.surface. connection.between.lakes.and.their.local.climate.while.their.local. climate.is.coupled.to.the.global.climate..In.this.general.context,. This. is. a. simplified. sketch. of. the. greenhouse. effect. showing. lakes.are.likely.sensitive.to.changes.in.the.global.climate. one. important. process. in. the. complicated. radiation. transfer. processes. in. the. atmosphere.. Including. all. processes,. the. Life. on. Earth. aspires. as. a. perfect. fitness. by. evolutionary. thermodynamic.balance.explains.the.long-term.global.surface. optimization.in.all.possible.environments.e.g..yielding.a.variety. temperature.of.about.15°C..Without.greenhouse.gases.such.as. of.living.beings.in,.on,.and.around.the.lakes..This.is.especially. water.vapor,.carbon.dioxide,.nitrous.oxide,.methane.and.ozone,. true.for.human.beings.and.their.cultural.and.economic.lake.uses.. the.mean.earth’s.surface.would.be.much.colder.-18°C. Consequently,.changes.in.global.climate.may.bring.life.either.in. a.more.vulnerable.position.or.adapt.and.redistribute.into.other. Lakes.exist.by.the.presence.of.greenhouse.gases.because.these. species. gases.store.a.part.of.the.Sun’s.energy.into.the.atmosphere.and.

Definitions

Climate. is. defined. as. the. worldwide. (global). connection. Upon.analyzing.changes.in.the.climate.in.terms.of.temporarily- between. numerous. atmospheric,. lithospheric,. hydrologic,. and. averaged. weather,. long. periods. have. to. be. considered.. anthropogenic.processes..Climate.is.the.mean.condition.within.a. Notably,.changes.in.climate.conditions.such.as.mean.local.air. period.that.is.long.enough.for.determining.reliable.statistics..The. temperature. cannot. be. detected. over. two. or. three. years,. but. IPCC.states:.«Climate in a narrow sense is usually defined as the rather. over. decades.. According. to. the. IPCC. the. term. climate. average weather, or more rigorously, as the statistical description change. «…refers to a change in the state of the climate that in terms of the mean and variability of relevant quantities over can be identified […] by changes in the mean and/or the a period of time ranging from months to thousands or millions variability of its properties, and that persists for an extended of years. The classical period for averaging these variables is period, typically decades or longer. Climate change may be due 30 years, as defined by the World Meteorological Organization. to natural internal processes or external forcing, or to persistent The relevant quantities are most often surface variables such as anthropogenic changes in the composition of the atmosphere or temperature, precipitation and wind. Climate in a wider sense in land use.».(IPCC,.2007a). is the state, including a statistical description, of the climate system.».(IPCC,.2007a).

Climate change in the past

For. understanding. and. discriminating. the. effect. of. anthropogenic. greenhouse. gas. emissions. on. climate. change. from.natural.causes.it.is.necessary.to.overview.the.history.of.the. earth’s.climate. In. the. Jurassic. and. cretaceous. period. dinosaurs. breathed.

warm.air.with.a.CO2.concentration.up.to.a.maximum.of.around.. 4000. ppm1. (Berner. and. Kothavala,. 2001). whereas. the. meanmeasured. concentration. at. the. Mauna. Loa. Observatory,. Hawaii,. was391.6. ppmin. the. year. 2011. (Tans. and. Keeling,. 2012).. These. low. and. precise. values. are. estimated. either. byrecalculations. of. the. past. climate. as. by. using. proxy. data. like. boron. isotopes.. Higher. precision. is. achieved. by. ice. core. analysis.dating.several.hundred.thousand.years.back.to.present. (e.g..Petit.et.al.,.1999)..From.a.geological.perspective.this.period. is. very. short. although. it. covers. the. period. of. the. evolution. of. Figure 1.1 > Source: IPCC (2007), original figure description:‘Variations of deuterium (dD) Homo.sapiens. in antarctic ice, which is a proxy for local temperature, and the atmospheric

concentrations of the greenhouse gases carbon dioxide (CO2 ), methane (CH4), Figure. 1.1. is. quoted. from. the. IPCC’s. 4th. assessment. report. and nitrous oxide (N2O) in air trapped within the ice cores and from recent (IPCC,.2007a)..One.can.see.periods.of.high.GHG.concentrations. atmospheric measurements. Data cover 650,000 years and the shaded bands indicate current and previous interglacial warm periods’. but.longer.periods.with.low.N2O.and.CO2.concentrations..

9 1 «ppm (parts per million) or ppb (parts per billion, 1 billion = 1,000 million) is the ratio of the number of greenhouse gas molecules to the total number of molecules ofdry air. For example, 300 ppm means 300 molecules of a greenhouse gas per million molecules of dry air.» (IPCC, 2007b) These changes correspond to the glacial-interglacial cycle. actual warm period there were several variations of GHG This ‘Milankovitch Cycle’ is caused by changes in the earth’s concentrations in the atmosphere and thus in the global mean orbit (Rahmstorf and Schellnhuber, 2006) and thus not by temperatures, e.g. the 8.2ka2 event (Alley et al., 1997). The . anthropogenic influences. Here, high GHG concentrations are 8.2 ka event was adecrease in global temperature approximately caused by high temperatures with a GHG delay of centuries 8200 years ago.When focusing onto the last 2000 years volcanic to a millennium. There are natural causes and anthropogenic aerosols and changing solar activity where the reasons for causesfor climate change. In the last decades anthropogenic global climate changes (Böhm, 2008). Higher solar activity with influences overlap dramatically these natural processes. the contribution of less active volcanos led to a ‘Medieval Warm Within the ice core verified period climate changed several times Epoch’ around the 10th century. At that time parts of Greenland very rapidly. There were changes within decades during glacial were settled by Norseman (Pettersson, 1914). The following periods, e.g. the Dansgaard-Oeschger events with a warming cold conditions with minimum temperatures in the 16th century in Greenland from 8°C to 16°C (Severinghaus and Brook, 1999; can be put down to the so called ‘Maunder Minimum’ that was Masson-Delmotte et al.,2006). These changes are determined characterized by less solar activity (e.g. Lean, 2000). from local proxies and often they did not occur on the entire So we conclude that notable changes in the global climate planet (Masson-Delmotte et al., 2005). Instead, the transitions occurred without anthropogenic influences. In other words, in from the glacial to the interglacial periods were slower but global. principle not all changes in lakes can be attributed to (global) the There were numerous cold and warm phases in the current anthropogenic influences presented in the next section. interglacial that began around 10 thousand years ago. In this

Human caused climate change

From thousand years ago until the middle of the 1950s solar radiation and volcanoes caused climate changes. Since then, around 50 years ago, human activities appear to overwhelm the natural causes of climate change. In the time series of global mean temperature there is a significant upward trend before the 1950’s (see Figure 1.2) but this trend is caused by increasing solar activity which reached a plateau in the middle of the 20th century. At that time humans caused their first noticeable climate signal. Firstly, from 1950s to 1970s a slight decrease in global temperature by extensive aerosol emissions (regarding sulfur see Stern, 2005). Subsequent efforts in reducing air pollution led Figure 1.2 > Global Land-Ocean Temperature Index; base period: 1951-1980. Graph is based on data source: NASA Goddard Space Flight Center, to fewer aerosols in the air while GHG emissions continued to Science and Exploration Directorate, Earth Science Division. increase year by year. http://data.giss.nasa.gov/gistemp

And that’s why between the 1980s and today there is a change A constant global mean temperature implies an atmosphere in of global climate in a way that is unique in the younger climate a global thermodynamic equilibrium. The earth’s atmosphere, history. The actual mean global CO2 concentration in the however, is far away from this equilibrium. Hence, climate atmosphere of 391.6 ppm in the year 2011 (Tans and Keeling, change is better assigned as ‘unbalancing’ rather than warming. 2012) is the highest since approximately half-million years The patterns of the human caused climate change around the (IPCC 2007). And this concentration rises with an exceptional globe show this issue (e.g. Figure 1.3). speed but is still much less than the 4000 ppm analyzed for the Jurassic and cretaceous period. Methane, being the major anthropogenic GHG beside nitrous oxide, reaches atmospheric concentrations exceeding those of the last 650ka (Spahni et al., 2005). The main source of methane are wetlands, rice agriculture, biomass burning, and ruminant animals (IPCC 2007). Since the end of the 1990s, the increase of methane concentration stopped (Simpson et al., 2002). Water vapor is the dominant but not anthropogenic GHG. Water vapor is very uniformly distributed around the globe and indirectly influenced by other human GHG emission. Similar to the positive feedback of methane (higher concentration → rising temperatures → more methane production), feedback of water vapor content and air temperature is also positive. Warmer Figure 1.3 > Temperature Anomaly of annual mean temperature 1980-2011 vs 1951-1980. air cause more evaporation as well as evapotranspiration and Data Data source: NASA Goddard Space Flight Center, Science and Exploration consequently larger water vapor concentrations. Recent studies Directorate, Earth Science Division. http://data.giss.nasa.gov/gistemp/maps show that because of this feedback the positive temperature trend in Europe exceeds the global one (Philipona, 2005).

2 1 ka = 1000 years. 10 Consequences of present climate change

There.are.numerous.effects.of.present.climate.change.around. Further. effects. concerning. ice. are. changes. in. snow. cover,. in. the. globe:. Thermal. expansion. and. freshwater. from. melting. river.and.lake.ice,.in.sea.ice,.and.in.the.stability.of.ice.sheets. inland. ice. enlarge. the. oceans. volume. (Folland. et. al.. 2001).. and.ice.shelves. Consequently.in.the.20th.century,.the.sea.level.rose.1.5.to.2.mm. Not.only.lakes.are.affected.by.climate.changes.(see Chapter 1.3). per.year.(Miller.&.Douglas,.2004). but.also.other.hydrologic.subsystems.e.g..extreme.events.like. Glaciers. react. sensitive. to. changes. in. atmospheric. boundary. floods.and.droughts.are.more.often.and.there.are.shifts.in.annual. conditions..Numerous.studies.yield.decreasing.glacier.masses. precipitation.(see.Bates.et.al.,.2008,.for.an.overview.regarding. around.the.world.(IPCC,.2007).since.the.1970s. water.sources.and.climate.change)..The.list.seems.to.be.endless. Frozen. grounds. in. the. meaning. of. seasonal. or. continuous. frozen.grounds.(permafrost).are.defrosting.due.to.changing.heat. balance.at.the.surface.(e.g.permafrost.in.the.Alps:.VonderMühll. et.al.,.2004).

Future climate change

Predictions. of. the. future. global. climate. are. highly. uncertain. the. periods. 1980-1999. and. 2090-2099 (see Table 1).. Recent. because. of. the. complexity. and. the. chaotic. character. of. the. studies.show,.that.actual.GHG.emissions.are.at.the.upper.limit. earth’s.atmosphere.and.atmosphere-surface.interactions.with.its. of.the.pessimistic.IPCC-Scenarios.(University.of.Copenhagen,. numerous.positive.and.negative.feedbacks..Furthermore,.human. 2009).... beings.and.their.future.GHG.emissions.are.unknown..There.are. infinite. theoretical,. possible. developments. of. anthropogenic. GHG. emissions. and. all. natural. climate. forcing.. Nevertheless,. Likely.range.of.. Sea.Level.Rise a. bundle. of. climate. scenarios. can. cover. a. broad. range. of. Temperature.Change Scenario (m.at.2090-2099.. possibilities..Scenarios.enable.the.assessment.of.the.shape.of. (°C.at.2090-2099.relative. relative.to.1980-1999) a.likely.21st.century.climate..For.comparison.of.different.model. to.1980-1999) approaches,. to. assess. future. climate,. the. Intergovernmental. B1 1.1.–.2.9 0.18.–.0.38 Panel. on. Climate. Change. (IPCC). published. a. Special. Report. on. Emission. Scenarios. (SRES,. Nakicenovic&. Swart,. 2000).. A1B 1.7.–.4.4 0.21.–.0.48 Out. of. about. 40. Scenarios. three. scenarios. usually. selected:. Table 1A2 > Adapted and shortened from2.0.–.5.4 IPCC (2007), Table SPM.3. Projected0.23.–.0.51 global average A1B.(moderate),.A2.(pessimistic).and.B1.(optimistic)..In.global. surface warming and sea level rise at the end of the 21st century. analyses. these. scenarios. enable. together. with. Atmosphere- Ocean. General. Circulation. Models. (AOGCMs). a. careful. view. into. the. 21st. century..These. calculations. show. a. rising. global. . average.surface.temperature.in.a.range.of.1.1.to.5.4°C.between.

Climate change and lakes

Several. long. time. series. studies. have. shown. close. coupling. The.main.impacts.of.climate.change.on.freshwater.ecosystem. between.climate,.lake.thermal.properties.and.individual.organism. result.fr om.changes.in.air.temperature,.precipitation.and.wind. physiology,. population. abundance,. community. structure. and. regimes.. Freshwater. system. respond. by. changes. in. their. food-web. structure. (Weyhenmeyer. et. al.,. 1999;. Straile. 2000;. physical. characteristics. including. stratification. and. mixing. Gerten.and.Adrian,.2000;.Arhonditsis.et.al.,.2004):.understanding. regimes. of. lake. water. columns,. catchment. hydrology. or. the.functioning.of.this.complex.system.of.interactions.is.essential. changes. in. ice-cover. which,. in. turn,. may. induce. chemical. to.assess.the.risk.linked.to.the.future.management.of.alpine.lake. changes. in. habitat. (Kernan. et. al.,. 2010).. Rising. temperatures. resources,.and.it.is.a.main.activity.of.SILMAS.project... favour.cyanobacteria.in.several.ways..Cyanobacteria.generally. Since.1970.freshwater.biodiversity.has.decreased.more.drastically. grow.better.at.higher.temperatures.(often.above.25°C).than.do. than.marine.or.terrestrial.biodiversity.(Loh.&.Wackernagel,.2004)... other.phytoplankton.species.such.as.diatoms.and.green.algae.. This. is. the. result. of. a. complex. mix. of. stressors. and. impacts. This.gives.cyanobacteria.a.competitive.advantage.at.elevated. (Stanner. &. Bordeaux,. 1995;. Malmquist. and. Rundle,. 2002).. temperatures.(Paerl.and.Huisman,.2008). The.major.drivers.can.be.summarized.as.multiple.use.(such.as. Model.studies.predict.that.lake.temperatures,.especially.in.the. fisheries,.navigation.and.water.abstraction),.nutrient.enrichment,. epilimnion,.will.increase.with.increasing.air.temperature,.so.that. organic,.acidification.and.habitat.degradation..Climate.change. temperature. profiles,. thermal. stability. and. mixing. patterns. are. is. adding. further. stresses. (temperature. increase,. Hydrological. expected. to. change. as. a. result. of. climate. change. (Hondzo. &. changes). and. interacts. in. complex. ways. with. existing. ones. (. Stefan.1993;.Stefan.et.al.,.1998). Huber.et.al.,.2008). 11 In the vast majority of lakes, the vertical temperature distribution The increase of air temperature can bring to an increase of and the intensity of vertical mixing are determined predominantly water temperature in the hypolimnio, as verified in Ambrosetti by meteorological forcing at the lake surface. A change in & Barbanti, 1999, in Lake Maggiore (Italy): the deep waters of climatic conditions affecting this local meteorological forcing will Lake Maggiore contain a sort of climatic memory, represented therefore alter both thermal structure and vertical transport by by variations in the caloric content. This analysis, relating to the mixing, which in turn will affect the flux of nutrient and dissolved period 1963 – 1998, demonstrates that the caloric content in oxygen, as well as the productivity and composition of the the hypolimnion at the end of limnological year (i.e. when deep plankton (Imboden 1990; Reynolds, 1997); Phenology is often waters are formed), depends strictly on winter meteorological strongly influenced by temperature and precipitation (Hughes, parameters (wind run, air temperature and solar radiation), as 2000). well as on the quality of heat that can reach the deep layers In particular the climate change can cause, among others, the before and after the onset of thermal stratification. following effects in the lakes: Ecologically, the enhances growth of cyanobacteria in warm, • Earlier water warming in spring (Gronskaya et al., 2001) calm summer can be directly related to systematic variation in • Increase in water temperature both on the surface and at the local weather: such “blooms” were once considered to be deeper levels in lakes (Endoh et al., 1999;). an inevitable consequence of eutrophication but changes in the • Lenghthening of period in summer when lake water temperature weather also play a major part in their seasonal development exceed 10°C (Jarvet, 2000); (Paerl and Huisman, 2008). In particular, even relatively small • Shortening of periods with ice cover and decrease in its changes in the thermal characteristics of lakes can cause major thickness (Todd and Mackay, 2003) shifts in phytoplankton, bacterioplankton and zooplankton Analyses of long term data series demonstrate that such a populations as well as altering the rates of metabolic processes change has already occurred in recent decades: American and (e.g. Steinberg and Tille-Backhaus, 1990; Tulonen et al., 1994; Asian lakes in particular are analysed, and the results are an Weyhenmeyer et al., 1999; Gerten and Adrian, 2000; Arvola et increase of water temperature of about 2°C in the last 10 years al., 2002; Jasser and Atvola, 2003). This is because organism (Schindler et al., 1990; Hampton et al., 2008; Coats et al., 2006). are often adapted to certain narrow temperature ranges and In Lake Constance (warm monomictic Lake in Central Europe) because their life-cycle strategies can be highly sensitive to the mean annual water temperature has increased by 0.17°C variations in ambient water temperature (Chen and Folt, 1996). per decade since 1960 (Straile et al., 2003). This warming The importance of lakes to our understanding of potential is strongly related to increasing winter air temperatures and effects of climate change has been demonstrated both from affected the duration and extended of winter lake mixing, the analyses of how biological components of lakes may respond heat content of the lake and the vertical distribution of oxygen (e.g. Meisner et al. 1987; Hill and Magnuson 1990; Shuter and and nutrients. Reduced winter cooling favours the persistence Post 1990; Meisner 1990; Minns and Moore 1992), and from of small temperature gradients and may result in an incomplete recent analyses showing that most lakes act as a source of CO2 mixing of the lake. because they are supersaturated relative to the atmosphere (Cole et al. 1994).

Impacts on physical, chemical and biological lake parameters

Alpine lake ecosystem are vital resource, on the hand for lake atmospheric input of momentum (wind and waves), radiation biodiversity, and on the other hand for human uses of the water and heat, excluding lakes that are forced also by other (natural bodies (navigation, bathing, irrigation, energy, tourism…). Any or man-made) sources of thermal energy. alteration could lead to implications from the ecological, cultural, In a clear lake the radiation penetrates deeper into the water. social and economic point of view. If there is an abundance of phytoplankton, a part of the solar Climate warming has direct effects on the physical, chemical, radiation is converted to biomass by photosynthetic/chemical and biological characteristics of lakes, and it also operates on reactions rather than absorbed as heat. Even the absorption of lakes indirectly via modifications in the surroundings watershed, solar radiation per unit surface area is the same, the surface e.g. through shifts in hydrological flow pathways, landscape temperature is higher than in clear lakes, when a fraction of the weathering, catchment erosion, soil properties, and vegetation. incoming radiation is also absorbed in deeper water layers, and The interaction between variables, the feedback effects that when energy is distributed vertically on a larger water volume accelerate or dampen environmental change, and threshold (Rinke et al., 2010). effects by which lakes may abruptly shift from one environmental Disregarding river outflow and withdrawal of heat by heat state to another are important topic to analyze in order to pumps, the heat loss is generally caused by wind-forced as preserve the lake environment. well as free convective heat transport and evaporation at the This chapter focuses some of main physical, chemical and free lake surface. Free convection occurs when air above the biological responses of lakes to climate change that have been lake’s surface is lighter than remote air. The rate of evaporation revealed by recent research. and related heat content depends on air and water temperature, wind speed and air humidity. The higher the wind speed and the Physical impacts deficit of the water vapor’s partial air pressure, the more energy is lost by evaporation heat, and vice versa. The heat balance A lake is an open physical system with a direct coupling to of a lake is also affected by the balance between thermal the atmosphere in terms of energy, water mass and mixing. infrared radiation to the lake and the lake’s back radiation to the The distribution and transportation of thermal as well as atmosphere. kinetic energy within the water body strongly depends on the

12 These.heat-exchange.processes,.very.briefly.summarized.above,. al.,.2008)..Between.1962.and.2005.Lake.Constance’s.surface. may.yield.thermal.water.stratification..For.the.latter.the.surface. water.temperature.increased.by.0.03°C.per.year(Wahl,.2007)..At. lake.water.needs.to.exceed.the.temperature.of.maximum.water. Lake.Geneva.the.surface.water.temperature.increased.by.about. density.(typically.4.→C).and.also.requires.mild.turbulent.mixing. 1°C.since.the.1970s.(Perroud.&.Govette,.2010). that.limit.downward.heat.transport.into.the.deeper.part.of.the.en attenteClimate.change.also.affects.the.hypolimnion.water.temperature.. lake. This.signal.is.not.that.intense.as.in.the.epilimnion,.but.it.is.as.well. In.summer,.colder.water.with.higher.density.accumulates.in.the. significant.(see.the.above.mentioned.studies). lake’s.depth.(hypolimnion),.while.warmer.water.floats.near.the. Vertical.mixing.and.thermal.stratification (mixing).surface,.called.epilimnion..The.transition.from.the.colder. hypolimnion. to. the. warmer. epilimnion. is. characterized. by. the. Stability. of. thermal. stratification. is. a. consequence. of. the. thermocline.–.the.maximum.temperature.gradient.indicating.is. density.differences.between.the.water.layers..The.warmer.the. a. barrier. for. mixing.. In. a. larger. alpine. or. pre. alpine. lake. (e.g.. epilimnion,. compared. to. the. hypolimnion,. the. more. energy. is. LakeConstance). in. winter/spring. there. is. an. approximately. required.for.mixing.water.against.its.buoyancy.effects..Stability. uniform.temperature,.and.consequently,.if.there.are.no.salinity. is. here. defined. as. the. work. that. has. to. be. spent. to. transfer. effects,. uniform. density. distribution.. In. this. case. the. input. of. stratification. into. a. complete. vertical. mixing. (Schmidt,. 1928).. momentum.due.to.wind.enables.sufficient.mixing.of.the.entire. The.lakes.stability.changes.increase.if.air.temperature.increases.. water.body,.what.is.essential.for.O2,.and.nutrient.contents.in. For.instance,.at.Lake.Zurich.the.thermal.stabilization.increased. the.deep.water.layers..And.in.winter.the.water.temperature.is. by.20%.in.the.period.1947-1998.(Livingstone,.2003). closer.to.its.maximum-density.temperature.so.that.temperature. Mixing. is. strongly. reduced. by. increasing. stability. so. that. the. changes.induce.negligible.density.changes. present.climate.changes.heat.up.the.epilimnion.more.than.the. We. conclude. that. the. lake’s. physics. strongly. depends. on. the. hypolimnion..Wahl.(2005).proved.for.Lake.Constance.a.shift.of. local.climate.and.the.local.climate.affected.by.the.global.climate.. the.maximum.mixing.from.April.to.March,.and.ascribed.this.to. Consequently,. changes. in. atmospheric. conditions. demand. to. warmer.winter.temperatures..Peeters.et.al..(2002).showed.that. study.in.detail.the.impact.of.these.changes.on.the.lake’s.mixing. increasing. air. temperatures. lead. to. a. suppression. of. deeply. status. and. its. ecological/biological. evolution.. If. the. climate. penetrative.winter.mixing.events.in.Lake.Zurich..De.Stasio.et.al.. conditions.change,.a.lake.is.directly.affected,.hence,.all.live.in,. (1996).calculated.an.earlier.and.longer.onset.of.stratification.in. on,.and.around.the.water.have.to.face.changes.in.their.ecological. north-temperate.lakes. nice.within.the.physical.framework. Despite.the.importance.of.vertical.mixing.to.the.trophic.status. Therefore,. lakes. are. sentinels,. indicators. and. depending. on. in.deep.lakes.(Salmaso.et.al..2003).there.are.few.publications. their.size.also.regulators.of.(local).climate.change.(Williamson. which. give. a. general. perspective. to. physical. issues. in. Alpine. et. al.,. 2009).. They. are. the. deepest. elements. in. a. landscape. lakes. and. accumulate. all. the. catchments. information,. e.g.. in. the. sediments.. Additionally. changes. in. lakes. physics. often. are. Chemical impacts reactions.by.a.climate.change..Furthermore,.lakes.are.important. Climate.is.a.master.variable.for.ecologically.important.chemical. accumulators.and.regulators.in.cycles.of.matter..For.instance,. processes.(Kernan.et.al.,.2010)..An.increase.in.water.temperature. each.year.all.lakes.and.reservoirs.store.more.carbons.than.all. has. an. important. impact. on. lakes. chemical. processes. with. ocean.sediments.together.(Dean.&.Gorham,.1998). increases.in.pH.and.greater.in.lake-alkalinity.generation.(Psenner. Numerous. studies. proved. an. often. drastic. relation. between. and.Schmidt,.1992). changes.in.a.lake.system.of.any.kind.and.climate.change..E.g.. The. pH,. ionic. strength,. ionic. composition,. and. conductivity. O’Reilly.et.al..(2003).found.this.sensitivity.for.Lake.Tanganyika,. are.very.sensitive.and.easily.measurable.indicators.of.changes. Africa.. LakeBaikal’s. environment. dramatically. changes. within. in.weathering.rate,.as.well.as.water.balance..For.many.lakes,. the.last.decades.(Hampton.et.al.,.2008). there. can. be. challenges. in. disentangling. the. roles. of. internal. Water.temperature and.catchment.changes.with.respect.to.water.chemistry,.which. may. be. further. complicated. by. confounding. factors. such. as. Due.to.the.heat.fluxes.at.a.lakes.surface,.water.temperature.can. eutrophication,.acidification,.or.atmospheric.nitrogen.deposition. be.highly.correlated.to.the.climate.parameters,.in.particular.to. (Hessen.et.al..2009). the.air .temperature..Thompson.et.al.(2005).analyzed.45.small. In.alpine.environment,.within.small.lakes.catchment.area,.with.a. lakes. in. an. altitude. range. of. 1500m. to. 2000m. in. the. eastern. bedrock.of.granitic.gneiss.(by.silica.nature),.the.pH.is.influenced. Alps. (Niedere. Tauern).. They. analyzed. that. the. epilimnion’s. mainly.by.solubility.of.rocks,.that.increases.with.the.temperature. water.temperature.in.summer.is.very.sensitive.to.a.change.in. (Zobrist.e.Drever,.1990).and.from.reduction.processes.that.could. air. temperature.. In. an. exceptional. case. a. 6°C. increase. of. air. be. able,. for. depletion. of. Oxygen,. to. determine. an. increase. in. temperature. is. followed. by. an. increase. in. the. lake’s. surface. pH.(Koinig.et.al,.1998)..In.European.rivers,.Zwolsman.and.van. water.of.about.12°C..In.this.case.the.distinct.sensitivity.is.caused. Boìkhoven.(2007).and. Van.Vliet.and.Zwolsam.(2008).observed.an.. by.the.dependency.of.the.duration.and.thickness.of.ice.to.the. average.increase.in.water.temperature.of.around.2°C.respectively. epilimnion. temperature.. With. a. smaller. although. significant. in.Rhine.and.Meuse.rivers.after.the.severe.drought.of.2003,.with. sensitivity,. all. other. lakes. exhibit. a. change. of. epilimnic. water. a.pH.increase.(reflecting.a.decrease.in.CO2.concentration).and. temperature..Some.examples:.Between.1964.and.1998.the.Lake. a.decrease.in.dissolved.oxygen..(DO).solubility.reflecting.a.lower. Washington’s.epilimnic.water.temperatureincreased.by.0.045°C. DO.solubility.under.higher.water.temperatures..A.DO.decrease. per.year.(Arhonditsis.et.al.,.2004)..At.Lake.Zurich.the.epilimnic. can. also. be. associated. to. an. increase. in. DO. assimilation. of. water.temperature.increased.by.0.24°C.per.decade.in.the.period. biodegradable.organic.matter.by.microorganisms.(linked.to.an. between. the. 1950s. and. the. 1990s. (Livingstone,. 2003).. The. increase. in. Dissolved. Organic. Carbon. (DOC). (Prathumrataba.. earth’s.deepest.lake,.Lake.Baikal,.is.warming.up.too:.since.1946. et.al,.2008). the.mean.water.temperature.increased.by.1.21°C.(Hampton.et.

13 DOC concentrations is an important constituent of many natural When applicable, hypolimnetic oxygen concentrations have waters: it may be stored in soils for varying lengths of time before added value as indicators of climate change because they transport to surface waters. The humic substances generated have widespread consequences for internal nutrient loading by organic matter decomposition impart a characteristic brown (Pettersson et al. 2003), habitat size, and refuge availability (De color to the water due to the absorption of visible light by Stasio et al. 1996; Jansen and Hesslein 2004). these compounds. DOC thus influences light penetration into surface waters, as well as their acidity, nutrients availability Biological Impacts metal transports and toxicity (Kernan et al., 2010) as well as Phytoplankton changes observed in the catchment related to increased run-

off, permafrost melting, shifts in vegetation, and changes in The increase of CO2 in atmosphere has a direct effect on wetlands (Evans et al. 2006; Benoy et al. 2007; Keller et al. global temperature and, consequently, on many biological lake 2008), and increased CO2 concentrations (Freeman et al. 2004). processes. Lake plankton is susceptible to the variation of The rising of DOC documented (Freeman et al., 2001; Evans temperature, although the plankton populations are relatively et al., 2005, 2006; Vuorenmaa et al., 2006; Monteith et al., well buffered against the frequent fluctuations in temperature. 2007) is interpreted as evidence of climate change impacts The response of plankton populations to the climate change on terrestrial carbon stores due to rising temperatures and the is more evident considering a long-time series the direct increasing frequency and severity of summer drought (Freeman ecological effects of the influence of increasing temperature and et al., 2001; Hejzlar et al., 2003; Worral et al., 2004). Increasing it is relatively easy to predict future scenarios. More difficult is precipitation could also lead to increasing DOC concentrations, to understand the complicate indirect effects and to elaborate a first by increasing the proportion of DOC-rich water derived future model of plankton response. from the upper organic horizons of mineral soils and secondly The most plankton algae are able to maintain a normal biological by reducing water residence time, and hence DOC removal , behaviour relatively wide range of temperature. Nevertheless in lakes (Hongve et al., 2004) Rising levels of atmospheric CO2 there are important studies that shown the positive relationship influencing plant growth and litter quality were also proposed between photosynthetic capacity and temperature for selected to explain increased rates of DOC production (Freeman et al., species of phytoplankton. Some species show an exponential 2004). increase of common physiological processes until the maximum Concerning nutrients concentration, an increase of N of 25°C water temperature, other show the maximum at 41°C mineralization in soil due to an increase in mean soil temperature (Talling, 1957; Jewison 1979; Straskraba & Gnauk 1985). But is expected (Durchane et al., 2007). Water bodies quality is most part of phytoplankton populations are able to growth subjected to weather seasonality which has an important impact ranging from 5-25°C with peaks between 10-20°C, while the on their nutrient patterns (Zhu et al., 2005). A warmer climate will population growth drops at temperature above 25°C. create indirect impacts on water bodies like an increase nutrients Phytoplankton growth is also correlated to the availability of load in surface and groundwater (Van Vliet and Zwolsman, 2008) nutrients into the lake. Nutrient turnover could increase in warmer and counteract policies effects of external nutrient loading climate and cause or worsen the eutrophic status, removing the reduction.. phosphorus stored in the sediments (Hamilton et al, 2001). Nutrient concentrations and ratios in lakes are likely to be Increasing water temperatures favours Diatoms, but the altered as a consequence of changes in terrestrial export related bacterial biomass in summer is evened with shifts in dominance to climatic influences on weathering rates, precipitation, run-off from Diatoms to Cyanobacteria and the seasonal development (Sommaruga-Wögrath et al. 1997; Rogora et al. 2003; Bergström of phytoplankton composition could be quite different and Jansson 2006), fire frequency (Kelly et al. 2006; Westerling (Weyhenmaeyer, 2002). This effect is evident at temperature > et al. 2006), or terrestrial primary productivity (Boisvenue and 20°C when the populations of Cyanobacteria and filamentous Running 2006). Nutrient concentrations can also be affected by algae are favoured at higher temperature (Straile, 2000). internal processes related to changes in thermal structure and/or Moderate increase of water temperature in winter and spring primary productivity (Jeppesen et al. 2005; Wilhelm and Adrian causes a faster shift in phytoplankton population in early 2008). Furthermore, the longer water renewal times predictable summer, when the algal species have a growth ratio lower than in a warmer and drier climate could lead to a decline of SO4 winter and spring algae that appear fast-growing (Vincon-Leite et al, 2002). This event often produce significant algal blooms and NO3 concentrations and an increase in base cations and alkalinity (Schindler et al., 1990; Webster and Brezonik, 1995). that troubles lake managers. As regards the solutes concentrations in high mountain lakes, The presence of ice cover and amount of snow fell in winter Thies et al. (2007) observed, during the last two decades, control the algal population (Petterson, 1990; Adrian et al, 1995) a substantial increase in two remote high mountain lakes and influence the nutrient availability responsable for spring algal (Rassass see, Italy, and Schwarzsee ob Solden, Austria): the blooms with consequent higher peak of planktonic biomass high concentrations can be explained by an increase in the (Muller - Navarra, 1997). mobilization and release of solutes from active rock glaciers in The limnological classic view of seasonal phytoplanktonic the lake catchment entering the lakes via melt water, related to presences and compositions implies an order of dominance of the observed increase in average air temperature in the region species or groups during the annual cycle. Normally the scheme over recent decades (Auer et al., 2007). of dominance succession is expressed in simple way: Oxygen concentrations in lakes can indicate climate shifts because oxygen levels are strongly influenced by temperature and thermal structure (Hanson et al. 2006). For example, the Pioneer B C D = Climax extremely warm European summer of 2003 resulted in a long Where B, C, D represent the succession of dominances period of thermal stratification and increased hypolimnetic of diverse phytoplanctonic groups. oxygen depletion in some Swiss lakes (Jankowski et al. 2006).

14 In.lacustrine.ecosystem,.the.above.mentioned.succession.can. difficult.to.separate.the.ef fects.of.increasing.temperature.from. undergo. a. deviation. or. interruption. caused. by. the. external. the.availability.of.food. perturbation.producing.a.reversed.succession: Nevertheless,.considering.that.the.phytoplanktonic.community. depends..on.the.increase.of.water.temperature.in.conditions.of. stable.nutrients.availability.,.it.is.possible.to.assume.that.there. Pioneer B C is..a. direct. relationship. between. increasing. phytoplanktonic. biomass. and. increasing. presence. of. grazing. zooplankton.. In. fact,.higher.water.temperature.in.the.spring.produce.an.increase. perturbation of.phytoplankton.with.consequent.increase.of.Cladocera,.with. precision.Daphnia,.that.leads.to.phytoplankton.suppression.in. B C D = Climax primis.and.shift.from.a.dominance.of.larger.to.smaller.species. Temperature. normally. controls. the. growth. rates,. while. the. The.lacustrine.ecosystem.can.be.subjected.to.human.pressure. availability. of. food. controls. the. reproductive. capacity. of. and.in.this.case.different.“climax”.can.be.changed.in.short-time,. population. and. a. relevant. increase. of. water. temperature. in. as.shown.below. winter. has. less. effect. than. a. smaller. increase. in. the. summer. (Edmondson. &. Winberg,. 1971).. Such. behaviour. is. typical. of. a. wide. part. of. aquatic. invertebrates. in. case. of. slow,. but. perturbation perturbation continuous,.increase.of.water.temperature. Fish.. Climax A Climax A’ Climax A’’ Many.species.of.fish.depend.on.changes.of.water.temperature. short time (De.Stasio.et.al.1996;.Magnusson,.1997).for.their.survival.and. growth.. Temperature-increased. could. lead. to. a. change. in. Imagining. a. lake. free. of. human. pressure,. the. depletion. of. structure.and.distribution.of.fish.population,.mainly.producing. succession.is.due.only.to.the.climate.whims.years.by.years..If. a. different. growth. rate. of. predatory. influencing. on. food-web. the.perturbation.is.not.due.to.the.sudden.events.but.determinate. (Carpenter.et.al,.1985)..Can.be .expected.that.a.warmer.spring. by.the.slow.and.continuous.increase.of.water.temperature,.then. induce. a. forward. shift. from. planktivory. to. piscivory. feeding. the.”climax”.will.be.different.and.different.will.be.the.composition. caused.to.the.increased.by.wider.pressure.of.piscivora.predation. of.phytoplankton.community. (Olson,. 1996).. In. an. other. hand,. the. rate. of. predation. by. planktivorous.fish.can.provoke..a.strong.drop.of.the.zooplankton. presence,.especially.of.Daphnia.(Meher,.2000).. Climax A Climax A’ Climax A’’ So,.the.water.temperature.could.induces.a.direct.effect.on.prey- predator. interaction,. modifying. the. structure,. dominance. and. slow water temperature increase in long-time behaviour.of.fish.population.and.conditioning.also.the.ratio.inter. and.intra.feeding.planktonic.groups.(.Dodson.&.Wagner,.1996;. Beisner.et.al,.1996). Zooplankton.. The.zooplanktonic.populations,.formed.mainly.of.Rotifers.and. microcrustaceans. (Cladocera. and. Copepods). account. the. secondary. production. in. lakes. (Morgan,. 1980).. Higher. water. temperatures. support. the. zooplankton. population. shift. and. change.community.compositions. The.abundance.of.Rotifers,.Cladocera.and.Copepods.depends. to. the. trophic. web. of. each. zooplanktonic. group.. Some. of. these.have.a.vegetarian.behaviour,.other.are.predator,.besides. every. group. has. different. tolerance. to. the. increase. of. water. temperature,. i.e.. Rotifers. have. a. very. broad. temperature. tolerance. and. many. common. species. can. live. easily. ranging. from. 1. to. 25°C. (Berzins&Pejler,. 1989)...therefore,. it. is. often.

Conclusions

There.is.strong.evidence.that.climate.change.had.and.will.have. hypolimnion,. in. stability,. and..in..vertical. mixing. are. of. high. in.the.future.an.effect.on.the.physical,.chemical.and.biological. importance. for. the. lakes. ecology...On. the. other. end,. climate. characteristics.of.lake.ecosystem,.both.directly.though.changes. driven. changes. may. lead. to. a. strong. variablitiy. in. chemical. in.climate.driven.both.indirectly.through.interaction.with.other. compounds.in.the.lakes...The.residence.time.of.a.lake.influences. stressors.. On. the. one. hand,. physical. impacts. as. unregulated. the.chemical.composition.of.lake.waters.by.controlling.the.time. water. levels,. changes. in. water. temperature. in. epi-. and. available. for. biogeochemical. and. photochemical. processes.

15 to operate, the extent of accumulation and loss of dissolved The weather conditions also influence the biological dynamics and particulate materials, and the duration of biogeochemical both the phytoplanktonic successions and the growth of interactions with the lake sediments and littoral zone. In lakes zooplankton (Cushing, 1982). The verified increase of water that experience anoxic bottom water conditions and nutrient temperature of the lakes in the last years has surely influenced little release from the sediments, a prolonged residence time caused changes in the compositions of the phytoplankton community by reduced precipitation and inflows can result in increased and has outcome the change of zooplankton community, but phosphorus accumulation (internal phosphorus loading) and often for our lakes it is very difficult to separate the influence of eutrophication. Conversely, in regions that experience increased warmer climate from the local impulse caused by human impact. precipitation and water flow, the increased flushing of nutrients Nevertheless for our lakes, that are substantially free of outcome and phytoplankton may result in reduced algal production nutrients, the analysis of long-term data sets could confirm a (Vincent, 2009). closed relationship between increasing water temperature and changes in community structure (Blenkner, 2005).

References Adrian R, Deneke R, 1996. Possible impact of mild winters Benoy G, Cash K, McCauley E, Wrona F. Carbon dynamics on zooplankton succession in eutrophic lakes of the Atlantic in lakes of the boreal forest under a changing climate. Environ. European area. Fresuwater Biology, 36, 757-770. Rev. 2007;15:175-189. Adrian R, Deneke R, Mischke U, Stellmaker L, Lederer l, 1995. Bergström A-K, Jansson M. Atmospheric nitrogen deposition A long-term study of the Heiligensee (1975-1992). Evidence has caused nitrogen enrichment and eutrophication of lakes in for effects climate change on the dynamics of eutrofied lake the northern hemisphere. Glob. Change Biol. 2006;12:1–9. ecosystems. Archiv. für Hydrobiologie, 133, 315-337. Berner, R.A., and Z. Kothavala, 2001. GEOCARB III: A revised Adrian, R., and others, 2009. Lakes as sentinels of current model of atmospheric CO2 over phanerozoic time. Am. J. Sci., climate change. Limnol. Oceanogr. 54: 2283–2297. 301(2), 182–204. Alley, R.B., et al., 1997. Holocene climatic instability: A large, Blenkner T, 2005. A conceptual model of Climate-related effects widespread event 8200 years ago. Geology, 25, 483–486. on lake ecosystems. Hydrobiologia, 533, 1-14. Ambrosetti & Barbanti, 1999. Deep water warming in lakes: an Böhm, R., 2008. Harte und weiche Fakten zum Klimawandel. indicator of climatic change. J. Limnol. 58: 1-9. Der Sachverständige, Heft 4/2008. Arhonditsis, G. B., M. Winder, M. T. Brett and D. E. Schindler. Boisvenue C, Running SW. Impacts of climate change on 2004. Patterns and mechanisms of phytoplankton variability in natural forest productivity - evidence since the middle of the Lake Washington (USA). Water Research 38: 4013-4027. 20th century. Glob. Change Biol. 2006;12:862–882. Arhonditsis, G.B., M.T. Brett, C.L. De Gasperi, D.E. Schindler, Bushaw, K. L., R. G. Zepp, M. A. Tarr, D. Schulz-Janders, 2004. Effects of climatic variability on the thermalproperties of R. A. Bourbonniere, R. E. Hodson, W. L. Miller, D. A. Bronk LakeWashington. Limnol. Oceanogr. 49: 256–270. and M. A. Moran. 1996. Photochemical release of biologically Arvola, L., Palomäki, A., Lehtinen, S. & Järvinen, M. 2002, available nitrogen from aquatic dissolved organic matter. Nature ‘Phytoplankton community structure and biomass in two basins 381:404-407. of a boreal lake in relation to local weather and the North Atlantic Carpenter S R, Kitchell J F, Hodgson J R, 1985. Cascading Oscillation’, Verhandlungen der Internationalen Vereinigung für trophic interactions and lake productivity. BioScience, 35, 634- theoretische und angewandte Limnologie, vol 28, s. 700-704. 639. Auer I, Böhm R, Jurkovic A, Lipa W, Orlik A, Potzmann R, Christian Steinberg and Rolanda Tille-Backhaus. Re-occurrence Schöner W, Ungersböck M, Matulla C, Briffa K, Jones PD, of filamentous planktonic cyanobacteria during permanent Efthymiadis D, Brunetti M, Nanni T, Maugeri M, Mercalli L, artificial destratification J. Plankton Res. (1990) 12 (3): 661-664. Mestre O, Moisselin J-M, Begert M, Müller-Westermeier Coats, R., Perez-Losada, J., Schladow, S. G., Richards, G, Kveton V, Bochnicek O, Stastny P, Lapin M, Szalai S, R.C. and Goldman, C. R. (2006). The Warming of Lake Tahoe. Szentimrey T, Cegnar T, Dolinar M, Gajic-Capka M, Zaninovic Climatic Change 76:121-148. K, Majstorovic Z, Nieplova E, 2007. HISTALP - Historical Cole, J. J., N. F. Caraco, G. W. Kling, and T. W. Kratz. 1994. instrumental climatological surface time series of the greater Carbon dioxide supersaturation in the surface waters of lakes. Alpine region 1760-2003. International Journal of Climatology Science. 265:1568-1570. 27: 17-46. Chen CY and Folt CL., Consequences of fall warming Bates, B.C., Z.W. Kundzewicz, S. Wu, J.P. Palutikof, Eds., for zooplankton overwintering success. Limnology and 2008. Climate Change and Water. Technical Paper of the Oceanography. 41(5):1077-1086, 1996. Intergovernmental Panel on Climate Change, IPCC Secretariat, Geneva, 210 pp. Cushing D H, 1982. Climate and Fisheries. Academic Press, London. Beisner B E, McCauley E E, Wrona F j, 1996. Temperature - . mediated dynamics of planktonic food chains - the effect of De Stasio BT, Hill DK, Kleinhans JM, Nibbelink NP, Magnuson invertebrate carnivore. Freshwater Biology, 35, 219-231. JJ. Potential effects of global climate change on small north- temperate lakes: Physics, fish, and plankton. Limnol. Oceanogr. 1996 ; 41:1136–1149.

16 Dean, W.E., E. Gorham, 1998..Magnitude.and.significance.of. Hanson PC, Carpenter SR, Armstrong DE, Stanley EH, carbon. burial. in. lakes,. reservoirs,. and. peatlands.. Geology26:. Kratz TK. Lake. dissolved. inorganic. carbon. and. dissolved. 535–538. oxygen:.Changing.drivers.from.days.to.decades..Ecol..Monogr.. Dodson S I, Wagner A E, 1996..Temperature.affects.selectivity. 2006;76:343–363. of.Chaoborus.larvae-eating.Daphnia..Hydrobiologia,.325,.157- Hejzlar J., Dubrovsky M., Buchtele, J., and Ruzicka. M. 161.. 21003. The. apparent. and. potential. effects. of. cliamte. change. Ducharne A., Baubion C., Beaudoin N., Benoit M., Billen on.the.inferred.concentration.of.dissolved.organic.matter.in.a. G., Brisson N et al.,.Lonf.term.prospective.of.the.Seine.river. temperature. stream. (the. Malse. River,. South. Bohemia).. The. system:.confronting.climatic.and.direct.anthropogenic.changes.. Sience.of.the.Total.Environment,.310,.142-152 Sci.Total.Environ.2007;.375:292-311. Hessen DO, Andersen T, Larsen S, Skjelkvale BL, de Wit Endoh, S., Yamashita, S., Kawakami, M.,AND Okumura, Y., HA. Nitrogen. deposition,. catchment. productivity. and. climate. 1999..Recent.Warming.of.Lake.Biwa.Water,.Jap..J..Limnol.,.60,. as. determinants. of. lake. stoichiometry.. Limnol.. Oceanogr.. 223.-.228. 2009;54:2520–2528. Evans C.D. Monteith D.T and Cooper D.M., 2005.. Long. Hill, D. K., and Magnuson J.J, 1990. Éptential.effects.of.global. term. increase. in. surface. water. dissolved. organic. carbon:. climate.warming.on.the.growth.and.prey.consumption.of.Great. Observations,. possible. causes. and. environemtnal. impacts.. lake.sfish..Trans..Am..Fish..Soc..199:.2656-275 Environemntal.Pollution,.137,.55-71. Hondzo, M. and StefanH. (1993)..Regional.water.temperature. Evans CD, Chapman PJ, Clark JM, Monteith DT, Cresser caracteristics. of. lakes. subjected. to. climate. change,. Climatic. MS. Alternative.explanations.for.rising.dissolved.organic.carbon. cgange,.24.187.-.211. export. from. organic. soils.. Glob.. Change. Biol.. 2006;12:2044– Hongve D., Riise G., and Kristiasen JF, 2004..Increased.colour. 2053. and.organic.acid.concentrations.in.Norwegian.forest.lakes.and. Folland, C.K., et al., 2001. Observed. climate. variability. drinking. water. -. A. result. of. increased. precipitation?. Aquatic. and. change.. In:. Climate. Change. 2001:. The. Scientific. Basis.. Science,.66,.231.-.238. Contribution.of.Working.Group.I.to.the.Third.Assessment.Report. Huber V., Adrian., R., and Gertne D., 2008.. Phytoplancton. of.the.Intergovernmental.Panel.on.Climate.Change.[Houghton,. response. to. climate. warming. modified. by. trophic. state.. J.T.. et. al.. (eds.)].. Cambridge. University. Press,. Cambridge,. Limnology.and.Oceanography,.53.(1).1-13. United.Kingdom.and.New.York,.N Y, .USA,.pp..99-181. Hughes, L. 2000..Biological.consequences.of.global.warming:. Freeman C., Evanns C.D., and Montheith, D.T. 2001..Export.of. is.the.signal.already.apparent?.Trends.in.Ecology.and.Evolution. organic.carbon.from.peat.soils..Nature,.412.785 15,.56-61 Freeman C.,. others. Export. of. dissolved. organic. carbon. from. IMBODEN,D.M.1990. Mixing. and. transport. in. lakes:. peatlands. under. elevated. carbon. dioxide. levels.. Nature.. Mechanisms.and.ecological.relevance,.p.47-80..In.M.Tilzerand. 2004;430:195-198. C..Serruya,[eds.],.Large.lakes:.Ecological.structure.and.function. Gergel, S. E., M. G. Turner and T. K. Kratz. 1999..Dissolved. IPCC, 2007. Climate.Change.2007:.The.Physical.Science.Basis.. organic. carbon. as. an. indicator. of. the. scale. of. watershed. Contribution. of. Working. Group. I. to. the. Fourth. Assessment. influence. on. lakes. and. rivers.. Ecological. Applications. 9:1377- Report. of. the. Intergovernmental. Panel. on. Climate. Change. 1390. [Solomon,.S.,.D..Qin,.M..Manning,.Z..Chen,.M..Marquis,.K.B.. Gerten, D., Adrian, R. 2000. Climate-driven. changes. in. Averyt,M.. Tignor. and. H.L.. Miller. (eds.)].. Cambridge. University. spring. phytoplankton. dynamics. and. the. sensitivity. of. shallow. Press,. Cambridge,. United. Kingdom. and. New. York,. N Y, . USA,. polymictic.lakes .to.the.North.Atlantic.Oscillation..Limnology.&. 996.pp. Oceanography.45,.1058-1066. Jankowski T, Livingstone DM, Forster R, Bührer H, Gronskaya, T.P., George, D.G. & Arvola, L. 2001.. The. Niederhauser P. Consequences. of. the. 2003. European. heat. influence.of.long-term.changes.in.the.weather.on.the.thermal. wave. for. lakes:. Implications. for. a. warmer. world.. Limnol.. characteristics. of. lakes. in. the. UK,. Finland. and. Russia.. –9th. Oceanogr..2006;51:815-819. International.Conference.on..the.Conservation.and.Management. Jansen W, Hesslein RH..Potential.effects.of.climate.warming. of.Lakes...Session.5..Partnerships.for.Sustainable.Life.in.Lake. on.fish.habitats.in.temperate.zone.lakes.with.special.reference. Environments:.Making.Global.Freshwater.Mandates.Work..Shiga. to.Lake.239.of.the.experimental.lakes.area.(ELA),.north-western. Prefectural.Government,.Japan,.pp..43-46. Ontario..Environ..Biol..Fish..2004;70:1-22. Hamilton D P, Spillman G, Prescott K L, Kratz K, Magnuson Jarvet, A., 2000..Biological.consequences.of.global.warming:.is. J J, 2001.. Effects. of. atmospheric. nutrient. inputs. on. trophic. the.signal.already.apparent?.Tree,.15,.56-61. status.of.Crystall.lake,.Wisconsin..Verhandlungen.Internationale. Jasser, I. & Arvola, L. 2003. Potential.effects.of.abiotic.factors. Vereinigung.Limnologie,.28,.467-470 on. the. abundance. of. autotrophic. picoplankton. in. four. boreal. Hampton, S.E., Izmest’Eva, L.R., Moore, M.V., Katz, S.L., lakes..-.J..Plankton.Res..25:.873-88 Dennis, B. & Silow E.A. (2008).. Sixty. years. of. environmental. Jeppesen E.,. others. Lake. responses. to. reduced. nutrient. change.in.the.world’s.largest.freshwater.lake. loading-an. analysis. of. contemporary. long-term. data. from. 35. Hampton, S.E., L.R. Izmenst’eva, M. V. Moore, S. L. Katz, E. A. case.studies..Freshwater.Biol..2005;50:1747–1771. Silow, 2008. Sixty.years.of.environmental.change.in.theworld’s. Jewson D H, 1976..The.interactions.of.components.controlling. largest.freshwater.lake-Lake.Baikal,.Siberia..Glob.Change.Biol.. net.phytoplankton.photosynthesis.in.a.well-mixed.lake.(Lough. 14:.1947-1958. Neagh,.North.Ireland)..Freshwater.Biology,.6,.551-576.

17 Keller W, Paterson AM, Somers KM, Dillon PJ, Heneberry Meisner, J. D. 1990. Effect of climatic warming on the southern J, Ford A. Relationships between dissolved organic carbon margin of the native range of brook trout, Salvelinus fontinalis. concentrations, weather, and acidification in small Boreal Shield Canadian Journal of Fisheries and Aquatic Sciences. 47: 1065- lakes. Can. J. Fish. Aquat. Sci. 2008;65:786-795. 1070. Kelly EN, Schindler DW, St. Louis VL, Donald DB, Vlaclicka Meisner, J.D., J.L. Goodier, H.A. Regier, B.J. Shuter, and KE. Forest fire increases mercury accumulation by fishes via W.J. Christie. 1987. An Assessment of the Effects of Climate food web restructuring and increased mercury inputs. Proc. Nat. Warming on Great Lakes Basin Fishes. Journal of Great Lakes Acad. Sci. USA. 2006;103:19380-19385. Research, 13: 340-352. Kernan M., Battarbee R., Moss B., 2010. Climate change Miller, L., B.C. Douglas, 2004. Mass and volume contributions impacts on freshwater ecosystems, Blackwell Publishing to 20th century global sea level rise. Nature, 428, 406–409. Koining K.A., Schmidt R., Sommaruga-Wograth S., Tessadri Minns, C.K. and Moore, J.E. 1992. Predicting the impact of R., Psenner R., 1998. Climate change as the primay cause for climate change on the spatial pattern of freshwater fish yield pH shifts in a high alpine lake. Water Air Soil Poll., 104: 167-180. capability in eastern Canada. Climatic Change 22:327-346. Lean, J., 2000. Evolution of the Sun’s Spectral Irradiance Since Montheith D.T., Stoddard J.L. Evans C.D., et al., 2007. the Maunder Minimum. Geophysical Research Letters, Vol. 27, Dissolved organic carbon trends resulting from changes in No. 16, Pages 2425-2428. atmospheric deposition . Nature, 450, 537 – 540. Lenters, J.D., T.K. Kratz, C.J. Bowser, 2005. Effects ofclimate Muller-Navarra D C, Guss S, von Storch H, 1997. Interannual variability on lake evaporation: Results from a longtermenergy variability of seasonal succession events in a temperate lake to budget study of Sparkling Lake, northernWisconsin (USA). J. its relation to temperature variability. Global Change Biology, 3, Hydrology 308: 168–195. 429-438. Livingstone, D.M., 2003. Impact of secular climate change on Nakicenovic, N., R. Swart (Eds.), 2000. IPCC Special Report thethermal structure of a large temperate central European lake. on Emission Scenarios, Cambridge University Press, UK. pp 570 Clim. Change 57: 205–225. O’Reilly, C.M., S.R. Alin, P.-D.Plisnier, A.S. Cohen, B. A. MCKEE, Loh J., Wackernagel , M. (2004). Living planet Report 2004, 40 2003. Climate change decreases aquatic ecosystemproductivity PP WWF Interational, Gland, Switzerland. of Lake Tanganyika, Africa. Nature 424:766–768. Magnuson J J, Webster K E, Assel R A, Bowser C J, Olson M H, 1996. Ontogenetic niche shift in large-mouth bass: Dillon P J, Eaton J G, Evan H E, Fee E F, Hall R I, Mortsch variability and consequewnces for first year growth. Ecology, 77, L R, Schindler D W, Quinn F H, 1997. Potential effects of 179-190. climate change on aquatic system: laurentian great lakes and Paerl, H.W., J. Huisman. 2008. Blooms like it hot. Science, Vol. Precambrian shiekl region. Hydobiological Processes, 11, 826- 320, April 4, 2008:57-58. 873. Peeters, F., D.M. Livingstone, G.-H.Goudsmit, R. Kipfer, R. Malmqvist B., and Rundle S., 2002. Threats to the running Forster, 2002. Modeling 50 years of historical temperature water ecosystems of the world. Environemtnal Consertvation, profiles in a large central European lake. Limnology and 29, 134-153. Oceanography, 47, 186-197. Masson-Delmotte, V., et al., 2005. Rapid climate variability Perroud, M., S. Goyette, 2010. Impact of warmer climate on during warm and cold periods in polar regions and Europe. Lake Geneva water-temperature profiles. Boreal Environment ComptesRendus Geoscience, 337(10–11), 935–946. Research 15: 255-178. Masson-Delmotte, V., et al., 2006. Past and future polar Petit, J.R., et al., 1999. Climate and atmospheric history of the amplification of climate change: climate model intercomparisons past 420,000 years from the Vostok ice core, Antarctica. Nature, and ice-core constraints. Clim. Dyn., 26 (5), 513–529. 399, 429–436.

18 Petterson K, 1990..The.spring.development.of.phytoplankton.in. Spahni, R., et al., 2005. Atmospheric. methane. and. nitrous. lake.Erken:.species.composition,.biomass,.primary.production. oxide.of.the.late.Pleistocene.from.Antarctic.ice.cores..Science,. and.nutrient.conditions.-.a.review..Hydrobiologia,.191,.9-14. 310(5752),.1317–1321. Pettersson K, Grust K, Weyhenmeyer G, Blenckner T. Stanner, D., and Bordeau, P. (eds) (1995).. Europe’s. Seasonality.of.chlorophyll.and.nutrients.in.Lake.Erken-effects.of. Environment.. The. Dobris. Assessment.. European. Environemtn. weather.conditions..Hydrobiologia..2003;506:75–81. Agency,.Copenaghen. Pettersson, O., 1914. Climate. variations. in. historic. and. Stefan, H.G., Fang, X and Hondzo M.. Simulated. climate. prehistoric. time.. SvenskaHydrogr.. -. Biol.. Komm.. Skriften,. 5,. change.effects.on.year.round.water.temperatures.in.temperatue. 1–26. zone.lakes..Climatic.change,.40,.353-381 Philipona, R., B. Dürr, A. Ohmura, C. Ruckstuhl, 2005. Stern, D.I., 2005..Global.sulfur.emissions.from.1850.to.2000.. Anthropogenic. greenhouse. forcing. and. strong. water. vapor. Chemosphere,.58,.163–175. feedback. increase. temperature. in. Europe.. Geophysical. Straile A & Adrian R, 2000. The.North.Atlantic.Oscillationand. Research.Letters,.Vol..32,.L19809,.4.PP.,.2005. phytoplankton.dynamics.in.two.European.lakes.–.two.variations. Prathumratana I., Sthiannopkao S., Kim KW..The.relationship. on.general.theme..Global.Change.Biology,.6,.663-670. of.climatic.and.hydrological.parameters.to.surface.water.quality. Straile D, 2000. Meteorological. forcing. of. plankton. dynamics. in.the.lower.Meking.River..Environ.Int.2008;.34:.860-6 in.large.and.deep.continental.European.lakes..Oecologia,.122,. Psenner R., Schmidt R..Climate.driven.pH.control.of.remote. 44-50.. Alpine.lakes.effects.of.acid.deposition..Nature.1992;.356:781-3. Straile, D., Jöhnk, K.D., Rossknecht, H., 2003..Complex.effects. Rahmstorf, S., and H.J. Schellnhuber, 2006. Der.Klimawandel.. of.winter-warming.on.the.physico-chemical.characteristics.of.a. Beck.Verlag,.Munich,.144.pp. deep.lake..Limnology.&.Oceanography.48:.1432-1438. Reynolds, C.S.1997. Vegetation. processes. in. the. pelagic:. A. Straskraba, M & Gnauk, A, 1985. Freshwater. ecosystem. model.for.ecosystem.theory,.p.1–371..In.O.Kinne[ed.],.Excellence. modelling. and. simulation.. Developmens. in. Environmental. in.ecology..Ecology.Institute,.Oldendorf. Modelling,.8,.1-309..Elsevier.Amsterdam. Rinke, K., P. Yeates, K.-O.Rothhaupt, 2010..A.simulation.study. Street, F.A., A.T. Grove, 1979.. Global. maps. of. lake-level. of.the.feedback.of.phytoplankton.on.thermal.structure.via.light. fluctuations.since.30,000.yrB.P...Quaternary.Research,.Volume. extinction..Freshwater.Biology,.Volume.55,.Issue.8,.pages.1674– 12,.Issue.1,.Pages.83-118. 1693,.August.2010. Tailling D F, 1957..The.phytoplankton.population.as.a.compound. Rogora M, Mosello R, Arisci S..The.effect.of.climate.warming. photosynthetic.system..New.Phytologist,.56,133-149. on. the. hydrochemistry. of. alpine. lakes.. Water,. Air,. Soil. Poll.. Tans, P., R. Keeling, 2012.. NOAA/ESRL. (www.esrl.noaa. 2003;148:347–361. gov/gmd/ccgg/trends/),. Scripps. Institution. of. Oceanography. Salmaso, N., R. Mosello, L. Garibaldi, F. Decet, M.C. Brizzio, (scrippsco2.ucsd.edu/). P. Cordella, 2003.. Vertical. mixing. as. a. determinant. of. trophic. Thies H., Nickus U., Mair V., et al.,. Unexpected. response. status. in. deep. lakes:. a. case. study. from. two. lakes. south. of. of. high. alpine. lake. waters. to. climate. warming.. Environemntal. the. Alps. (Lake. Garda. and. Lake. Iseo).. Journal. of. Limnology,. Science.and.Technology,.41.7424.–.7429. 62(Suppl..1):.33-41. Todd M.C., and Mackay A.W., 2003.. Large. –scale. climate. Schindler D.W., Beaty K.G., Fee, E.J., et al., 1990..Effects.of. controls.on.the.Lake.Bajkal.ice.cover..J..Climate,.16.3186-3199. climatic.warming.on.lakes.of.the.central.boreal.forest..Science,. Tulonen T; Kankaala P; Ojala A; Arvola L,.Factors.Controlling. 250,.967-970. Production. Of. Phytoplankton. And. Bacteria. Under-Ice. In. A. Schindler, D. W. and P. J. Curtis. 1997.. The. role. of. DOC. in. Humic,.Boreal.Lake...Journal.of.plankton.research.16(10),.1994,. protecting. freshwaters. subjected. to. climatic. warming. and. pp..1411-1432 acidification.from.UV.exposure..Biogeochemistry.36:1-8. University of Copenhagen, 2009.. Synthesis. Report. from. Schindler, D. W., P. J. Curtis, S. E. Bayley, B. R. Parker, K. G. Climate.Change,.Global.Risks,.Challenges.&.Decisions. Beaty and M. P. Stainton. 1997..Climate-induced.changes.in. Van der Kamp, G., D. Keir, M.S. Evans, 2008..Long-termwater. the.dissolved.organic.budgets.of.boreal.lakes..Biogeochemistry. level. changes. in. closed-basin. lakes. of. the. Canadianprairies.. 36:9-28. Can..Water.Res..J..33:.23–38. Schmidt, W., 1928.. Über. die. Temperatur-. und. Van Vliet MTH, Zwolsman JJG. Impact of summer droughts Stabilitätsverhältnisse.von.Seen,.Geogr..Ann..10,.Heft.1.und.2. on the water quality of the Meuse river. J. Hydrol 2008..353:. (1928). 1-.17. Severinghaus, J.P., and E.J. Brook, 1999. Abrupt. climate. Vincent, W F, Effects of Climate Change on Lakes, Elsevier, change.at.the.end.of.the.last.glacial.period.inferred.from.trapped. 2009. air.in.polar.ice..Science,.286(5441),.930–934. Vincon-Leite B., Tassin B., Druart J., 2002..Phytoplankton. Shuter B.J., and Post J.R.,.Climate.population.viability,.and.the. dynamics. an. meteorology.. Lakes&Reservoirs:. Research. an. zoogeography.of.temperature.fishes..Trans..Am..Fish..Soc..199:. management,.7,.93-102.. 314.–.336.pringer. VonderMühll, D., J. Nötzli, K. Makowski, and R. Delaloye, Simpson, J.I., et al., 2002..Implication.of.the.recent.fluctuations. 2004.. Permafrost. in. Switzerland. 2000/2001. and. 2001/2002.. in. the. growth. rate. of. tropospheric. methane.. Geophysical. Glaciological. Report. (Permafrost). No.. 2/3,. Glaciological. Research.Letters,.Vol..29,.No..10,.10.1029/.2001GL014521. Commission.of.the.Swiss.Academy.of.Sciences,.Zurich,.86.pp. Sommaruga-Wögrath S, Koinig K, Schmidt R, Tessadri R, Sommaruga R, Psenner R..Temperature.effects.on.the.acidity. of.remote.alpine.lakes..Nature..1997;387:64–67. 19 Vuorenmaa, J., Forsius M., and Mannio J. 2006. Increasing Wilhelm S, Adrian R. Impact of summer warming on the trends of total organic carbon concentrations in small forest thermal characteristics of a polymictic lake and consequences lakes in Finland from 1987 to 2003. The Science of the total for oxygen, nutrients and phytoplankton. Freshwater Biol. Environment, 365, 47 – 65. 2008;53:226–237. Wahl, B., 2007. Kovarianzanalysen physikalischer, chemischer Williamson, C.E., J.E. Saros, W.F. Vincent, J.P. Smol, 2009. und biologischer Langzeitdaten des Bodensees mit Lakes and reservoirs as sentinels, integrators, and regulators of hydrometeorologischen Parametern. KLIWA-Berichte Heft 11: climate change. Limnol. Oceanogr., 54(6, part 2), 2009, 2273- Zum Einfluss des Klimas auf den Bodensee, Teil 1. 2282. Webster K.E., and Brezonik, P.L.: 1995. Climate confound Worral F., Harriman, R., Evans C.D., et al., 2004. Trends in detection of chemical trends related to acid deposition in upper dissolved organic carbon in UK rivers and lakes. Biogeochemistry, Midwest lakes in the USA. Water, air and soil pollution, 85, 1575- 70, 369 – 402. 1580. Zhu Z., Arp PA., Mazumder A., Meng F., Bourque CPA, Foster Westerling AL, Hidalgo HG, Cayan DR, Swetnam TW. NW. Modeling stream water nutrient concentrations and loadings Warming and earlier spring increases western US forest wildfire in respondse to weather condition and forest harvesting. Ecol. activity. Science. 2006;313:940–943. Model 2005: 185: 231 – 43. Weyhemejer G.A., Adrian R, Gaedke U, Livingstone D M, Zobrist J., Drever J.J., 1990. Weathering processes in Alpine Maberly S C, 2002. Response of phytoplankton in European watersheds sensitive to acidification. IN: Proceedings of the EEC lakes to a change in the North Atlantic Oscillation. Verhandlungen Workshop “Acidification processes in remote mountain lakes”, Internationale Vereinigung Limnologie, 28, 1436-1439. Pallanza, 20-22 June 1989. Air Pollution Research Report,. Weyhenmeyer, G.A., Blenckner,T. and Pettersson,K. (1999). 20: 179 - 161 Changes of the plankton spring outburst related to the North Zwolsman JJG, van Bokhoven AJ., Impact of summer Atlantic Oscillation. Limnol. Oceanogr., 44, 1788–1792. droughts on water quality of the Rhine River – a preview of climate change? Water Sci Technol 2007; 56:44 – 55.

List of figures

Figure 1.1...... 7 Figure 1.3...... 8 Source: IPCC (2007), original figure description:‘Variations Temperature Anomaly of annual mean temperature 1980-2011 of deuterium (dD) in antarctic ice, which is a proxy for local vs 1951-1980. Data Data source: NASA Goddard Space Flight temperature, and the atmospheric concentrations of the Center, Science and Exploration Directorate, Earth Science

greenhouse gases carbon dioxide (CO2), methane (CH4), and Division. http://data.giss.nasa.gov/gistemp/maps

nitrous oxide (N2O) in air trapped within the ice cores and from recent atmospheric measurements. Data cover 650,000 years and the shaded bands indicate current and previous interglacial warm periods’.

Figure 1.2...... 8 Global Land-Ocean Temperature Index; base period: 1951- 1980. Graph is based on data source: NASA Goddard Space Flight Center, Science and Exploration Directorate, Earth Science Division. http://data.giss.nasa.gov/gistemp

List of tables

Table 1.1...... 9 Adapted and shortened from IPCC (2007), Table SPM.3. Projected global average surface warming and sea level rise at the end of the 21st century.

20

2. Survey on climate change

22 Introduction

The. biological. -. ecological. approach. of. Work. Package. 4. of. The.data.source.for.each.lake.are.the.following:. SILMAS. project. foresaw. the. collection. of. chemical,. physical,. •..Lake.Annecy.-.SILA.–.INRA.for.quality.data,.Meteo.France.-. biological.and.meteorological.data.concerning.the.alpine.lakes. SILA.for.meteorological.data in.order.to.highlight.the.most.significant.ecological.events,.and. •..Lake. Viverone,. Lake. Avigliana. Grande,. Lake. Sirio,. Lake. the. analysis. of. the. ecological. trends.. Arpa. Piemonte,. as. WP. Candia,.Lake.Mergozzo,.Lake.Orta:.Agenzia.Regionale.per.la. leader,.tried.to.collect.the.most.relevant.alpine.lakes.quality.data.. Protezione.Ambientale.del.Piemonte.(Arpa.Piemonte) Unfortunately,. most. of. the. investigated. lakes. have. very. poor. •..Lake. Bodensee:. ©. BOWIS. -. Daten. aus. dem. Bodensee- data.and.almost.no.environmental.information,.and.between.the. Wasserinformationssystem. der. Internationalen. lakes.with.data,.there.are.very.important.difference.in.monitoring. Gewässerschutzkommission.für.den.Bodensee.(IGKB). timing.. •..Lake.Caldonazzo,. Lake. Levico,. Lake. Santa. Giustina,. Lake. Piazze,.Lake.Toblino,.Lake.Cavedine,.Lake.Ledro,.Lake.Garda:. In.this.chapter,.it.is.possible.to.observe,.through.the.available. Agenzia.Provinciale.per.la.Protezione.dell’Ambiente.(APPA) data,. the. long. term. changes. measured. in. ten. investigated. alpine. lakes,. located. in. different. regions:. three. Austrian. lakes. These.lakes.are.very.different.concerning.morphology,.altitude,. (Wörthersee,.Klopeiner.See.and.Ossiacher.See),.1.lake.located. climate,.water.use..Different.types.of.lakes,.and.lakes.in.different. in. Germany,. Switzerland. and. Austria. (lake. Constance),. five. geographical.and.climatic.regions.respond.in.different.ways.to. Italian. lakes. (3. located. in. Piedmont. region. -. Lake. Sirio,. Lake. the.observed.variations.in.the.weather.. Grande.di.Avigliana.and.Lake.Viverone.-.and.two.lakes.Trentino. The.Table 2.1.resume.the.main.characteristics.of.the.investigated. Alto. Adige. region. -. Lake. Caldonazzo. and. Lake. Levico). and. a. lakes. French.lake.(Lake.Annecy)..

MAX. MAX. AVERAGE RESIDENT TROPHIC SURFACE MAX. AVERAGE VOLUME NAME LENGTH WIDTH ALTITUDE TIME STATE (km2) DEPTH (m) DEPTH (m) (Mm3) (km) (km) (m a.l.s.) (years)

Lake.Ossiacher. MESOTROPHIC 163 10,4 1,5 501 52,6 19,9 206,3 1,8 See Lake.Wörther. MESOTROPHIC 19,38 16,1 1,7 493 85,2 42,1 816,32 10,5 See Lake.Klopeiner. MESOTROPHIC 1,11 1,80 0,80 446 48 23 25 12 See

Lake.Annecy OLIGOTROPHIC 27 14,6 3,2 446,0 65,0 41,0 1125,0 3,5

Lake.Constance MESOTROPHIC 535,0 63,0 13,0 395,0 254,0 90,0 48,4 about.4,5

Lake.Caldonazzo MESOTROPHIC 5,3 4,735 1,87 449,0 49,0 27,0 149,0 3,6

Lake.Viverone EUTROPHIC 6 3 3 229 50 23 132 ~.35.(*)

Lake.Grande.di. EUTROPHIC 0,9 1,2 0,8 346,0 26,0 19,5 17,2 2,3 Avigliana

Lake.Sirio EUTROPHIC 0,3 0,9 0,5 266,0 43,5 18,0 5,2 5,7

Lake.Levico MESOTROPHIC 1,16 2,84 0,95 440 38 11

Table 2.1 > Main characteristics of the investigated lakes.

23 The past trend

The statistical methods used to analyse the collected data is the bottom of the lakes. In Europe, lake surface waters are typically linear regression analysis, used to calculate rates of change of at their warmest in July and August. temperature, and the Hurst exponent. For Austrian and Trentino lakes the data unfortunately are very All the statistical analysis were performed on the raw, measured poor, and it is difficult to show a significant trend, comparable data. Details of the measurements, such as the time periods to others lakes. For this lakes, we showed in this framework a on which the analysis were based, are given in the appropriate monthly trend, both for winter and summer temperatures. figures. Particular attention is paid to the interannual and seasonal water temperature variations at the surface and at the

Water temperatures

The water temperature has extremely important ecological Aim of this survey is to provide an overview of our knowledge of consequences. Temperature exerts a major influence on the impact of climate change on thermal characteristic of alpine aquatic organisms with respect to selection/occurrence and investigated lakes. level of activity of the organisms. In general, increasing water temperature results in greater biological activity and more rapid Observed climate trends: winter growth. All aquatic organisms have preferred temperature in The analysed data indicate that during the last years there has which they can survive and reproduce optimally. For example, been a common rising trend in winter water temperatures in the trout typically need cold water which may not be available in lakes in the alpine area, at the surface and at the bottom, as shallow waters during the summer. showed in the following diagrams (Figure 2.1). Temperature is also an important influence on water chemistry. Depending on the data availability, for all lakes the winter Rates of chemical reactions also generally increase with period analysed is December, January and February; except for increasing temperature. Temperature is a regulator of the solubility Austrian lakes, that is the month of December, and for lakes of of gases and minerals (solids) - or how much of these materials Trentino Alto Adige, that for which is available only the month of can be dissolved in water. The solubility of important gases, March. such as oxygen and carbon dioxide increases as temperature Lake Constance, a worm monomictic lake, has the longest decreases. For example, warm water contains less dissolved temperature data series, with a temperature profile measured at oxygen (DO) than cold water. Inversely the solubility of most approximately monthly intervals since 1960s. The annual water minerals increases with increasing temperature. Meteorological temperature mean has increased by 0,17°C per decade since forcing at the air –water interface is the main determinant of the 1960s (Straile et al., 2003). Winter warming rate for this lake is heat balance of most lakes (Edinger et al., 1968; Sweers, 1976). + 0,19°C per decade at the surface and + 0,10°C at the bottom. However, lakes that differ with respect to their morphometry Very similar results for Annecy lake, a French oligotrophic lake. respond differently to these changes (Gorham, 1964). For Piedmontese lakes (Grande di Avigliana, Viverone and The depth of the thermocline has important consequences on Sirio Lakes) the data available are only for the last decade. The the biology and chemistry of a lake. positive winter trend of the last decade for these lakes is line with the trend of the other alpine lakes.

Lake Grande di Avigliana Lake Sirio

Surface (°C per year) > + 0,16 DJF Bottom (°C per year) > + 0,12 DJF Surface (°C per year) > + 0,08 DJF Bottom (°C per year) > + 0,08 DJF

Figure 2.1 > Warming winter trend in alpine lakes investigated, showing linear regression lines and their gradients, at the surface and at the bottom (Data source: see paragraph 2.1)

24 Lake Viverone Lake Constance

Surface (°C per year) > + 0,68 DJF Bottom (°C per year) > + 0,08 DJF Surface (°C per year) > + 0,19 DJF Bottom (°C per year) > + 0,10 DJF

Lake Annecy Wörthersee

Surface (°C per year) > + 0,13 DJF Bottom (°C per year) > + 0,05 DJF Surface (°C per year) > + 0,066 December Bottom (°C per year) > + 0,003 December Klopeiner See Ossiacher See

Surface (°C per year) > + 0,16 December Surface (°C per year) > + 0,025 December Bottom (°C per year) > + 0,12 December Bottom (°C per year) > + 0,008 December Lake Levico Lake Caldonazzo

Surface (°C per year) > + 0,006 March - April Surface (°C per year) > + 0,007 Month of March Bottom (°C per year) > + 0,012 March - April Bottom (°C per year) > + 0,01 DJF Month of March

25 Observed climate trends: summer In summer, Lake Constance presents a warming summer trend The analysed data indicate that during the last years there has of about + 0,54°C per decade at the surface and + 0,06°C per been a common rising trend in summer water temperatures decade at the bottom. For Lake Grande di Avigliana and Lake (Figure 2.2) in the investigated alpine lakes during the last Sirio the summer trend is increasing, but slightly less compared decade, at the surface and at the bottom. with the winter trends. For Lake Viverone and Lake Annecy the summer trend is similar to winter trend. For Trentino lakes the available data show a slight increase for the months of July – August.

Lake Viverone Lake Constance

Surface (°C per year) > + 0,68 DJF Bottom (°C per year) > + 0,08 DJF Surface (°C per year) > + 0,19 DJF Bottom (°C per year) > + 0,10 DJF

Lake Viverone Lake Constance

Surface (°C per year) > + 0,67 JJA Bottom (°C per year) > + 0,11 JJA Surface (°C per year) > + 0,54 JJA Bottom (°C per year) > + 0,06 JJA

Lake Annecy Wörthersee

Surface (°C per year) > + 0,009 JJA Bottom (°C per year) > + 0,12 JJA Surface (°C per year) > + 0,01 May Bottom (°C per year) > + 0,001 May

Figure 2.2 > Warming summer trend in alpine lakes investigated, showing linear regression lines and their gradients, at the surface and at the bottom. (Data source: see paragraph 2.1).

26 Klopeiner See Ossiacher See

Surface (°C per year) > + 0,02 May Bottom (°C per year) > + 0,03 May Surface (°C per year) > + 0,008 May Bottom (°C per year) > + 0,002 May

Lake Levico Lake Caldonazzo

Surface (°C per year) > + 0,008 July - August Surface (°C per year) > + 0,13 July - August Bottom (°C per year) > + 0,01 July - August Bottom (°C per year) > + 0,07 July - August

27 Results of the Hurst exponent applied to water temperatures data Lake H exp The.Hurst.exponent.has.been.applied.to.the.lakes.investigated. Avigliana 0.79 (Avigliana. Grande,. Sirio,. Viverone,. Constance,. Annecy,. Sirio 0.79 Wörthresee,. Ossiacher. See,. Klopeiner. See,. Caldonazzo. and. Viverone 0.75 Levico). Bodensee 0.84 For. each. one. it. has. been. used. the. water. temperature. data. Annecy 0.92 measured.at.the.bottom.of.the.last.longest.and.homogeneous. time.period.to.show.the.trend.line.in.order.to.verify.the.increase. Klopeiner.See 0.60 or.decrease.of.temperature.on.the.long.time.series..The.choice.of. Ossiacher.See 0.64 the.water.bottom.temperature.data.is.because.the.temperature. Wörthersee 0.76 level.is.homogeneous.for.all.ten.lakes..The.temperature.of.water. Caldonazzo 0.53 bottom.is.normally.around.the.4-5°C.degrees.independently.of. Levico 0..99 the. depth,. while. for. the. water. at. different. depth. and. near. the. surface,. the. temperature. can. be. influenced. by. the. seasonal. MEAN 0.76 conditions.of.the.weather..It.is.the.reason.of.the.choice.to.prefer.to.

consider.the.water.temperature.of.maximum.depth,.because.the. Table 2.2 > results of the Hurst exponent applied to the investigated lakes. variations.are.caused.mainly.by.the.increase.of.air.temperature. due. to. the. Climate. Change.. From. the. analysis. of. the. water. bottom.temperature,.as.showed.in.the.previous.diagrams,.there. is.the.evidence.that.exist.an.increasing.trend.line,.that.shows.an. alarming..increase.for.the.next.decades..But.the.question.is:.is. credible.and.reliable.this.behaviour.of.temperature.or.it.is.only. a.temporary.increase.not.related.with.the.Climate.Change?.The. Hurst.exponent.is.useful.to.reply.to.this.question,.because.it.is.a. robust.method.to.evaluate.the.trends.of.long.time.series....

The. Hurst. exponent. is. used. as. a. measure. of. the. long. term. memory.of.time.series..It.relates.to.the.autocorrelations.of.the. time. series. and. the. rate. at. which. these. decrease. as. the. lag. Figure 2.3 > Hurst exponent values. between.pairs.of.values.increases..Studies.involving.the.Hurst. exponent.were.originally.developed.in.hydrology.for.the.practical. matter. of. determining. optimum. dam. sizing. for. the. Nile. river’s. volatile. rain. and. drought. conditions. that. had. been. observed. The. results. of. the. application. of..H. exp. is. very. different,. over.a.long.period.of.time..The.name.«Hurst.exponent».or.Hurst. compared.between.the.different.lakes...Some.lakes.have.H.exp. coefficient.derives.from.Harold.Edwin.Hurst.(1880–1978),.who. very. high,. as. Caldonazzo. with. H=0.99,. other. lower. as. Levico. was. the. lead. researcher. in. these. studies,. and. the. use. of. the. with.H=0.53..The.value.of.other.eight.lakes.are.in.the.range.of. standard.notation.H.for.the.coefficient.relates.to.this.name.also. Levico.and.Caldonazzo.Hurst.exponent.values..No.one.shows. The.Hurst.exponent.is.referred.to.as.the.«index.of.dependence»,. an. H. exp. smaller. than. 0.5. and. this. means. that. the. variations. or.«index.of.long-range.dependence»..It.quantifies.the.relative. have,.at.different.level,.a.positive.autocorrelation.with.the.past. tendency.of.a.time.series.either.to.regress.strongly.to.the.mean. data.and.the.trend.of.the.increasing.temperature.is.credible. or.to.cluster.in.a.direction. The.variations.of.the.water.bottom.temperature.depends.by.the. Practically: variations. of. temperature. of. the. entire. water. body. due. to. the. •..a.value.in.the.range.0

28 Air Temperature

The. thermal. regime. of. water. bodies. is. mainly. determined. by. meteorological. station,. situated. near. the. investigated.. lakes the.local.weather..The.net.heat.exchange.across.the.air-water. for. Lake. Grande. di. Avigliana,. Lake. Sirio,. Lake. Viverone,. Lake. interface.is.given.by.the.sum.of.energy.fluxes.related.to.radiation,. Annecy,. Lake. Klopeiner. See;. where. these. data. were. not. latent.and.sensible.heat.(Edinger.et.al.,.1968;.Imboden.&.Wüest,. available,.it.has.been.used.for.the.analysis.the.trend.of.long.term. 1995).. A. shift. in. climate. variables. such. as. air. temperature,. data.of.HISTALP.Project.(for.Lakes.Caldonazzo.and.Levico,.Lake. radiation,.cloud.cover,.wind.or.humidity.will.influence.these.heat. Constance,.Lake.Ossiacher.See.and.Lake.Wörthersee)..For.all. fluxes.and.thus.alter.the.heat.balance.of.lakes..Model.studies. lakes.the.name.of.the.station.is.indicated.for.each.lake.under.the. predict.that.lake.temperatures,.especially.in.the.epilimnon,.will. name.of.the.lake,.in.Figure 2.4 and Figure 2.5. increase. with. increasing. air. temperature,. so. that. temperature. profiles,.thermal.stability.and.mixing.patterns.are.expected.to. Observed climate trends: winter change.as.a.result.of.climate.change.(Hondzo.and.Stefan,.1993;. The. long. term. data. (referred. to. lake. Constance,. Wörthersee,. Stefan.et.al.,.1998). Ossiacher.See,.Klopeiner.See,.Lake.Annecy,.lake.Caldonazzo. Analyses.of.long.term.meteorological.data.series.demonstrate. and. lake. Levico). showed. a. positive. trend. in. winter. air. that.such.a.change.has.already.occurred.in.recent.decades.. temperatures..The.Piedmontese.lakes.data.referred.to.a.short. In the following panels period. (about. the. last. two. decades),. so. their. slightly. negative. (Figure 2.4, Figure 2.5),. it. is. showed. the. trend. of. the. air. trend.is.not.relevant.as.indicator.of.a.decrease.of.air.temperature. temperatures. in. the. areas. interested. by. investigated. alpine. in.winter.in.the.last.decades,.and.this.trend.is.in.line.with.the. lakes,.for.the.winter.season.(mean.of.the.values.of.months.of. trend.of.the.last.decade.of.the.other.investigated.alpine.lakes..In. December,.January.and.February).and.for.the.summer.season. In.Figure 2.4.it.is.possible.to.consult.the.t.the.linear.regression. (mean.of.the.values.of.the.months.of.June,.July.and.August). and.the.lake.air.temperature.gradient.for.each.lake.investigated. In.particular.it.has.been.analysed..the.air.temperature.of.local.

Lake Grande di Avigliana (Avigliana station) Lake Sirio (Borgofranco d’Ivrea station)

- 0,16 °C per decade - 0,15 °C per decade

Lake Viverone (Piverone station) Lake Constance – Bregenz station (HISTALP Project)

- 0,51 °C per decade + 0.18 °C per decade

Figure 2.4 > Air temperature winter trend in alpine lakes investigated, showing linear regression lines and their gradients, at the surface and at the bottom. (Data source: see paragraph 2.1).

29 Wörthersee - Klopeiner See Lake Annecy (Cran- Gevrier station) (Klagenfurt – Flugafen station HISTALP Project)

+ 0,05 °C per decade (Min air temperature) + 0,07 °C per decade + 0,4 °C per decade (Max air temperature)

Lake Levico - Lake Caldonazzo (Trento Histalp Station) Ossiacher See (Ossiach station)

+ 0,07 °C per decade + 0,38 °C per decade

Figure 2.4 > Air temperature winter trend in alpine lakes investigated, showing linear regression lines and their gradients, at the surface and at the bottom. (Data source: see paragraph 2.1).

Observed climate trends: summer referred to the last decades (for Lake Grande di Avigliana, lake Viverone and Lake Sirio). In general, surface water temperatures The air temperature summer trend (months of June, July and approximately reflect the behaviour of the regional summer air August) showed an increase in air temperature in all areas temperature, in particular the strong warming from 1970 to 2009 interested by the investigated lakes, both for the long term data (+ 0,5 °C per decade). (lake Constance, Wörthersee, Ossiacher See, Klopeiner See, Lake Annecy, lake Caldonazzo and lake Levico) and for data

30 Lake Grande di Avigliana (Avigliana station) Lake Sirio (Borgofranco d’Ivrea station)

+ 0,59 °C per decade + 0,73 °C per decade

Lake Viverone (Piverone station) Lake Constance - Bregenz station (HISTALP Project)

+ 0,66 °C per decade + 0,08 °C per decade

Lake Annecy Cran- Gevrier station Wörthersee - Klopeiner See (Source: Meteo-France - SILA) (Klagenfurt - Flugafen station HISTALP Project)

Min Air Temperature: + 0,05 °C + 0,05 °C Max Air Temperature: + 0,9 °C

Lake Levico - Lake Caldonazzo (Trento station HISTALP Project) Ossiacher See (Ossiach station)

+ 0,05 °C + 0,28 °C

Figure 2.5 > Air temperature summer trend in alpine lakes investigated, showing linear regression lines and their gradients. (Data source: see paragraph 2.1).

31 Precipitation

Changes in precipitation may vary considerably at local In general, from the analysed data, in the investigated lakes scales, in particular in the alpine space, where there are strong we observed a slight decrease in precipitation in particular in orographic effects. Long term precipitation trends are highly summer season (Figure 2.7). variable between seasons, and have large spatial differences. In the Alpine space have been found weak or non-significant Observed climate trends: winter trends for precipitation in the past century (Beniston, 2005) The long term data (referred to lake Constance, Wörthersee, Precipitation changes may not only be caused by global Ossiacher See, Klopeiner See, lake Caldonazzo and lake warming but may affect global warming as well (i.e., they provide Levico) of HISTALP project showed a negative trend in winter a feedback). precipitation, except for lake Constance (but this trend becomes In particular within SILMAS project we analysed the precipitation negative from 1960 on). The Piedmontese lakes data referred of local meteorological station, situated near the investigated to a short period (about the last two decades), so their slightly lakes for Lake Grande di Avigliana, Lake Sirio, Lake Viverone, negative trend is not relevant as indicator of a decrease of Lake Annecy, Lake Klopeiner See; where precipitation data were precipitation in winter in the last decades, but this trend is in line not available, we used for the analysis the trend of long term with the trend of the last decade of the other investigated alpine data of HISTALP Project (for Lakes Caldonazzo and Levico, lakes. In Figure 2.6 it is possible to consult the linear regression Lake Constance, Lake Ossiacher See and Lake Wörthersee) For and the winter precipitation gradient for each lake investigated. all lakes the name of the station is indicated under the name of the lake, in Figure 2.6 and Figure 2.7).

Lake Grande di Avigliana (Avigliana station) Lake Sirio (Borgofranco d’Ivrea station)

+ 2,18 mm per decade + 16,5 mm per decade

Lake Viverone (Piverone station) Lake Constance - Bregenz station (HISTALP Project)

+ 3,74 mm per decade + 4,20 mm per decade

Figure 2.6 > Precipitation winter trend in alpine lakes investigated, showing linear regression lines and their gradients (Data source: see paragraph 2.1).

32 Lake Annecy Cran- Gevrier station Wörthersee – Klopeiner See - Ossiacher See (Source: Meteo-France – SILA) (Klagenfurt – Flugafen station HISTALP Project)

- 3,8 mm per decade - 1,5 mm per decade

Lake Levico - Lake Caldonazzo (Trento station HISTALP Project)

- 0,92 mm per decade

Figure 2.6 > Precipitation winter trend in alpine lakes investigated, showing linear regression lines and their gradients (Data source: see paragraph 2.1).

Observed climate trends: summer period. (about. the. last. two. decades),. so. their. slightly. negative. trend..is.not.relevant.as.indicator.of.a.decrease.of.precipitation. The. long. term. data. (referred. to. lake. Constance,. Wörthersee,. in. winter. in. the. last. decades,. but. this. trend. is. in. line. with. the. Ossiacher. See,. Klopeiner. See,. lake. Caldonazzo. and. lake. trend.of.the.last.decade.of.the.other.investigated.alpine.lakes..In. Levico).of.HISTALP.project.showed.a.negative.trend.in.winter. Figure 2.7.it.is.possible.to.consult.the.linear.regression.and.the. precipitation,..The.Piedmontese.lakes.data.referred.to.a.short. summer.precipitation.gradient.for.each.lake.investigated.

33 Lake Grande di Avigliana (Avigliana station) Lake Sirio (Borgofranco d’Ivrea station)

- 17.64 mm per decade - 0.6 mm per decade Lake Viverone (Piverone station) Lake Constance - Bregenz station (HISTALP Project)

- 7,25 mm per decade - 0,15 mm per decade Lake Annecy Cran- Gevrier station Wörthersee - Klopeiner See - Ossiacher See (Source: Meteo-France – SILA) (Klagenfurt - Flugafen station HISTALP Project)

+ 9,4 mm per decade - 1,8 mm per decade Lake Levico - Lake Caldonazzo (Trento station HISTALP Project)

- 1,75 mm per decade

Figure 2.7 > Precipitation summer trend in alpine lakes investigated, showing linear regression lines and their gradients (Data source: see paragraph 2.1).

34 Transparency Lake Avigliana Grande Water. column. transparency. is. an. important. feature. of. water. quality. and. has. ecological. implications.. Traditionally.transparency. has. been. measured. with. a. Secchi.disc.. An. increase. in. transparency. indicates. a. reduction. in. productivity.as.a.result.of.the.reduced.vertical.mixing,. due.to.an.increase.in.water.temperatures..The.collected. data. on. alpine. lake. investigated. on. transparency. are. very. poor;. the. gradients. of. the. measured. data. indicated.a.slight.increase.in.transparency,.except.for. Lake.Klopeiner.See..

+ 0,0007 m per year Lake Viverone Lake Sirio

+ 0,000005 m per year + 0,0002 m per year

Lake Annecy Lake Klopeiner See

+ 0,000005 m per year -0,0003 m per year

Lake Ossiacher See Lake Woerthersee

+ 0,0003 m per year + 0,0024 m per year

Figure 2.8 > Trasparency trend in alpine lakes investigated, showing linear regression lines and their gradients (Data source: see paragraph 2.1).

35 Lake Caldonazzo Lake Levico

+ 0,039 per year + 0,031 m per year

Figure 2.8 > Trasparency trend in alpine lakes investigated, showing linear regression lines and their gradients (Data source: see paragraph 2.1).

Oxygen

The concentration of dissolved oxygen [DO, units of milligram increase in deep water temperatures and a simultaneous per litre (mg·L-1)] is perhaps the single most important feature of decrease in deep water oxygen, with important ecological water quality. It is an important regulator of chemical processes consequences (see chapter 3). Oxygen concentrations in lakes and biological activity. Most forms of aquatic life require oxygen can indicate climate shifts because oxygen levels are strongly (DO). For example, certain combinations of low temperature influenced by temperature and thermal structure (Hanson et and high DO concentrations are required for the maintenance al. 2006). For example, the extremely warm European summer of a cold water sport fishery (such as trout and salmon). Plant of 2003 resulted in a long period of thermal stratification and photosynthesis produces oxygen within the region below the increased hypolimnetic oxygen depletion in some Swiss lakes water surface with adequate light (photic zone). Microbial (for (Jankowski et al. 2006). When applicable, hypolimnetic oxygen example, bacteria) respiratory and organic decay processes concentrations have added value as indicators of climate consume oxygen. Near the reservoir surface, oxygen can change because they have widespread consequences for move between the water and air. The rate and direction of this internal nutrient loading (Pettersson et al. 2003), habitat size, exchange is dependent on the wind speed and status of the and refuge availability (; Jansen and Hesslein 2004). In the surface waters with respect to the equilibrium or saturation context of SILMAS project, It has been investigated the summer concentration. Upon the onset of thermal stratification the (months of June, July and August) and winter trend (where the hypolimnion (see thermal stratification) becomes isolated data are available for the months of December, January and from sources of oxygen. DO is widely observed to decrease February) of Oxygen for lakes Avigliana Grande, Viverone, Sirio, progressively in the hypolimnion over the period of stratification, Annecy and Constance). For others, because of a lack of oxygen because the demand for oxygen associated with respiration and data, it has been considered the monthly trend. In figure 2.9 is decay exceeds the sources. This is illustrated by the time plot indicated the specific period of investigation for each lake. In of DO concentration in the hypolimnion. Oxygen demand in the the lakes investigated, there is a common decrease of summer hypolimnion is usually localized at the interface between the lake oxygen at the bottom (Figure 2.9) for the lake Viverone, lake Sirio bottom (sediments) and the overlying water column. This bottom lake Annecy, lake Caldonazzo and Levico, whereas there is an demand is described as sediment oxygen demand (SOD). DO increase of summer oxygen at the bottom in Lake Constance concentrations are widely observed to decrease progressively and a slight increase in Lake Avigliana Grande. with depth in the hypolimnia of mesotrophic and eutrophic lakes as the sediments are approached, because of the localized demand at the bottom and the limited vertical mixing at these depths. DO concentrations in metalimnia and hypolimnia of oligatrophic (low productivity) lakes are often higher than in the overlying epilimnia. These differences are temperature based, as saturation DO values are higher than in the overlying epilimnia. These differences are temperature based, as saturation DO values are higher at lower temperatures of the metalimnia and hypolimnia. DO minima and/or maxima may at times be observed in the metalimnia of certain lakes and reservoirs generally associated with high localized concentrations of plankton (see plot of metalimnetic DO minimum in profile to right).The occurrence of several consecutive mild winters leads to incomplete mixing in such lakes, which results in a gradual

36 Lake Avigliana Grande Lake Viverone

Data not available for Lake Avigliana Grande - DJF

Surface (mg/l per year):+ 0.13 (JJA) Bottom (mg/l per year):+ 0.006 (JJA)

Lake Annecy

Surface (mg/l per year): - 0.06 (JJA); + 0.08 (DJF) Bottom (mg/l per year): -0.07 (JJA); - 0.16 (DJF) Lake Avigliana Grande

Lake Sirio

Data not available for Lake Avigliana Grande - DJF

Surface (mg/l per year): + 0.085 (JJA); + 0.0006 (DJF) Surface (mg/l per year): - 0.04 (JJA) Bottom (mg/l per year): - 0.22 (JJA); + 0.003 (DJF) Bottom (mg/l per year): - 0.0007 (JJA)

Figure 2.9 > Oxygen trend in alpine lakes investigated, showing linear regression lines and their gradients (Data source: see paragraph 2.1).

37 Lake Klopeiner See Lake Ossiacher See

Surface (mg/l per year): - 0.02 (May); - 0.013 (D) Surface (mg/l per year): - 0.02 (May); + 0.02 (D) Bottom (mg/l per year): + 0.003 (May); - 0.001 (D) Bottom (mg/l per year): - 0.04 (May); - 0.02 (D)

Lake Wörthersee Lake Caldonazzo

Surface (mg/l per year): Surface (mg/l per year): - 0.012 (March); + 0.006 (Jul – Aug) Bottom (mg/l per year): Bottom (mg/l per year): + 0.004 (March); - 0.013 (Jul – Aug)

Figure 2.9 > Oxygen trend in alpine lakes investigated, showing linear regression lines and their gradients (Data source: see paragraph 2.1).

38 Lake Levico Lake Constance

Surface (mg/l per year): -0.05 (Mar); + 0.08 (Jul-Aug) Surface (mg/l per year): - 0.010 (JJA); +0.002 (DJF) Bottom (mg/l per year): - 0.009 (Mar); - 0.02 (Jul – Aug) Bottom (mg/l per year): + 0.044 (JJA); + 0.037 (DJF)

39 Phosphorus

Phosphorus has been studied intensively for more than half a In general lake catchment act as filter for phosphorus and can century. Excess phosphorus in lakes causes eutrophication, behave as sinks, attenuators or sources depending on the spatial resulting in fish kills, decreased water quality and decreased water and temporal distribution of phosphorus as well as its speciation. recreation enjoyment. In urban areas the main factor responsible Climatic changes in the air temperature and precipitation can for the increase has been the influx of phosphorus from municipal influence the retention of phosphorus in the catchment area and its and industrial point sources (Forsberg ., 1994); for example since transport in the lake. The physical characteristics of individual lakes the fifties, a significant degradation of Lake Constance state was influence the phosphorus dynamics. Lakes with a short retention observed as a result of the pollutants of more than 1.2 million time are typically more sensitive than lakes with a long retention inhabitants in its catchment area. To counteract this negative time whilst the internal dynamics of phosphorus is very different in development, international cooperation was realized in the isothermal and thermally stratified lakes (Søndergaard et al., 2003). International Commission for the Protection of Lake Constance In the following panels (Figure 2.10) the linear regression and the (IGKB). Thereby the phosphorus concentration in the lake water, gradients of total phosphorus for the investigated lakes have been after a maximum value of 87 mg/m3 phosphorus in 1979, was showed, both for the whole period and, if data are available, for reduced under 10 mg/m3 until 2009. the summer period (months of June, July and August).

Lake Avigliana Grande

Surface: – 0.01 µg/l per year / Bottom: – 0.006 µg/l per year Surface: - 0.4 µg/l per year / Bottom: - 13.8 µg/l per year Lake Viverone

Surface: – 0.005 µg/l per year / Bottom: – 0.009 µg/l per year Surface: - 0.4 µg/l per year / Bottom: - 13.8 µg/l per year Lake Sirio

Surface: + 0.002 µg/l per year / Bottom: + 0.028 µg/l per year Surface: + 3.31 µg/l per year / Bottom: - 2.44 µg/l per year

Figure 2.10 > Total phosphorus trend in alpine lakes investigated, showing linear regression lines and their gradients, at the surface and at the bottom. (Data source: see paragraph 2.1).

40 Lake Annecy

Surface: - 0.0002 µg/l per year / Bottom: - 0.0002 µg/l per year Surface: + 0.09 µg/l per year / Bottom:+ 0.187 µg/l per year

Lake Klopeiner See Lake Ossiacher See

Surface: + 0.0006 µg/l per year / Bottom: + 0.002 µg/l per year Surface: - 0.0013 µg/l per year / Bottom: + 0.0004 µg/l per year

Lake Wörthersee Lake Caldonazzo

Surface: + 0.07 µg/l per year / Bottom: - 0.002 µg/l per year Surface: + 0.003 µg/l per year / Bottom: + 0.058 µg/l per year

Lake Levico Lake Constance

Surface: - 0.02 µg/l per year / Bottom: + 0.03 µg/l per year Surface: - 0.8 µg/l per year / Bottom: - 1.83 µg/l per year

41 Chlorophyll “a”

Chlorophylls are complex molecules found in all photosynthetic to anthropogenic inputs of critical plant nutrients (particularly plants, including phytoplankton (microscopic plants dispersed phosphorus). The concentration of phytoplankton is widely used in the waters). Chlorophyll, contained within the plant’s as an indicator of the level of production of these microscopic cells, allows the plant to utilize sunlight as part of the their plants (e.g., primary production or trophic state). Phytoplankton metabolism. There are several types of chlorophyll identified by biomass is the primary regulator of clarity and colour of water slight differences in their molecular structure and constituents. in many lakes and reservoirs. The most widely used measure of These include chlorophyll a, b, c, and d. Chlorophyll a is phytoplankton biomass is chlorophyll a. Phytoplankton growth the principal photosynthetic pigment and is common to all depends on nutrients and water temperature. Chlorophyll-a phytoplankton. Chlorophyll a can thus be used as a measure of concentrations might increase under reference conditions with phytoplankton biomass. The distribution and concentration of increasing water temperature and nutrients. In the following phytoplankton is of major water quality and ecologic concern. panels (Figure 2.11) is shown the trend of chlorophyll “a” for the Managers are particularly concerned with the occurrences lakes where the data are available. Except for Lake Viverone, for of excess concentrations of phytoplankton and associated other lakes the trend showed is negative. nuisance conditions in surface waters, that occur in response

Lake Avigliana Grande Lake Viverone

Lake Sirio Lake Annecy

Lake Constance

Figure 2.11 > Chlorophyll “a” trend in alpine lakes investigated, showing linear regression lines and their gradients. (Data source: see paragraph 2.1).

42 Phytoplankton - Chlorophyceae

Green.algae,.or.Chlorophyceae,.are.among.the.most.numerous. hydrochemistry.such.an.increase.in.nutrient.availability.(Findlay. and.diverse.of.all.freshwater.algae..At.least.11.orders.of.green. et.al.,.2001)..The.better.light.consitions.during.warmer.winters. algae.are.recognised.and.sometimes.up.to.19.–.depending.on. with. shorter. ice. cover. and. less. snow. promote. phytoplankton. the.author..They.often.comprise.the.majority.of.the.planktonic. growth.in.winter.(Petterson.et.al,.2003). species. of. algae. present. in. healthy. freshwater. ecosystems.. In. deep. alpine. lakes,. the. internal. recycling. of. nutrients. and. Although.some.species.can.form.blooms.at.times.in.nutrient- the. consequently. development. of. phytoplankton. are. strongly. enriched. waters,. none. are. toxic.. The. Chlorophyceae. are. influences. by. the. duration. and. intensity. of. vertical. mixing. in. primarily.a.fresh-water.group,.with.about.90%.of.representatives. winter. and. early. spring. (Salmaso,. 2002,. 2005).. Prolonged. occurring. in. freshwater. environments.. Attached. and. benthic. thermal. stratification,. occurred. in. particular. in. summer,. can. species.are.common.in.many.shallow.streams.and.rivers,.while. influence. hypolimnetic. oxygen. conditions,. dissolved. nutrient. planktonic.species.occur.in.lakes,.reservoirs,.ponds.and.other. concentration. and. phytoplankton. composition. (Wilhem. and. open.water.environments,.as.well.as.in.rivers.and.streams. Adrian,. 2008).. Oxygen. depletion. and. higher. temperatures. Phytoplankton. community. composition. may. be. altered. by. increase. nutrient. release. processes. at. the. sediment. water. changes. in. winter. and. spring. temperatures,. depending. on. interface.(Søndergaard.et.al.,.2003). lake.type.and.location.(Elliott.et.al.,.2006)..Generally,.increased. In.the.lakes.investigated (Figure 2.12),.the.analysed.data.indicate. phytoplankton. productivity. and. biomass. are. correlated. a.slight.decrease.in.Chlorophyceae.for.the.considered.period. with. higher. spring. water. temperatures. as. well. as. change. in.

Lake Avigliana Grande Lake Viverone

Lake Sirio Lake Annecy

Lake Constance

Figure 2.12 > Chlorophyceae trend in alpine lakes investigated, showing linear regression lines and their gradient. (Data source: see paragraph 2.1).

43 Climate Driven Scenarios

The lake ecosystems are extremely sensitive to human impacts The first step in order to evaluate the role of different processes and climate change (Skjelkvaele and Wright, 1998) (at physical, chemical and biological level) and to assess the response of lake ecosystem to climate change at level of lake The hydrochemical characteristics of high altitude Alpine area is necessarily to evaluate climate change in the area lakes and mountain lakes generally in remote areas, far from interested by lakes. In particular, it is necessary to estimate the local pollution sources, are strongly influenced by the effects increase/decrease concerning air temperature and precipitation of atmospheric transport of type long range of pollutants, climate driven. changes in the processes of deposition and general changes in hydrological and thermal regimes. These lakes, influenced Many projects have already addressed to the quantitative more by regional and synoptic scale conditions that the local assessment of likely future evolutions of temperature circumstances can be considered excellent laboratories to and precipitation in the alpine space. In this context determine the effects of global changes and regional freshwater we used the available data from ENSEMBLES project . ecosystems. (www.ensembles-eu.org) and the Multimodel SuperEnsemble method (Krishnamurti T.N. et al., 1999). The lake ecosystems are characterized by numerous feedback mechanisms between biological processes, external forcing and In particular we used the dataset E-obs (Haylock et al., 2008), circulation hydrodynamics, which can be modified by climate a high resolution European land-only daily gridded data set for change and sometimes contradictory effects (IPCC assessment precipitation and mean, minimum, maximum temperature for report, 2001), causing changes in water quality and biodiversity the period 1950 - 2006. and leading to the possible occurrence of «Catastrophic shifts» The Multimodel SuperEnsemble method was applied to the in the ecosystem (Scheffer et al. 2001). A rise in temperature RCMs outputs form ENSEMBLES project to downscale the (Mooij et al. 2005; Johnk et al. 2008) can for example facilitate scenarios over complex terrain regions like Alpine Space. the development of phytoplanktonic bloom due to increased stability of the water column and the dependence of algal growth rate on temperature.

Regional Climatic Scenarios in the Alpine Space

The study of the impact of climate change is essential to predict observed past data with data produced for the future by RCM or its future effects. In this perspective, the main contribution is that are able to optimize intrinsic statistical errors. given by simulation and numerical models for climate prediction, In this sense, the E-obs dataset is the base to create future mathematical tools based on the physical atmospheric laws and climatic scenarios in the european alpine space. their interaction with the Earth surface (especially with the sea), and especially with the future evolution of the human society and Using Multimodel technique, already widely used for weather of greenhouse gas emission scenarios (having anthropological forecasting (Krisnamurty et al. 1999; Dog & Milelli, 2010), it is source) and on the basis on the land use. possible to estimate the errors that control run climate models The greenhouse gas emission scenarios depend strongly on (period: 1961-2000) commit in the description of temperature the future energetic and economic growth of the human activity. and precipitation on the Alpine Space. The most plausible scenario appears here to be the IPCC A1B, Applying the bias and the weights calculated in the control i.e. a quick raise of the world growth, with an increasing use period it is possible to correct the outputs of the models

of energies with low CO2 emissions (nuclear energy, renewable (period 2000-2100), assuming that the bias and the weights energies) and the fast introduction of new efficient technologies. do not have significant differences. You can then obtain more The A1B scenario seems to be the closest from the forecast of accurate scenarios for the Alpine Space region, with the ability the International Energy Agency at the year 2050. to distinguish behaviour of the plains and mountain areas an The numerical climate models, that consider in mathematical estimate of the main indicators predicted climate scenario. terms the implications arising from a specific socio-economic In order to calculate the Multimodel SuperEnsemble, it has scenario, are the following: been used the regional climatic models of European project • general circulation models (GCM General Circulation Model) ENSEMBLES, and in particular the reanalysis on ECMWF with a resolution of about 100 km; ERA-40 (1961-2000) and A1B scenario runs (1961-2100) of the • limited area models (RCM Regional Climate Models), with a following RCMs (daily data): resolution of about 25 km ; • HIRHAM5 - DMI (GCM: Arpege) The RCM derive from the GMC and are useful to assess a limited • REGCM3 - ICTP (GCM: ECHAM5) area. • HadRM3Q0 - Hadley Center (GCM: HadCM3Q0) Despite their high level of detail, the information provided by • RM4.5 - CNRM (GCM: Arpege) RCM can be further detailed, adapted to specific local area (as • CLM - ETH Zurich (GCM: HadCM3Q0) an area interested by a lake), and combined each other through • RACMO2 - KNMI (GCM: ECHAM5) statistical downscaling techniques that are able to correlate the • REMO - Max Plank Institute (GCM: ECHAM5)

44 Multimodel Super Ensemble Technique

As. suggested. by. the. name,. the. Multimodel. SuperEnsemble. The. technique. of. Multimodel. SuperEnsemble. can. not. be. method.(Krishnamurti.T.N..et.al.,.1999).requires.several.model. applied. to. the. values. of. precipitation,. because. the. scenarios. outputs,. which. are. weighted. with. an. adequate. set. of. weights. are.not.necessarily.correlated..to.each..other..and.the.resulting. calculated.during.the.so-called.training.period.. precipitation.are.the.weighted.average.of.correlated.events.(with. an.underestimation.of.the.precipitation.in.the.region). We. applied. this. technique. to. a. wide. number. of. weather. parameters.in.Alpine.region.with.a.very.good.reduction.of.the. It.was.therefore.applied.the.technique.of.Probabilistic.Multimodel. forecast.errors SuperEnsemble.Dressing.developed.by.ARPA.Piemonte.(Cane. &.Milelli,.2010).to.obtain.a.consistent.estimate.of.precipitation.. We.interpolated.the.model.and.control.files.on.the.E-obs.grid.via. Initially.probability.density.of.the.observations.conditioned.to.the. bilinear.interpolation.and.for.each.grid.point.we.compared.the. scenario. distributions. have. been. calculated,. then. the. weights. control.runs.in.the.period.1961-2000.with.the.observations.from. have.been.calculated.with.the.inverse.of.the.Continuous.Ranked. E-obs.dataset..We.then.obtained.the.Multimodel.SuperEnsemble. Probability.Score.in.the.training.period.and.probability.density. weights.with.a.Gauss-Jordan.minimisation. distributions.of.the.Multimodel,.drawing.at.the.end.a.summary. The. unbiased. Multimodel. evaluation. in. the. forecast. period. which.was.used.as.a.single.deterministic.model. (2071-2100).is.then.given.by.the.equation. The. Multimodel. SuperEnsemble. Dressing. reproduces. in. a. strongly.way.the.seasonal.component.of.the.precipitation. N ∑ ( −+= FFaOS iii ) i =1 where. N. is. the. number. of. models,. ai. are. the. SuperEnsemble. weights,.Fi.is.the.forecast.value,.Fi.is.the.mean.forecast.value.in. the.training.period.and.O.is.the.mean.observation.in.the.training. period.

Climate Driven Alpine Scenarios

In. the. following. panels. are. showed. the. air. temperature. and. Air Temperature Scenarios precipitation. scenarios. (25. x. 25. km. resolution). for. the. areas. All.the.climate.change.scenarios.imply.that.the.areas.interested. interested. by. investigated. lakes,. for. the. period. 2001. –. 2050.. by.lakes.in.the.alpine.space.will.become.warmer.(warming.mean. The. scenario. projection. showed,. obtained. with. Multimodel. summer.trend:.+.0,25.°C.per.decade)..The.differences.between. SuperEnsemble,. thanks. to. the. use. of. the. high. resolution. the.sites.are.principally.attributable.to.their.location.(Figure 2.13).. analysis,. allows. a. better. characterization. of. the. temperature. and.precipitation.variations.in.the.alpine.area,.with.differences. between.mountainous.and.plain.regions.

Lake Grande di Avigliana Lake Sirio - Viverone

+ 0,26 °C per decade + 0,27 °C per decade

Figure 2.13 > Air temperature scenario in alpine lakes investigated, showing linear regression lines and their gradients, for the period 2001 - 2050.

45 Lake Wörthersee Lake Constance

+ 0,27 °C per decade + 0,23 °C per decade

Lake Annecy Klopeiner See

+ 0,25°C per decade + 0,25 °C per decade

Lake Caldonazzo - Lake Levico Lake Ossiacher See

+ 0,27°C per decade + 0,23 °C per decade

Figure 2.13 > Air temperature scenario in alpine lakes investigated, showing linear regression lines and their gradients, for the period 2001 - 2050.

Precipitation Scenarios Trentino lakes) will be involved in a decreasing of precipitation; on the other hand, more precipitation will influence Austrian Precipitation is projected to decrease slightly for some higher lakes (Ossiacher See, Klopeiner See and Worthersee) and Lake latitude regions. Warmer surface temperature would speed Constance (respectively affected in the future by the following up the hydrological cycle at least partially, resulting in faster precipitation trends: +1,1, +6,3, +4,8 and +20,4 mm per decade) evaporation and more precipitation in other areas. The lakes (Figure 2.14). located in the Western area (Piedmont Lakes, Annecy Lake,

46 Lake Grande di Avigliana Lake Sirio - Viverone

- 14,5 mm per decade - 17,4 mm per decade

Lake Wörthersee Lake Constance

+ 4,8 mm per decade + 20,4 mm per decade

Lake Annecy Klopeiner See

- 5,4 mm per decade + 6,3 mm per decade

Lake Caldonazzo - Lake Levico Lake Ossiacher See

- 10,4 mm per decade + 1,1 mm per decade

Figure 2.14 > Precipitation scenarios in alpine lakes investigated, showing linear regression lines and their gradients, for the period 2001 - 2050.

47 References

Beniston M (2005). Mountain Climates and Climatic Change: Johnk K.D., Huisman J., Sharples J., Sommeijer B., Visser An Overview of Processes Focusing on the European Alps, Pure P.M. & Stroom J.M. Summer heatwaves promote blooms of and Applied Geophysics 162(8): 1587-1606. harmful cyanobacteria. Global Change Biology, 14, 495-512 (2008). Skjelkvaele BL, Wright RF. Mountain Lakes; Sensitivity to Acid Deposition and Global Climate Change, Ambio. Stockholm Krishnamurti, TN and Kishtawal, CM and LaRow, Timothy [Ambio], vol. 27, no. 4, pp. 280-286 (1998). E and Bachiochi, David R and Zhang, Zhan and Williford, Eric C and Gadgil, Sulochana and Surendran, Sajani (1999) Cane D., Milelli M., “Can a Multimodel SuperEnsemble Improved Weather and Seasonal Climate Forecasts from technique be used for precipitation forecasts?”, Advances in Multimodel Superensemble. In: Science, 285 (5433). pp. 1548- Geoscience, 25, 17-22, 2010. 1550., Forsberg, C. (1994). The large scale flux of nutrients from land Mooij W.M., et al. The impact of climate change on lakes in the to water and the eutrophication of lakes and marine waters. Netherlands: a review. Aquatic Ecology, 39, 381-400 (2005). Marine Pollution Bulletin 29, 409 - 413. Pettersson K, Grust K, Weyhenmeyer G, Blenckner T. Hanson PC, Carpenter SR, Armstrong DE, Stanley EH, Seasonality of chlorophyll and nutrients in Lake Erken-effects of Kratz TK. Lake dissolved inorganic carbon and dissolved weather conditions. Hydrobiologia. 2003;506:75-81. oxygen: Changing drivers from days to decades. Ecol. Monogr. 2006;76:343–363. Hondzo, M. Stefan, H., 1993. Regional watrer temperature characteristics of lakes subjected to climate change . Climatic Haylock, M., Hofstra, N., Klein Tank, A., Klok, L., Jones, P. Change, 24, 187-211. and New, M. 2008 A European daily high-resolution gridded dataset of surface temperature, precipitation and sea-level Stefan, HG, Fang, X., Hondzo, M. 1998. Simulated climate pressure. Journal of Geophysical Research 113, D20119, doi: change effects on year round water temperatures in temperate 10.1029/2008JD010201 zone lakes. Climatic Change, 40, 547-576. IPCC assessment report 2001. Climate Change 2001: Impacts, Edinger, J. E., D. W. Duttweiler, and J. C. Geyer (1968). The Adaptation and Vulnerability, http://www.ipcc.ch/ipccreports/ Response of Water Temperatures to Meteorological Conditions, tar/wg2/262.htm Water Resour. Res., 4(5), 1137-1143. Jankowski T, Livingstone DM, Forster R, Bührer H, Imboden, D. M., Wüest, A. (1995). Mixing Mechanisms in Lakes. Niederhauser P. Consequences of the 2003 European heat In Lerman, A. , Imboden, D. M. , Gat, J. R. (Eds.), Physics and wave for lakes: Implications for a warmer world. Limnol. Chemistry of Lakes (pp. 83-138). Heidelberg: Springer Verlag. Oceanogr. 2006;51:815–819. Jansen W, Hesslein RH. Potential effects of climate warming on fish habitats in temperate zone lakes with special reference to Lake 239 of the experimental lakes area (ELA), north-western Ontario. Environ. Biol. Fish. 2004;70:1–22.

List of figures

Figure 2.1...... xx Figure 2.4...... xx Warming winter trend in alpine lakes investigated, showing Air temperature winter trend in alpine lakes investigated, showing linear regression lines and their gradients, at the surface and at linear regression lines and their gradients, at the surface and at the bottom (Data source: see paragraph 2.1) the bottom. (Data source: see paragraph 2.1)

Figure 2.2...... xx Figure 2.5...... xx Warming summer trend in alpine lakes investigated, showing Air temperature summer trend in alpine lakes investigated, linear regression lines and their gradients, at the surface and at showing linear regression lines and their gradients. (Data source: the bottom. (Data source: see paragraph 2.1). see paragraph 2.1).

Figure 2.3...... xx Figure 2.6...... xx Hurst exponent values Precipitation winter trend in alpine lakes investigated, showing linear regression lines and their gradients (Data source: see paragraph 2.1).

48 Figure 2.7...... xx Figure 2.11...... xx Precipitation.summer.trend.in.alpine.lakes.investigated,.showing. Chlorophyll. “a”. trend. in. alpine. lakes. investigated,. showing. linear. regression. lines. and. their. gradients (Data source: see linear. regression. lines. and. their. gradients.. (Data. source:. see.. paragraph 2.1).. paragraph.2.1).

Figure 2.8...... xx Figure 2.12...... xx Trasparency. trend. in. alpine. lakes. investigated,. showing. Chlorophyceae. trend. in. alpine. lakes. investigated,. showing. linear. regression. lines. and. their. gradients (Data source: see linear. regression. lines. and. their. gradient.. (Data. source:. see.. paragraph 2.1). paragraph.2.1).

Figure 2.9...... xx Figure 2.13...... xx Oxygen. trend. in. alpine. lakes. investigated,. showing. linear. Air.temperature.scenario.in.alpine.lakes.investigated,.showing. regression. lines. and. their. gradients. (Data source: see linear. regression. lines. and. their. gradients,. for. the. period.. paragraph 2.1). 2001.-.2050.

Figure 2.10...... xx Figure 2.14...... xx Total. phosphorus. trend. in. alpine. lakes. investigated,. showing. Precipitation. scenarios. in. alpine. lakes. investigated,. showing. linear.regression.lines.and.their.gradients,.at.the.surface.and.at. linear. regression. lines. and. their. gradients,. for. the. period.. the.bottom. (Data source: see paragraph 2.1). 2001.-.2050.

List of tables

Table 2.1...... 21 Main.charachteristics.of.the.investigated.lakes

Table 2.2...... 22.. Results.of.the.Hurst.exponent.applied.to.the.investigated.lakes.

49 3. A model ecosystem for small mesotrophic and eutrophic sub-alpine lakes

50 Model formulation - single layer and spatially homogeneous conditions

The. model. proposed. here. includes. SIX. compartments,. in. particular,. is. assumed. not. to. limit. remineralization. and. describing.the.concentration.of.nutrients.N.(mainly.phosphorus),. recycling,.constituting.a.large.pool.which.reacts.rapidly.to.the. two. compartments. of. phytoplankton. ( . and. ),. one. availability.of.organic.material..In.this.way,.the.role.of.bacteria.is. P1 P 2 zooplankton. compartment. (Z,. mainly. Daphnia),. planktivorous. parameterized.in.the.nutrient.remineralization.terms.. fish. ( ),. and. piscivorous. predator. fish. ( ).. For. a. general. F C The. (possibly). limiting. nutrient. is. assumed. to. be. phosphorus,. introduction.to.mathematical.models.in.ecology.see.Kot.(2001),. as. is. typical. for. most. freshwater. ecosystems.. We. expect. for.an.introduction.to.lake.ecology.see.Wentzel.(2001).and.for.a. phosphorus.to.be.either.organic.and.contained.in.living.plankton. discussion.of.phytoplankton.ecology.see.Reynolds.(2006)..For. cells,. included. in. the. detritus,. or. available. as. dissolved. or. .models.and.their.use.in.describing.aquatic.ecosystems.see. NPZ colloidal.organic.phosphorus. Bracco.et.al..(2000),.Martin.et.al..(2002),.Pasquero.et.al..(2005),. Koszalka.et.al..(2007).and.references.therein.. Another. important. limiting. factor. is. light. intensity,. I.,. which is.assumed.to.vary.seasonally.and.also.to.be.affected.by.the. All.variables.are.measured.in.terms.of.concentrations,.i.e.,.mass. presence. of. high. phytoplankton. concentrations. which. can. per.unit.volume..The.concentration.of. .is.given.in.phosphorus. N shadow.the.lower.water.layers. content,. while. the. plankton. and. fish. compartments. are. measured.in.carbon.content.(or.biomass,.to.be.decided)...We. We.denote.by. .and. .the.growth.efficiencies.of.the. g1, g2, gZ, gF gC then.use.the.appropriate.conversion.factor.in.the.form.of.fixed. respective.kinds.of.plankton.and.fish,.i.e..the.fraction.of.ingested. P:C (or.P:biomass).ratio.Q.to.convert.from.carbon.(or.biomass). food.(or.nutrient,.for.phytoplankton).used.for.biomass.growth.and. to. phosphorus.. For. simplicity,. we. assume. Q. to. be. the. same. reproduction..The.fractions.1-g.are.instead.used.for.metabolism.. for.all.organisms..Hereinafter,.the.subscript. .and. .refer.to.the. The.parameters. and. .denote.linear.mortality.rates.of. 1 2 m1, m2, mZ mF two.phytoplankton.compartments,.Z.to.the.zooplankton.and F. phytoplantkton,.zooplankton.and.planktivorous.fish.respectively,. and.C.to.two.compartments.of.fish..In.the.model,.the.nutrient.is. and.are.assumed.to.be.constant..Planktivorous.fish.is.assumed. measured.in.μg-P/L,.while.phytoplankton,.zooplankton.and.fish. to. feed. on. zooplankton. and. to. be. subject. to. linear. mortality.. are.measured.in.μg-C/L..Here.P.stands.for.Phosphorus.and.C Piscivorous.fish.feed.on.planktivorous.fish.and.are.affected.by. for.Carbon. quadratic.mortality,.representing.also.the.feeding.of.carnivorous. fish.on.its.same.compartment. We.do.not.explicitely.describe.the.dynamics.of.bacteria,.protozoa,. macroinvertebrates. and. amphibians.. Bacterial. concentration,..

Nutrient

The.dynamics.of.nutrient.concentration.is.described.by The. term. ρ indicates. the. remineralization. rate. of. the. organic. phosphorus,. implicitly. representing. the. activity. of. bacteria... (1.1) Phosphorus. in. the. ecosystem. comes. from. rapid. recycling. in. the. water. column,. in. the. form. of. release. from. plankton. (secretion.and.waste.during.the.growth.process),.and.from.rapid. where.N.denotes.the.concentration.of.bioavailable.phosphorus,. remineralization.of.dead.organic.matter.in.the.water.column.(part. i.e.. orthophosphate. and. other. forms. of. organic. or. inorganic. of.which.is.lost.to.the.sediment)..The.term ρ is.given.explicitly.in. phosphorus. that. may. be. immediately. utilized,. and. I. is. light. equation.(1.9).discussed .below..In.the.model,.this.contribution.is. intensity.. The. Liebig. functions. . are. given. below.. The. added.instantaneously.to.the.P.pool. first. term. on. the. right. hand. side. of. this. equation. describes. phosphorus. uptake. and. the. parameters. . and. . are. the. The.parameter. .is.the.external.influx.of.soluble.phosphorus,.by. V1 V2 Φ maximal.nutrient.uptake.rates.by.phytoplankton.. rainfall,.runoff.or.bottom.resuspension..It.can.be.a.constant.flux. of.nutrient,.independent.of.the.nutrient.level.inside.the.system,. The.carbon.content.of. .and. .is.multiplied.by.the.molar . or.a.relaxation.term.(chemostat.model).of.the.kind P1 P2 P:C ratio.Q.of.phytoplankton,.to.obtain.the.phosphorus.contained. therein.. Stoichiometric. ratios. in. lakes. do. not. generally. obey. Redfield.ratios,.and.can.differ.between.different.species/groups. (e.g.,.Andersen.and.Hessen.1991,.Touratier.et.al..2001,.Vrede.et. al.2002,.Reynolds.2006)..For.simplicity,.here.we.assume.a.fixed. where . is. the. nutrient. level. in. the. chemostat. and. . is. the. N0 1/τ P:C.ratio.of.1/100..Experimentation.with.different.values.do.not. relaxation.rate. lead.to.relevant.variations.in.the.results,.especially.because.the. regeneration.terms.play.a.very.limited.role.in.the.high-nutrient. conditions.of.the.lakes.considered.here.

51 Phytoplankton

The. model. adopted. here. is. developed. for. considering. two. The.parameters .and. .are.the.maximal.nutrient.uptake.rates. V1 V2 different. phytoplankton. compartments,. distinguished. for. of.phytoplankton..The.parameter. .is.the.maximal.grazing.rate. rZ example. by. their. size. (dinoflagellates. versus. diatoms). or. by. of.zooplankton.. their.nutrient.uptake.variability,.or.by.their.response.to.light..As. The.functions.F1, F2.are.Liebig.functions.defined.as discussed.in.the.following,.in.general.only.one.compartment.of. phytoplankton. survives. in. the. homogeneous. model.. However,. we.keep.two.compartments.to.allow.for.future.explorations.of. (1.4) very.different.conditions..

The. parameters. and. are. the. half-saturation. constants. The.equations.for.phytoplankton.dynamics.are κi λi for. nutrient. and. light. respectively.. The. two. phytoplankton. compartments. may. respond. to. lack. of. nutrient. and/or. light. in. (1.2) different.ways.(i.e.,.have.different.values.for..and..). The. parameter. α represents. the. preference. of. zooplankton. (1.3) for. the. . phytoplankton. compartment. and. it. depends. on. P1 phytoplankton.size.and.ability.to.form.colonies.and.on.general. phytoplankton.palatability..If.zooplankton.does.not.consume.P ,. where. . and. . are. the. phytoplankton. concentrations.. In. 1 P1 P2 then. . ..If.zooplankton.does.not.consume. ,.then. .. α =0 P2 α=1 equations. (1.2,1.3),. the. first. term. corresponds. to. growth. of. . phytoplankton. by. nutrient. uptake,. the. second. term. to. loss. of. The. function. ..is. a. generalization. of. the. Holling.. phytoplankton. due. to. consumption. by. zooplankton.. The. third. type-III. functional. form. for. predation. over. two. compartments. term.represents.linear.phytoplankton.mortality,.which.includes. and.it.can.be.written.as both.natural.mortality.and.the.sinking.of.phytoplankton.cells.out. of.the.water.column. (1.5)

Zooplankton

The.dynamical.equation.for.zooplankton.is for. an. introduction. to. predation. functional. types. and. Sarnelle. and. Wilson. 2008. for. the. case. of. Daphnia). and. the. third. term. (1.6) is.linear.zooplankton.mortality..The.parameter. .is.the.maximal. rZ grazing.rate.of.zooplankton.and.the.function.G.has.been.defined. where. the. first. term. on. the. r.h.s.. is. zooplankton. grazing. on. in.eq..(1.5)..The.parameter. .is.the.maximal.grazing.rate.of.fish. rF phytoplankton,. the. second. term. represents. a. Holling. type-III. on.zooplankton.and.η.is.the.saturation.constant.in.the.Holling. predation. of. zooplankton. by. planktivorous. fish. (see. Kot. 2001. type-III.functional.form.

Planktivorous fish

We. assume. that. planktivorous. fish. feed. on. zooplankton... where. the. first. term. represents. fish. grazing. on. zooplankton,. The.dynamical.equation.for.planktivorous.fish.is the.second.term.is.planktivorous.fish.predation.by.piscivorous. fish.and.the.third.term.is.linear.mortality..The.parameter. .is.the. rF maximal.grazing.rate.of.zooplankton.by.planktivorous.fish.and. ..(1.7) r .is.the.maximal.grazing.rate.of.piscivorous.fish,.again.using.a...... C Holling.type-III.functional.form.

Piscivorous fish

The.dynamical.equation.for.the.piscivorous.fish.compartment.is which.includes.a.parametrization.of.piscivorous.fish.predation. on.its.same.compartment.and.mathematically.allows.for.a.finite. value.of.piscivorous.fish.at.equilibrium..The.parameter.r .is.the. ..(1.8) C ...... maximal. grazing. rate. of. planktivorous. fish. by. piscivorous. fish. and. . is. the. quadratic. mortality. parameter. (having. different. dC where.the.first.term.is.the.growth.of.piscivorous.fish.by.grazing. units.with.respect.to.the.linear.mortality.rates)..See.Steele.and. on.planktivorous.fish.and.the.second.term.is.quadratic.mortality,. Henderson.(1992).for.a.discussion.of.quadratic.mortality.

52 Direct nutrient recycling

The.fraction.of.food .that.is.not.used.for.biomass.growth.enters. where. ,. ,. ,. and. . indicate. the. fraction. of. 1-g1 1-g2 1-gz 1-gF 1-gC the.metabolism.of.the.consumer/predator..This.biomass.is.then. resource/prey. which. is. used. for. metabolism, ... egested. (i.e.. respired,. excreted,. or. secreted. from. the. body. indicates. the. fraction. of. dead. biomass. which. is. remineralized. surface).. The. phosphorus. content. of. the. egested. biomass. is. in. the. water. column. before. sinking. into. the. bottom. sediment. assumed. to. be. rapidly. transformed. into. soluble,. bioavailable. and.. ..indicates. the. fraction. of. excreta. which. are. phosphorus.by.the.bacterial.and.enzymatic.activity..This.soluble. remineralized.in.the.water.column. reactive. phosphorus. is. indicated. by. the. term. ρ. and. is. added. instantaneously.to.the.nutrient.compartment,.see.eq..(1.1). . The.term.ρ,.representing.direct.nutrient.recycling,.is.expressed. as.:

(1.9)

Temperature dependence

Where. . is. the. maximum. temperature. attained. by. the. Tmax consider.only.the.temperature.dependence.of.the.phytoplankton. system.for.the.linear.case.and.θ is.the.hald-saturation.constant. growth. rate,. and. use. either. a. linear. or. a. nonlinear. form. of. for. the. nonlinear. dependence.. Previous. results. indicate. the. temperature.dependence,.expressed.as appropriateness.of.the.nonlinear.form.(Thébault.and.Rabouille. 2003).

(1.10)

Seasonal dependence

light,.we.make.the.hypothesis.that.. Notice.that,.in.the.current.formulation.of.the.model,.we.consider. only.the.seasonal.variation.and.we.do.not.take.into.account.the. (1.11) daily. or. inter-annual. variations. in. temperature,. light. intensity. (cloudiness),.nutrient.influx.or.wind.conditions.(associated.with. so.that.light.has.a.minimum. .on.Dec.31.and.a.maximum. . mixing)..For.this.reason,.the.comparison.between.model.outputs. I0 Imax on.June.30,.with.the.agreement.that.the.time.t.in.days.is.t =1.on.. and.observations.described.in.the.following.chapters.should.be. Jan.1st.of.the.first.year.of.the.simulation. taken.as.purely.qualitative.

Temperature.is.expressed.by.a.similar.function,

(1.12) where. δ. is. a. time. delay. due. to. the. fact. that. temperature. lags. light..In.this.way,.temperature.has.an.annual.minimum.on.day.δ.

53 Simulation results for Lake Viverone - homogeneous model

We consider the case with no light absorption by phytoplankton. The parameters used in the simulation are reported in Table 3.1.

Parameter Symbol Value Units

Constant nutrient input Φ 0 μg-P L-1 d-1

-1 Nutrient pool N0 100 μg -P L Nutrient relaxation time τ 60 d

Max light intensity (summer) Imax 1 non dimensional

Min light intensity (winter) I0 0.3 non dimensional

Max temperature (summer) Tmax 25 °C

Min temperature (winter) T0 5 °C Half-sat for temperature θ 25 (nonlinear T dep.) °C Delay between light and temp δ 40 d P-C ratio Q 1/100 μg -P/ μg -C

Half-sat constant for light, 1 λ1 0.4 non dimensional

Half-sat constant for light, 2 λ2 0.4 non dimensional

-1 Half sat constant for phyto 1 κ1 30 μg -P L

-1 Half sat constant for phyto 2 κ2 30 μg -P L

-1 Phyto growth rate, 1 V1 1.2 d -1 Phyto growth rate, 2 V2 1 d Zoo growth rate r 0.4 d-1 Half-sat constant for zooplankton ε 60 μg-C L-1 Pref of zoo for phyto 1 α 0.5 non dimensional Pref of zoo for phyto 2 1-α 0.5 non dimensional

Growth efficiency for phyto 1 g1 0.60 non dimensional

Growth efficiency for phyto 2 g2 0.60 non dimensional

Growth efficiency for zoo gz 0.70 non dimensional -1 Linear mortality phyto 1 m1 1/8 d -1 Linear mortality phyto 2 m2 1/6 d -1 Linear mortality zoo mz 1/40 d Fraction of remin. dead biomass ϒ 0 non dimensional

Fraction of remin. excreta ϒ0 0.8 non dimensional -1 Growth rate, planktivorous fish rF 1/5 d Half-sat constant, plankt. fish η 600 μg-C L-1

-1 Linear mortality, plankt. fish mF 1/300 d

Growth efficiency, plankt. fish gF 0.75 non dimensional -1 Growth rate, piscivorous fish rc 1/10 d Half-sat constant, pisciv. fish ω 1200 μg-C L-1

Growth efficiency, pisciv. fish gc 0.75 non dimensional -1 -1 Quadratic mortality, pisciv. fish dc 0.0001 (μg-C) d

Table 3.1 > Parameter values for the homogeneous model.

54 In.general,.the.results.of.the.simulations.indicated.that: Figure 3.1.shows.the.long-time.(asymptotic).system.behavior.for. 1...The.two. phytoplankton. compartments. do. not. both. survive:. the.parameter.values.reported.in.Table 1..For.these.parameter. either.one.or.the.other.dominates,.depending.on.parameter. values,.a.simple.oscillation.with.annual.periodicity.is.obtained. values. 2...The.values.of.phyto.and.zoo.mortalities.are.quite.important,. as.the.values.of.the.efficiencies. 3...The.preference.of.zoo.for.phyto,.defined.by.α,.is.also.important. in.determining.which.phyto.compartment.survives. 4...There. are. some. differences. between. linear. and. nonlinear. temperature. dependence,. which. however. can. be. modified. by.different.choices.of.the.parameters..We.use.the.nonlinear. temperature.dependence.because.it.is.more.realistic.(Thébault. and..Rabouille.2003). 5...Light.dependence. is. almost. never. active,. and. the. system. dynamics. is. mainly. determined. by. the. temperature. dependence. of. the. phytoplankton. growth. rates.. Light. limitation.may.become.important.only.for.unrealistic.values.of. the.light.dependence.of.the.growth.rates.or.for.unrealistically. low.values.of.light.in.winter. 6...Zooplankton. parameters. are. important. and. must. be. Figure 3.1 > Model carbon concentrations for phytoplankton (solid line), zooplankton (long-dashed line), planktivorous fish (short-dashed line) and determined. piscivorous fish (dotted line), for the parameter values reported in Table 1. All concentrations are expressed in μg-C/L.

Model validation

To. validate. the. model. and. compare. its. output. with. measured. To. express. phytoplankton. concentration. in. units. of. μg-C. we. data,.we.need.to.express.the.model.output.and.the.data.in.the. need. to. use. the. conversion. factor. between. carbon. and. wet. same.units.. biomass,.which.we.fix.as.R=0.2 μg-C / μg-wet-biomass.(Bratbak. and. Dundas. 1984,. Straškrabová. et. al.. 1999).. Thus. we. obtain. In.the.data,.nutrient.concentration.is.expressed.in.μ ,.so.it. g-P/L that.the.phytoplankton.concentration.P.provided.by.the.model. is. directly. comparable. with. the. model. output.. Notice. that. the. (in.μ ).is data. in. general. provide. information. on. the. total. phosphorus. g-C/L concentration,. that. is,. the. phosphorus. which. is. dissolved. P=R* BP =0.2 NP /2 μg-C /L = NP /10 μg-C /L in. the. water. (what. we. called. bioavailable. phosphorus. and. (numerically,.P= N /10)..Of.course,.this.depend.on.the.value.of. enter. the. model. equations). plus. the. phosphorus. content. of. P the.average.radius:.if.we.assume.a.radius.of.about.10.μm.then.it. the. phytoplankton. and. zooplankton. compartments.. Thus,. becomes.(numerically).P ≈ 4/5 N ..If.we.assume.a.radius.of.about. in. the. comparison,. we. have. to. sum. the. different. phosphorus. P 15.μm.then.we.obtain.that.(numerically).P ≈ 2.7 N ..If.we.assume. contributions.from.the.model.(N + Q*P + Q*P +Q*Z).to.obtain. P 1 2 a.radius.of.about 20 μm.then.we.obtain.that.(numerically).P ≈ 6 N .. the.model.TOTAL.phosphorus.which.is.to.be.compared.with.the. P So.the.comparison.critically.depends.on.the.size.assumed.for. data. phytoplankton..Notice,.also, .that.the.ratio.of.carbon-to-biomass. For. biomass,. we. first. need. to. estimate. the. ratio. carbon-to- can. vary. significantly. from. that. adopted. here.. However,.given. biomass.for.phyto-.and.zoo-plankton. the. coarseness. of. our. description. and. the. poor. knowledge. about.the.values.of.the.cell.radii,.we.should.assume.the.chosen. In. the. data,. phytoplankton. and. zooplankton. concentrations. relationship.between. .and. .(which.depend.on.R.and.the.cell. are.expressed.either.in.number.of. .or.as.a.biovolume,.. P NP cells / mL radius).as.purely.indicative.of.an.order-of-magnitude.estimate. mm3/m3.. When. the. biomass. is. expressed. in. number. of. cells. per. mL,. we. need. to. convert. into. biomass. concentration.. Taking. the. When. the. concentration. is. expressed. as. biovolume,. i.e.. in.. phytoplankton. measurement,. ,. in. cells/mL,. if. we. assume. 3 3,. this. is. equivalent. to. 3 3 NP mm /m mg/m = 103 μg/m = 103 μg/1000 phytoplankton.cells.to.have.an.average.radius.of.about.5.μm,.. dm3=μg/L. in. biomass. concentration.. Thus,. the. model. this. gives. a. volume. per. cell. of. about. 500. μm3.. Assuming. concentration. in. μg-C/L..must. be. multiplied. by 1/R to. obtain. the. density. of. water,. ρ=1 kg /L= 109 μg/L=109 μg/dm3= 103 the. biomass. concentration.. For. phytoplankton,. R=0.2 μg-C μg/mm3 = 10-6 μg/μm3,.a.single.cell.with.volume.500.μm3..weights.. / μg-biomass,. while. for. zooplankton R=0.5 μg-C / μg-biomass M=500*10-6 μg=5*10-4μg.. Thus,. if. we. have. NP. cells. per. mL. it. (Straškrabová.et.al..1999). means. we. have. 1000*N cells/L. and. a. total. phytoplankton. wet. biomass.concentration.of

-4 -1 BP(wet)= 1000*N*5*10 μg/L = 5*10 * N μg/L= N/2 μg/L

55 Figure 3.2 shows a comparison between the total phosphorus model at describing specific processes, the fact that we lumped concentration measured in Lake Viverone at a depth of. together all zooplankton species in just one compartment, 10 meters (similar results are found at 5 and 20 meters) and that we ignored zooplankton stage structure or the fact that the phosphorus concentration N produced by the model. Notice zooplankton could feed on other sources. On the other hand, the the agreement between the model output and the data, which data ara available only for one year. Presumably, a longer time display a periodic oscillation with minima in correspondence of series would be required to truly disentangle the zooplankton phytoplankton blooms, as in the model output. population dynamics. 

Figure 3.3 > Comparison between measured chlorophycea concentrations (solid line), assuming cells with approximate equivalent radius of 15 μm, and the phytoplankton Figure 3.2 > Total Phosphorus concentration in μg-P/L from measurements at a depth of 10 meters in concentration produced by the homogeneous model (dashed line). Lake Viverone (solid line) and from the homogeneous model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.2.

Figure 3.3 shows the phytoplankton concentration produced by the homogeneous model, P, together with the total concentration derived from chlorophycea measurements in Lake Viverone, integrated in the euphotic layer. The measured concentration has been multiplied by a factor of 2.5, roughly corresponding to cells with typical radius of about 15 μm. There is a good agreement between the model output and the data in terms of timing of the oscillations, which display an approximately periodic fluctuations with maxima in late summer. However, the model blooms have a much shorter duration than the measured phytoplankton blooms. The date also show a much larger interannual variability, while the model displays a strictly periodic behavior. It remains to be seen whether the irregularity in the data is generated by intrinsic chaotic, rather than periodic, ecosystem dynamics or it is generated by a fluctuating, irregular external forcing (temperature variations, interannual variability in the precipitation and wind regimes, differences in the vertical Figure 3.4 > Comparison between measured zooplankton concentrations, including Rotifera, Copepoda and Cladocera (solid line), and the zooplankton concentration produced mixing of the lake, irregular nutrient input). by the homogeneous model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.5. Figure 3.4 shows a comparison between the zooplankton concentration produced by the model, Z, and the total zooplankton concentration obtained from measurements in Lake Viverone, which include Rotifera, Copepoda and Cladocera. There is a general qualitative agreement in the integrated concentration values but the data remind more of a strongly fluctuating average population rather than an oscillating population with an annual zooplankton concentration peak as in the case of the model. This may point at inadequacies in the

56 Homogeneous model with an alternative set of parameter values

One. of. the. problems. of. the. homogeneous. model. discussed. zooplankton. which. immediately. stops. phytoplankton. blooms.. above. is. the. fact. that. phytoplankton. blooms. -. and. the. This.effect.could.be.reduced.if.we.assume.a.rapid.response.of. corresponding. periods. of. nutirent. depletion. -. are. shorter. planktivorous.fish.to.the.zooplankton..Thus,.we.considered.also. than. in. the. observations.. Looking. into. the. model. functioning,. a.different.parameter.set,.reported.in.Table 3.2. one. realizes. that. this. effect. is. due. to. an. excessive. growth. of.

Parameter Symbol Value Units

Constant.nutrient.input Φ 0 μg-P L-1 d-1

-1 Nutrient.pool N0 120 μg -P L Nutrient.relaxation.time τ 60 d

Max.light.intensity.(summer) Imax 1 non dimensional

Min.light.intensity.(winter) I0 0.3 non dimensional

Max.temperature.(summer) Tmax 20 °C

Min.temperature.(winter) T0 5 °C Half-sat.for.temperature θ 15.(nonlinear.T.dep.) °C Delay.between.light.and.temp δ 60 d P-C.ratio Q 1/100 μg -P/ μg -C

Half-sat.constant.for.light,.1 λ1 0.4 non dimensional

Half-sat.constant.for.light,.2 λ2 0.4 non dimensional

-1 Half.sat.constant.for.phyto.1 κ1 30 μg -P L

-1 Half.sat.constant.for.phyto.2 κ2 30 μg -P L

-1 Phyto.growth.rate,.1 V1 1.4 d -1 Phyto.growth.rate,.2 V2 1 d Zoo.growth.rate r 0.2 d-1 Half-sat.constant.for.zooplankton ε 60 μg-C L-1 Pref.of.zoo.for.phyto.1 α 0.5 non dimensional Pref.of.zoo.for.phyto.2 1-α 0.60 non dimensional

Growth.efficiency.for.phyto.1 g1 0.60 non dimensional

Growth.efficiency.for.phyto.2 g2 0.70 non dimensional

Growth.efficiency.for.zoo gz 1/8 non dimensional -1 Linear.mortality.phyto.1 m1 1/6 d -1 Linear.mortality.phyto.2 m2 1/30 d -1 Linear.mortality.zoo mz 0.3 d Fraction.of.remin..dead.biomass ϒ 0.8 non dimensional

Fraction.of.remin..excreta ϒ0 1 non dimensional -1 Growth.rate,.planktivorous.fish rF 600 d Half-sat.constant,.plankt..fish η 1/300 μg-C L-1

-1 Linear.mortality,.plankt..fish mF 0.75 d

Growth.efficiency,.plankt..fish gF 1/10 non dimensional -1 Growth.rate,.piscivorous.fish rc 1200 d Half-sat.constant,.pisciv..fish ω 0.75 μg-C L-1

Growth.efficiency,.pisciv..fish gc 0.75 non dimensional -1 -1 Quadratic.mortality,.pisciv..fish dc 0.0001 (μg-C) d

Table 3.2 > Alternative parameter values for the homogeneous model. The values differing from Table 1 a have been indicated in bold.

57 The main differences with respect to the previous parameter set is (1) a much higher growth rate of planktivorous fish; (2) a reduced zooplankton growth rate; (3) a slightly larger nutrient input; and (4) maximum temperature at and half-sat Tmax=20 °C constant for temperature at θ=15 °C, values which are closer to observations. Also for these parameter values, the homogeneous model has a periodic solution, with summer phytoplankton blooms that repeat regularly over the years. In this case, however, the blooms last longer and one obtains a better agreement with the observations. Figure 3.5 shows the Phosphorus data together with the model output for nutrient. A goos agreement between data and model results is observed, and the periods of low nutrient conditions now are longer and closer in duration to the observations.

Figure 3.6 > Comparison between measured chlorophycea concentrations (solid line), assuming cells with approximate equivalent radius of 15 μm, and the phytoplankton concentration produced by the homogeneous model with the alternative set of parameter values (dashed line). Concentrations are in μg-C/L assuming a carbon- to-biomass ratio R=0.2.

Figure 3.2 > Total Phosphorus concentration in mg-P/L from measurements at a depth of 10 meters in Lake Viverone (solid line) and from the homogeneous model with the alternative set of parameter values (dashed line).

Figure 3.6 shows the phytoplankton concentration produced by the homogeneous model with the alternative set of parameter values and the total concentration derived from chlorophycea measurements in Lake Viverone, integrated in the euphotic layer. A good agreement in the timing of the bulk phytoplankton increase for the model output and the observations is obtained, and the duration of the model summer phytoplankton bloom is now longer, and closer to that observed in the data. However, the model displays a short intense peak at the beginning of the bloom which is not seen in the data, and the overall shape of the bloom is different between the model and the data. Figure 3.7 shows a comparison between the model zooplankton and the observations. In the model, we have only one compartment of zooplankton, while data allow for a distinction between rotifers, copepods and cladocerans. We can compare the model output either with the total zooplankton biomass, as done above, or with the cladoceran compartment, which is the

one of main interest to planktivorous fish. The two panels of Figure 3.7 > Left panel: Total zooplankton concentration from observations (solid line) and model figure 9 show the model zooplankton compared respectively to zoopankton concentration (dashed line). Right panel: Cladoceran concentration from observations (solid line) and model zooplankton concentration (dashed line). the total zooplankton biomass from data (left panel) and to the Concentrations are in mg-C/L assuming a carbon-to-biomass ratio R=0.5. cladoceran biomass (right panel).

58 Model formulation - two vertical layers

In.this.formulation.we.consider.a.model.system.composed.of. The.two-layer.model.has.seven.compartments,.namely.nutrient. two.vertical.layers,.one.above.the.thermocline.(epilimnion).and. and. phytoplankton. in. the. upper. and. lower. layer,. zooplanton,. one. below. (hypolimnion).. The. upper. layer. is. characterized. by. planktivorous. fish,. and. piscivorous. fish.. The. equations. are. abundant.light.while.the.lower.layer.is.characterized.by.darker. written.as: conditions. Nutrient.is.supposed.to.enter.the.upper.layer.through.a.given. (3.1) external. flux. associated. with. the. input. of. water. streams. and. precipitation. and. leave. the. upper. layer. by. a. flux. associated. (3.2) to. water. flow. out. of. the. lake.. In. this. simplified. formulation,. we. assume. a. given. input. flux. . independent. of. the. nutrient. Φ0 concentration. in. the. lake,. and. an. output. flux. that. depends. (3.3) linearly. on. the. nutrient. concentration. in. the. lake,. −βN.. This. choice.is.based.on.the.assumption.that.there.is.an.independent. (3.4) nutrient.source.from.the.water.input,.and.a.fixed.water.volume. leaving.the.upper.layer.of.the.lake;.thus,.the.loss.of.nutrient.is. proportional.to.the.nutrient.concentration.N. (3.5) From. the. bottom,. there. is. a. nutrient. input. in. the. lower. layer. due.to.decomposition.of.organic.material.in.the.sediment.and. (3.6) nutrient.release.from.the.chemostat.represented.by.the.nutrient- rich. bottom. sediment.. This. is. obtained. by. using. a. relaxation. (3.7) term.of.the.form. .where. .is.the.nutrient. N0 reservoir.in.the.bottom. where.the.indices.i=1,2.refer.to.the.populations.respectively.in. There. is. an. additional. nutrient. mixing. term. between. the. two. the.upper.and.lower.layer,. layers,.expressed.as. .for.the.upper.layer.and. μ (N2-N1) μ (N1-N2) for.the.lower.layer,.associated.with.turbulent.exchanges.between. (3.8) the. two. layers.. Mixing. is. assumed. to. vary. seasonally,. from. a. maximum.μ .in.winter.to.a.minimum.in.summer..For.simplicity,. 0 (3.9) we. assume. that. the. mixing. intensity. varies. sinusoidally. with. time,.with.a.delay.δ .with.respect.to.the.temperature.variation. 1 (3.10) We. assume. that. the. excreta. produced. by. the. methabolism. and. .and. .are.respectively.the.temperature.in.the.upper. and.the.dead.organic.material.sink.to.the.bottom..There,.it.will. T1, T2, I1 I2 participate.in.the.nutrient.reservoir.of.the.bottom.which.re-enters. and. lower. layer. and. the. light. intensity. in. the. upper. and. lower. layer..For.simplicity,.we.assume. ..where. ,. . the.lower.layer.by.the.relaxation.term. I2=ξ I1 ξ =exp(-H/D) H is.related.to.the.depth.of.the.epilimnion.and.D.is.related.to.the. Phytoplankton. is. split. into. two. populations,. P1. and. P2,. living. extinction.depth.of.light.penetration..The.seasonal.variation.of. respectively. in. the. upper. and. lower. layer.. The. parameters. of. light.intensity,.temperature.and.mixing.rate.are.written.as the.two.populations.are.the.same,.but.the.temperature.and.the. light.level.of.the.two.layers.are.different..One.can.also.include.a. (3.11) mixing.term.between.the.two.populations,.i.e..a.coupling.term.of. the.form.μ .for.the.upper.layer.and.μ .for.the.lower. 0(P2-P1) 0(P1-P2) (3.12) layer,. due. to. turbulent. mixing. and/or. phytoplankton. vertical. movements.between.the.two.layers..Here.we.assume.that.only. (3.13) turbulent.mixing.is.active,.and.thus. . μP=μ Zooplankton.are.supposed.to.be.able.to.freely.move.between. the.two.layers.and.to.feed.on.phytoplankton,.wherever.they.are.. Although.visual.predators.may.have.a.preference.for.the.upper. layer.where.light.is.more.abundant.and.there.is.a.diurnal.cycle. of. zooplankton. predation,. we. shall. ignore. these. effects. here.. Consequently,.only.one.compartment.of.zooplankton.is.kept.in. Figure 3.7 > Left panel: Total zooplankton concentration from observations (solid line) and model the.model..Analogous.behavior.is.assumed.for.fish. zoopankton concentration (dashed line). Right panel: Cladoceran concentration from observations (solid line) and model zooplankton concentration (dashed line). Concentrations are in mg-C/L assuming a carbon-to-biomass ratio R=0.5.

59 Simulations for Lake Viverone - Two layer model

Table 3.3 reports the values of the parameters for the two-layer model.

Parameter Symbol Value Units

-1 -1 Nutrient input, upper layer Φ0 0.01 μg -P L d Nutrient loss rate, upper layer β 1/30 d-1

-1 Nutrient pool at the bottom N0 200 μg -P L Nutrient relaxation time (bottom) τ 30 d

-1 Maximum turbulent exchange (winter) μ0 1/10 d -1 Minimum turbulent exchange (summer) μ1 1/720 d

Max light intensity (summer) Imax 1 non dimensional

Min light intensity (winter) I0 0.3 non dimensional Light intensity extinction ratio ξ 0.3 non dimensional

Max temperature (summer, upper) T1,max 20 °C

Min temperature (winter, upper) T1,0 5 °C

Max temperature (summer, lower) T2,max 10 °C

Min temperature (winter, lower) T2,0 5 °C Half-sat for temperature θ 15 °C Delay between light and temperature δ 60 d

Delay between temperature and mixing δ1 20 d P-C ratio Q 1/100 μg -P/ μg -C Half-sat constant for light, 1 λ 0.4 non dimensional Half sat constant for phyto κ 30 μg -P L-1 Phyto maximum growth rate V 1.4 d-1 Zoo growth rate r 0.2 d-1 Half-sat constant for zooplankton ε 60 μg-C L-1 Growth efficiency for phyto 1 G 0.60 non dimensional

Growth efficiency for zoo gZ 0.70 non dimensional Linear mortality phyto M 1/8 d-1

-1 Linear mortality zoo mZ 1/30 d -1 Growth rate, planktivorous fish rF 1 d Half-sat constant, plankt. fish η 600 μg-C L-1

-1 Linear mortality, plankt. fish mF 1/300 d

Growth efficiency, plankt. fish gF 0.75 non dimensional -1 Growth rate, piscivorous fish rC 1/5 d Half-sat constant, pisciv. fish ω 1200 μg-C L-1

Growth efficiency, pisciv. fish gC 0.75 non dimensional -1 -1 Quadratic mortality, pisciv. fish dC 0.0001 (μg-C) d

Table 3.3 > Parameter values for the two-layer model.

60 The.upper.panel.of.Figure 3.8.reports.the.dynamics.of.the.two. Figure 3.10.shows.a.comparison.between.the.upper-layer.model. layer.model,.where.we.show.phytoplankton.in.the.upper.layer,. phytoplankton.and.the.observations.in.Lake.Viverone..A.good. zooplankton,.and.the.two.fish.compartments..The.lower.panel. correspondence. between. model. output. and. data. is. obtained,. of.figure.8.shows.phytoplankton.in.the.upper.and.lower.layer. and. the. model. phytoplankton. bloom. now. lasts. for. a. longer. period,.with.a.closer.agreement.with.the.observations.

Figure 3.10 > Phytoplankton concentration from measurements in lake Viverone (solid line) and upper-layer phytoplankton from the two-layer lake ecosystem model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.2.

Figure 3.11. shows. the. comparison. between. the. model. zooplankton. and. the. observations. in. Lake. Viverone,. for. the. cladoceran. compartment. (upper. panel). and. for. the. total. zooplankton. concentration. (Rotifera. plus. Copepoda. plus. . Cladocera).. A. reasonable. agreement. is. found. between. model. zooplankton.and.cladoceran.concentration,.while.the.agreement. with.the.total.zooplankton.data.is.poorer.

Figure 3.8 > Upper panel: Phytoplankton in the upper layer (solid curve), zooplankton (long-dashed curve), planktivorous fish (short-dashed curve) and piscivorous fish (dotted curve) for the two-layer lake ecosystem model. Lower panel: Phytoplankton in the upper layer (solid curve) and in the lower layer (dashed curve) for the two-layer lake ecosystem model.

Figure 3.9 shows.a.comparison.between.the.total.Phosphorus. concentration. produced. by. the. model. and. observations. from. Lake.Viverone..A.good.agreement.between.model.and.data.is. obtained,.both.in.the.duration.of.the.low.summer.Phosphorus. conditions. and. in. the. general. shape. of. the. total. Phosphorus. variability.

Figure 3.11 > Upper panel: Total zooplankton concentration from observations (solid line) and model zoopankton concentration (dashed line). Lower panel: Cladoceran concentration61 Figure 3.9 > Total Phosphorus concentration in μg-P/L from measurements in lake Viverone from observations (solid line) and model zooplankton concentration (dashed line). (solid line) and from the two-layer lake ecosystem model (dashed line). Concentrations are in mg-C/L assuming a carbon-to-biomass ratio R=0.5. Effects of climate and environmental change

To estimate the response of the lake ecosystem to climate changes. Owing to its flexibility and overall better performances, and/or environmental changes, we can vary some of the key for this study we consider only the two-layer model. parameters of the model and analyze how the model behavior

Temperature

Temperature enters the lake ecosystem dynamics in various layer. In the model, the result of the warming is an anticipation ways. Here, we consider only the temperature dependence of of the phytoplankton and of the zooplankton blooms, and a net the phytoplankton growth rate. increase of the phytoplankton biomass. Figure 3.12 illustrates the model behavior for the higher temperatures, compared with To illustrate the possible effects of an intense climate the model behavior for the temperatures assumed valid today. warming, we assume a (rather extreme) temperature increase of about 3°C, that is, we assume that the minimum seasonal Interestingly, a similar result is obtained also when increasing water temperature in mid February is 8°C instead of the 5°C only the winter temperature and, to slightly lesser extent, only approximately observed today, and the maximum temperature the summer temperature. in summer is 23 °C in the surface layer and 13°C in the bottom

Figure 3.12 > Comparison between the two-layer model dynamics for warmer conditions (solid lines) and model dynamics for current conditions (dashed lines). Left panel: phytoplankton concentrations in the upper layer; right panel: zooplankton concentration. All concentrations are expressed in μg-C/L.

Nutrient relaxation rate

Changing the nutrient relaxation time with the bottom reservoir, τ, leads to moderate changes in the maximum amplitude of the phytoplankton, with no change in the timing of the bloom and minor effects on zooplankton and fish concentrations. . Figure 3.13 shows the phytoplankton concentrations for three values of the bottom relaxation times, namely τ = 15 days, . τ = 30 days and τ = 60 days. Analogous results are found for variations in the export rate of nutrient from the surface layer, β. A significant (eg, two-fold) increase in the value of β (that is, a faster export) makes the phytoplankton bloom less intense, while decreasing β makes the phytoplankton bloom slightly more intense. On the other hand, the model is rather insensitive to variations in the surface nutrient input, Φ, as most of the nustrient comes from the lower layer by turbulent relaxation.

Figure 3.13 > Comparison between the two-layer model dynamics for faster relaxation to the bottom nutrient pool, τ = 15 days (solid line), for the standard value of the relaxation 62 time, τ = 30 days (long-dashed line) and for slower relaxation, τ = 60 days (short-dashed line). All concentrations are expressed in μg-C/L. Turbulent exchange rate

Variations.in.the.(small).summer.turbulent.exchange.rate.between. the.upper.and.the.lower.lake.layers.do.not.lead.to.significant. changes. in. the. model. ecosystem. dynamics.. Conversely,.

changes.in.the.winter.turbulent.exchange.rate.μ0.lead.to.changes. in.the.intensity.of.the.phytoplankton.bloom..Figure 3.14.shows. the.phytoplankton.concentration.for.three.values.of.the.winter.

turbulent.exchange.rate.μ0.between.the.surface.layer.and.the. lower.layer.of.the.lake..

Figure 3.14 > Comparison between the two-layer model dynamics for faster winter turbulent relaxation between the two layers, τ = 5 days (solid line), for the standard value of the relaxation time, τ = 30 days (long-dashed line) and for slower relaxation, τ = 60 days (short-dashed line). All concentrations are expressed in μg-C/L.

A comment on the use of one- or two-layer models

Although.a.quantitative.comparison.between.the.performances. results,.albeit.not.in.a.dramatic.way.with.respect.to.the.one-layer. of.the.one-layer.and.two-layer.models.is.beyond.the.goal.of.this. model..Both.models,.in.addition,.had.troubles.in.reproducing.the. report,. the. results. reported. above. suggest. that. the. two-layer. details.of.the.observed.dynamics.and.do.display.a.significant. model.is.more.flexible.and.could.presumably.provide.more.useful. dependence.on.the.parameter.values.

Application of the lake ecosystem models to other lakes

In.the.following.we.discuss.the.application.of.the.homogeneous. In. the. following. we. considered. the. period. 2002-2007. for. and.of.the.two-layer.lake.ecosystem.models.to.other.lakes.for. comparison.with.phytoplankton. which.measurements.are.available. Compared.to.Lake.Viverone,.Lake.Avigliana.Grande.has.a .much. Simulation results for Lake Avigliana Grande earlier.phytoplankton.bloom.and.a.slightly.lower.nutrient.input.. In.addition,.the.number.densities.of.Chlorophyta.cells.are.larger. The.following.data.are.available:. than.for.Lake.Viverone..Presumably,.we.can.assume.a.slightly. smaller.average.size.of.the.phytoplankton.cells. Measurements Period Units Homogeneous.model from.22/02/2001.. Phosphorus.. μg/l We.use.the.same.parameter.values.reported.in.Table 1b,.with.the. to.16/01/2008 exceptions.reported.below: From.the.data,.we.find.that.Phosphorus.levels.in.Lake.Avigliana. from.19/02/2002.. Grande.are.lower.than.for.Lake.Viverone,.and.we.set.the.nutrient. to.11/12/2007 3 3 pool.at. -1. Phytoplankton. mm /m and N0=100 μg -P L . We.assume.a.smaller.growth.rate.for.planktivorous.fish,.which.is. from.11/03/2009.. cells/mL now.set.at. -1. to.03/11/2009 rF = 0.5 d We.assume.no.delay.between.light.and.temperature,.that.is, δ=0. Zooplankton not.available We.assume.phytoplankton.cells.with.size.10 μm.

63 Figure 3.15 shows the Phosphorus concentration for Lake Figure 3.17 shows the Phosphorus concentration for Lake Avigliana and the homogeneous model output. The overall Avigliana, comparing the two-layer model output with the dynamics is well reproduced. observed data. A good agreement between the two curves is obtained.

Figure 3.15 > Avigliana Grande: Total phosphorus concentration in μg-P/L from measurements at a depth of 10 meters (solid points and connecting line) and from the homogeneous model (dashed line). Figure 3.17 > A vigliana Grande: Total phosphorus concentration in μg-P/L from measurements at a depth of 10 meters (solid points and connecting line) and from the two-layer model (dashed line).

Figure 3.16 shows the phytoplankton concentration from data, assuming cells with size 10 μ , and from the homogeneous m Figure 3.18 shows the phytoplankton concentration from data, model. In general, the model blooms seem to have a slight delay assuming cells with size μ , and from the two layer model. with respect to the observed ones. 10 m

Figure 3.18 > Avigliana Grande: Measured chlorophycea concentrations (solid points and Figure 3.16 > Avigliana Grande: Measured chlorophycea concentrations (solid points and connecting line) and phytoplankton concentration produced by the two-layer model connecting line) and phytoplankton concentration produced by the homogeneous (dashed line). Concentrations are in μg-C/L using a carbon-to-biomass ratio R=0.2 and assuming cells with size model (dashed line). Concentrations are in μg-C/L using a carbon-to-biomass ratio 10 μm. R=0.2 and assuming cells with size 10 μm. For phytoplankton, the agreement between the data and the model output is less satisfactory. By assuming slightly smaller Two layer model phytoplankton cells, we could obtain a closer agreement in the typical values of the concentrations. However, the phytoplankton We use the same parameter values reported in Table 2, with the data display a huge, irregular variability which is not found in the following exceptions: model. It is curious, however, that the nutrient measurements For the bottom nutrient reservoir, we use N =100 μg -P L-1. 0 do not dispay such an irregular variability and, for nutrient, the The maximum growth rate of phytoplankton is fixed at -1. 1.8 d model output is rather close to the data. The maximum growth rate of planktovorous fish is fixed at 0.5 d-1. The delays between light and temperature and between temperature and mixing are kept as for Lake Viverone.

64 Effects.of.temperature.and.environmental.changes Simulation results for Lake Sirio Figure. 3.19. shows. the. effects. of. increasing. temperatures. by.. The.following.data.are.available:. 3. -°C. and. increasing. the. winter. turbulent. exchange. rate. to.. 1/5 day-1.on.the.two-layer.model.dynamics.for.Lake.Avigliana..In. Measurements Period Units this.case,.warmer.conditions.and.more.intense.winter.turbulent. exchange.lead.to.a.very.slight.anticipation.of.the.bloom,.with.a. from.10/01/2001. Phosphorus.. μg/l smaller.maximum.concentration.but.with.a.significantly.longer. to.27/10/2009.. period.of.high.phytoplankton.concentrations..On.the.other.hand,. for. the. zooplankton. one. observes. a. significant. anticipation. Phytoplankton. from.13/05/2002.. . to.26/10/2009 cells/mL of. the. period. of. large. abundance. and. a. slight. increase. in. the. average. zooplankton. concentration.. Note. that. the. results. Zooplankton not.available reported.here.should.be.taken.as.purely.indicative.and.they.refer. only.to.the.response.of.the.model.adopted.here..These.results. should.be.seen.more.as.a.test.on.the.parameter.sensitivity.of.the. In. the. following. we. consider. the. period. 2002-2009. for. the. model.than.a.projection.on.the.future.behavior.of.the.real.lake. comparison.with.phytoplankton. ecosystem. Similarly. to. Avigliana,. from. the. data. we. see. that. Phosphorus. levels.in.this.lake.are.lower.than.in.Lake.Viverone. Homogeneous.model In.the.following.comparison,.we.use.the.same.parameter.values. as.in.Table.1b,.with.the.following.exceptions. From.the.data,.we.find.that.Phosphorus.levels.in.Lake.Sirio.are. lower. than. for. Lake. Viverone,. and. we. set. the. nutrient. pool. at. -1. N0=70 μg -P L We.assume.a.smaller.growth.rate.for.planktivorous.fish,.which.is. now.set.at. -1. rF = 0.5 d We.assume.phytoplankton.cells.with.size.10.μm.

Figure 3.20.shows.the.Phosphorus.concentration.for.Lake.Sirio,. comparing. the. output. of. the. homogeneous. model. with. the. observed.data..A.good.agreement.between.the.two.curves.is. obtained.

Figure 3.19 > Avigliana: Upper panel: Comparison between the two-layer model phytoplankton concentration in the upper layer for faster winter turbulent exchange between the two layers, -1, and warmer summer and winter conditions (+3 °C) (solid line) and Figure 3.20 > Sirio: Total phosphorus concentration in μg-P/L from measurements at a depth of μ0=0.2 day current conditions (dashed line). Lower panel: the same for zooplankton concentra- 10 meters (solid points and connecting line) and from the homogeneous model tion. Concentrations are in μg-C/L. (dashed line).

65 Figure 3.21.shows.the.phytoplankton.concentration.from.data,. Figure 3.23.shows.the.phytoplankton.concentration.from.data,. assuming. cells. with. size. 10 μm,. and. from. the. homogeneous. assuming.cells.with.size.10 μm,.and.from.the.two.layer.model. model..The.timing.of.the.blooms.is.well.captured,.although.the. data.display.larger.variability.than.the.model.output.

Figure 3.23 > Sirio: Comparison between measured chlorophycea concentrations (solid points and connecting lines) and the phytoplankton concentration produced by the Figure 3.21 > Sirio: Comparison between measured chlorophycea concentrations (soild points and homogeneous model (dashed line). Concentrations are in μg-C/L assuming a connecting lines) and the phytoplankton concentration produced by the homoge- carbon-to-biomass ratio R=0.2. The concentration was derived from the number neous model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-bio- concentration assuming an equivalent cell radius of 10 μm. mass ratio R=0.2. The concentration was derived from the number concentratation assuming an equivalent cell radius of 10 μm.

Effects.of.climate.and.environmental.changes Two .layer.model Since. the. parameter. values. are. the. same. used. for. Lake. We.use.the.same.parameter.values.reported.in Table 2,.with.the. Avigliana,.the.response.of.the.two-layer.model.to.a.3.°C.increase. following.exceptions: in.temperatures.and.a.stronger.turbulent.exchange.between.the. For.the.bottom.nutrient.reservoir,.we.use. μ -1. N0=100 g -P L two.layers..is.the.same.reported.in figure A5. The.maximum.growth.rate.of.phytoplankton.is.fixed.at.1.8 d-1. The.maximum.growth.rate.of.planktivorous.fish.is.fixed.at.0.5 d-1. The. delays. between. light. and. temperature. and. between. Simulation results for Lake Levico temperature.and.mixing.are.kept.as.for.Lake.Viverone. The.following.data.are.available:. That.is,.for.Lake.Sirio.we.use.the.same.parameters.already.used. for.Lake.Avigliana.Grande. Measurements Period Units Figure 3.22.shows.the.Phosphorus.concentration.for.Lake.Sirio,. from.2000.to.2007. comparing.the.two-layer.model.output.with.the.observed.data.. and.in.2009,.only.. A.good.agreement.between.the.two.curves.is.obtained. Phosphorus.. 2.measurements/year,.. μg/l 6.measurements.. in.2008.

6.measurements.. .(NB:./liter). Phytoplankton. cells/L in.2008,.2.in.2009,.. and.biovolume . 4.in.2010 (mm3/m3)

Zooplankton not.available

Lake. Levico. has. a. much. lower. sampling. frequency. than. the. other.lakes.considered.here,.and.we.compare.Chlorophytes.in. the.period.2008.to.2010. This.lake.is.characterized.by.rather.low.nutrient.concentration. (compared. to. Lakes. Viverone,. Avigliana. and. Sirio). and. a. correspondingly.low.phytoplankton.biomass. Homogeneous.model

Figure 3.22 > Sirio: Total phosphorus concentration in μg-P/L from measurements at a depth of N =100 μg -P L-1. 10 meters (solid points and connecting line) and from the two-layer model (dashed line). 0 In.the.following.comparison,.we.use.the.same.parameter.values. as.in.Table.1b,.with.the.following.exceptions. From.the.data,.we.find.that.Phosphorus.levels.in.Lake.Sirio.are. much.lower.than.for.Lake.Viverone,.and.we.set.the.nutrient.pool. at. 1. N0=30 μg -P L- We.assume.a.smaller.growth.rate.for.planktivorous.fish,.which.is. now.set.at. -1. rF = 0.5 d 66 We. assume. a. larger. maximum. temperature. in. summer,.. We. assume. a. larger. maximum. temperature. in. summer,.. ,. as. suggested. by. the. available. data (see ,. as. suggested. by. the. available. data. (see Tmax=28°C Tmax=28°C http://www.meteolevicoterme.it/Temperature-Lago.aspx). http://www.meteolevicoterme.it/Temperature-Lago.aspx). We. assume. a. delay. δ=0. between. the. seasonal. variation. of. We.assume.phytoplankton.cells.with.size 10 μm. temperature.and.light. We.assume.phytoplankton.cells.with.size.10 μm. Figure 3.26. shows. the. Phosphorus. concentration. for. Lake. Levico,.comparing.the.two-layer.model.output.with.the.observed. Figure 3.24. shows. the. phosphorus. concentration. for. Lake. data..A.good.agreement.between.the.two.curves.is.obtained. Levico,.comparing.the.output.of.the.homogeneous.model.with. the.observed.data..A.good.agreement.between.the.model.output. and.the.scattered.available.data.is.obtained.

Figure 3.26 > Levico: Total phosphorus concentration in μg-P/L from measurements at a depth of 10 meters observed (solid points) and from the two-layer model (dashed line).

Figure 3.24 > Levico: Total phosphorus concentration in μg-P/L from measurements at a depth of 10 meters observed (solid points) and from the homogeneous model (dashed line). Figure 3.27.shows.the.phytoplankton.concentration.from.data,. assuming.cells.with.size.10 μm,.and.from.the.two-layer.model.. Figure 3.25.shows.the.phytoplankton.concentration.from.data,. It. is. difficult. to. compare. data. and. model. output. owing. to. the. assuming. cells. with. size. 10 μm,. and. from. the. homogeneous. very.few.data.available..However,.the.general.range.of.values. model..It.is.difficult.to.compare.data.and.model.output.owing. observed.is.consistent.with.the.model.output. to. the. very. few. data. available.. However,.the. general. range. of. values.observed.is.consistent.with.the.model.output.

Figure 3.27 > Levico: Comparison between measured chlorophycea concentrations (solid points) and the phytoplankton concentration produced by the two-layer model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.2. Figure 3.25 > Levico: Comparison between measured chlorophycea concentrations (solid points) The concentration was derived from the cell number concentration assuming an and the phytoplankton concentration produced by the homogeneous model (dashed equivalent radius of 10 μm. line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.2. The concentration was derived from the cell number concentration assuming an equivalent radius of 10 μm.

Two .layer.model We.use.the.same.parameter.values.reported.in.Table 2,.with.the. following.exceptions: For.the.bottom.nutrient.reservoir,.we.use . N0=50 μg -P L-1 The.maximum.growth.rate.of.phytoplankton.is.fixed.at.1.8 d-1. The.maximum.growth.rate.of.planktivorous.fish.is.fixed.at.0.5 d-1. We. assume. a. delay. δ=0. between. the. seasonal. variation. of. temperature.and.light. 67 Effects.of.temperature.and.environmental.changes Simulation results for Lake Caldonazzo Figure 3.28. shows. the. effects. of. increasing. temperatures. by.. The.following.data.are.available:. 3°C. and. increasing. the. winter. turbulent. exchange. rate. to. 1/5. day-1. on. the. two-layer. model. dynamics. for. Lake. Levico.. Measurements Period Units In. this. case,. warmer. conditions. and. more. intense. winter. turbulent. exchange. lead. to. anticipation. of. the. phytoplankton. Phosphorus.. from.2000.to.2009 μg/l and. zooplankton. blooms,. with. a. slightly. smaller. maximum. .(NB:./liter). concentration.for.phytoplankton..On.the.other.hand,.the.model. cells/L Phytoplankton. from.2008.to.2010 and.biovolume fish. abundances. display. significant. increase,. as. shown. in. the. 3 3 figure. for. the. planktivorous. fish. compartment.. Note. that. the. (mm /m ) results.reported.here.should.be.taken.as.purely.indicative.and. Zooplankton not.available they. refer. only. to. the. response. of. the. model. adopted. here.. These.results.should.be.seen.more.as.a.test.on.the.parameter. sensitivity.of.the.model.than.a.projection.on.the.future.behavior. Lake.Caldonazzo.has.a.low.sampling.frequency.and.a.limited. of.the.real.lake.ecosystem. period. for. phytoplankton;. we. compare. Chlorophytes. in. the. period.2008.to.2010.and.nustrients.in.the.period.2000-2009. This.lake.is.characterized.by.rather.low.nutrient.concentration. (compared. to. Lakes. Viverone,. Avigliana. and. Sirio). and. a. correspondingly.low.phytoplankton.biomass. Homogeneous.model. In.the.following.comparison,.we.use.the.same.parameter.values. as.in.Table 1b,.with.the.following.exceptions.(we.use.the.same. parameter.values.as.for.Lake.Levico). From.the.data,.we.find.that.Phosphorus.levels.in.Lake.Sirio.are. much.lower.than.for.Lake.Viverone,.and.we.set.the.nutrient.pool. at. -1. N0=30 μg -P L We.assume.a.smaller.growth.rate.for.planktivorous.fish,.which.is. now.set.at. -1. rF = 0.5 d We. assume. a. larger. maximum. temperature. in. summer,.. . Tmax=28 °C We. assume. a. delay. δ=0. between. the. seasonal. variation. of. temperature.and.light. Figure 3.29 shows. the. phosphorus. concentration. for. Lake. Levico,.comparing.the.output.of.the.homogeneous.model.with. the. observed. data.. A. good. agreement. between. the. model. output.and.the.available.data.is.obtained...

PB visuel

Figure 3.29 > Caldonazzo: Total phosphorus concentration in μg-P/L from measurements at a depth of 10 meters observed (solid points) and from the homogeneous model (dashed line).

Figure 3.28 > Levico: Upper panel: Comparison between the two-layer model phytoplankton concentration in the upper layer for faster winter turbulent exchange between the two layers, -1, and warmer summer and winter conditions (+3 °C) (solid line) and μ0=0.2 day current conditions (dashed line). Mid panel: the same for zooplankton concentration. Lower panel: The same for the planktivorous fish abundance. Concentrations are in μg-C/L.

68 Figure 3.30.shows.the.phytoplankton.concentration.from.data. Figure 3.32 shows.the.phytoplankton.concentration.from.data,. and.from.the.homogeneous.model..It.is.difficult.to.compare.data. assuming.cells.with.size.10 μm,.and.from.the.two-layer.model.. and.model.output.owing.to.the.few.data.available..However,.the. It. is. difficult. to. compare. data. and. model. output. owing. to. the. general.range.of.values.observed.is.consistent.with.the.model. very.few.data.available..However,.the.general.range.of.values. output. observed.is.consistent.with.the.model.output.

Figure 3.30 > Caldonazzo: Comparison between measured chlorophycea concentrations (solid Figure 3.32 > Caldonazzo: Comparison between measured chlorophycea concentrations (solid points and connecting solid line) and the phytoplankton concentration produced points) and the phytoplankton concentration produced by the two-layer model by the homogeneous model (dashed line). Concentrations are in μg-C/L assuming (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio a carbon-to-biomass ratio R=0.2. R=0.2.

Two .layer.model. Effects.of.temperature.and.environmental.changes. We.use.the.same.parameter.values.reported.in.Table.2,.with.the. Since.the.model.parameters.are.the.same.as.those.used.for.Lake. following.exceptions.(we.use.the.same.parameter.values.as.for. Levico,.also.for.Lake.Caldonazzo.we.find.the.same.dependence. Lake.Levico): of.the.model.dynamics.on.temperature.and.parameter.changes. For.the.bottom.nutrient.reservoir,.we.use. -1. as.discussed.for.Lake.Levico. N0=50 μg -P L The.maximum.growth.rate.of.phytoplankton.is.fixed.at.1.8 d-1. -1 Simulation results for Lake Annecy The.maximum.growth.rate.of.planktivorous.fish.is.fixed.at.0.5 d . We. assume. a. delay. δ=0. between. the. seasonal. variation. of. The.following.data3.are.available: temperature.and.light. Measurements Period Units We. assume. a. larger. maximum. temperature. in. summer,.. . Tmax=28 °C Phosphorus.. from.1991.to.2009 μg/l Figure 3.31. shows. the. Phosphorus. concentration. for. Lake. Levico,.comparing.the.two-layer.model.output.with.the.observed. cells/L.(NB:./liter). and.biovolume data..A.good.agreement.between.the.two.curves.is.obtained. Phytoplankton. from.1996.to.2010 biovolume.in μm3/ ml3=10-3 mm3/m3

individuals/m2... NB:.the.description.in. Zooplankton from.1995.to.2009 the.file.says.ind/m2:. these.are.the.individuals. collected.over.50m.of. depth.

PB visuel The. phytoplankton. data. for. Lake. Annecy. are. given. in. terms. of. biovolume. (μm3/ml) which. are. converted. into. mm3/m3.. (the. standard. unit. used. in. this. work). dividing. the. data. values. by.1000..Then,.to.compare.with.the.carbon.content.we.have.to. multiply.the.biovolume.by.the.factor.R=0.2.. For.Lake.Annecy,.there.are.several.phytoplankton.compartments. with.large.abundance. Figure 3.33.shows.the.temporal.dynamics. of.the.different.phytoplankton.compartments.recorded.for.Lake. Annecy.. In. the. following,. we. shall. compare. the. model. output. Figure 3.31 > Caldonazzo: Total phosphorus concentration in μg-P/L from measurements at a depth of 10 meters observed (solid points) and from the two-layer model (dashed line). with.the.TOTAL.phytoplankton.abundance.

69 3 Data Source: SILA - INRA. PB visuel

Figure 3.34 > Annecy: Total phosphorus concentration in μg-P/L from measurements at a depth of Figure 3.33 > Annecy: Temporal dynamics of the different phytoplankton compartments recorded in 10 meters observed (solid points and connecting line) and from the homogeneous Lake Annecy. Compartments are identified in the legend. model (dashed line).

For zooplankton, the available data are the number counts of Figure 3.35 shows the total phytoplankton concentration from rotifers, copepods and cladocerans per square meter in the data and from the homogeneous model. The general range upper 50 meters of the water column. We converted these of values observed is consistent with the model output. The counts to number concentrations per cubic meter (dividing by model is much more regular than the data, it remains to be seen the sampling depth of 50m). Then we further converted the data whether the irregularity present in the measurements is due to to biovolumes, using the average biovolume of the different the variability of the forcings and the nutrient input, or whether zooplankton compartments from the estimates obtained for it is an intrinsically irregular dynamics of the ecosystem which is Lake Viverone. This conversion is equivalent to assuming a not captured by the model. radius of 20 micron for rotifera, 160 micron for copepods and 120 micron for cladocerans. Biomasses are then converted to carbon concentrations by multiplying by the conversion ratio R=0.5. Available temperature data (http://www.ilec.or.jp/database/eur/ eur-45.html) indicate standard values for the seasonal minimum (winter) and maximum (summer) temperature of 5°C and 20°C, as for Lake Viverone. From the available data, we find extremely low phosphorus levels (indeed this lake is known to be very clean owing to strict environmental regulations since about fifty years). Phytoplankton data suggest a rather irregular dynamics, and the annual dynamics are highly variable from one year to another. Homogeneous model In the following comparison, we use the same parameter values as in Table 1b, with the following exceptions. From the data, we find that Phosphorus levels in Lake Annecy are much lower than for Lake Viverone, and we set the nutrient Figure 3.35 > Annecy: Comparison between measured total phytoplankton concentration (solid points and connecting line) and the phytoplankton concentration produced by pool at -1. the homogeneous model (dashed line). Concentrations are in μg-C/L assuming a N0=10 μg -P L carbon-to-biomass ratio R=0.2. We assume a delay δ=40 days between the seasonal variation of temperature and light. We assume a larger growth rate of phytoplankton, -1. V1=3 day In Figure 3.36 we compare the total observed zooplankton We assume a larger zooplankton growth rate, -1. rZ=0.4 day (cladocerans, copepods plus rotifers) with the zooplankton Figure 3.34 shows the phosphorus concentration for Lake produced by the homogeneous model. Overall, the model Annecy, comparing the output of the homogeneous model with produces much lower zooplankton concentrations and a much the observed data. A good agreement between the concentration more regular dynamics than what is observed in the data. levels of the model output and the available data is obtained. However, the data display a much more irregular variability than the model.

70 Figure 3.39.shows.the.total.phytoplankton.concentration.from. data.and.from.the.two-layer.model..A.fair.agreement.between. data.and.model.output.is.obtained,.even.though.the.data.display. much.more.irregular.fluctuations.than.the.model.output.and.tend. to.reach.larger.concentrations.

Figure 3.36 > Annecy: Comparison between measured zooplankton concentrations, including Rotifera, Copepoda and Cladocera (solid points and connecting line), and the zooplankton concentration produced by the model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.5.

Interestingly,. the. zooplankton. concentration. from. the. homogeneous. model. is. closer. to. the. measured. cladoceran. compartment..Figure 3.37 shows.the.cladoceran.measurements. Figure 3.39 > Annecy: Comparison between measured total phytoplankton concentrations (solid points and connecting lines) and the phytoplankton together.with.the.model.output. concentration produced by the two-layer model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.2.

Figure 3.40.shows.the.total.observed.zooplankton.concentration. together.with.the.zooplankton.produced.by.the.two-layer.model.. A.good.agreement.between.data.and.model.output.is.obtained.. For.the.two-layer.model.with.these .parameter.values,.the.best. agreement.is.with.the.total.zooplankton.concentration.data..By. reducing.the.zooplankton.growth.rate.to.about 0.1 day-1,.a.better. agreement.is.obtained.with.the.cladoceran.compartment.

Figure 3.37 > Annecy: Comparison between measured cladoceran concentrations (solid points and connecting line), and the zooplankton concentration produced by the model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.5.

Two .layer.model We.use.the.same.parameter.values.reported.in.Table.2,.with.the. following.exceptions: For.the.bottom.nutrient.reservoir,.we.use. -1. N0=18 μg -P L We.assume.a.delay.δ=30.days.between.the.seasonal.variation.of. temperature.and.light. -1 The.maximum.growth.rate.of.phytoplankton.is.fixed.at.3 d . Figure 3.40 > Annecy: Comparison between measured total zooplankton concentration (solid points and connecting line), and the zooplankton concentration Figure 3.38. shows. the. Phosphorus. concentration. for. Lake. produced by the two-layer model (dashed line). Concentrations are in Annecy,. comparing. the. two-layer. model. output. with. the. μg-C/L assuming a carbon-to-biomass ratio R=0.5. observed.data..A.good.agreement.between.the.general.range.of. variation.of.the.model.output.and.of.the.data.is.observed..Also. for.the.two-layer.model,.the.dynamics.seem.to.be.more.regular. than.those.observed.in.the.data.

Figure 3.38 > Annecy: Total phosphorus concentration in μg-P/L from measurements 71 at a depth of 10 meters (solid points and connecting line) and from the two-layer model (dashed line). Effects.of.temperature.and.environmental.changes Simulation results for Lake Woerthersee Figure 3.41. shows. the. effects. of. increasing. temperatures. The.following.data.are.available: by. 3°C. and. increasing. the. winter. turbulent. exchange. rate. to.. 1/5 day-1.on.the.two-layer.model.dynamics.for.Lake.Annecy..In. Measurements Period Units this.case,.warmer.conditions.and.more.intense.winter.turbulent. exchange. lead. to. a. decrease. in. the. maximum. phytoplankton. Phosphorus.. from.2004.to.2010 μg/l concentration. at. the. bloom. and. to. higher. concentrations. in. the.inter-bloom.period,.while.the.zooplankton.concentration.is. biovolume.in almost.unaffected.in.the.peak.and.it.is.increased.in.the.inter- Phytoplankton. from.2007.to.2009 μm3/ml3 = 10-3mm3/m3 bloom.periods..In.addition,.there.is.a.significant.increase.in.the. fish.compartments..Note.that.the.results.reported.here.should.be. taken.as.purely.indicative.and.they.refer.only.to.the.response.of. Nutrient.levels.in.the.surface.layer.of.Lake.Woerthersee.are.not. the.model.adopted.here..These.results.should.be.seen.more.as. large. a.test.on.the.parameter.sensitivity.of.the.model.than.a.projection. on.the.future.behavior.of.the.real.lake.ecosystem. The. phytoplankton. data. for. Lake. Woerthersee. are. given. in. terms. of. biovolume. (μm3/ml) which. are. converted. into. mm3/m3 (the. standard. unit. used. in. this. work). dividing. the. data. values. by.1000..Then,.to.compare.with.the.carbon.content.we.have.to. multiply.the.biovolume.by.the.factor.R=0.2.. For. Lake. Woerthersee,. there. are. several. phytoplankton. compartments. with. large. abundance.. Figure 3.42 shows. the. temporal.dynamics.of.the.different.phytoplankton.compartments. recorded. for. Lake. Worthersee.. Notice. the. dominance. of. the. Cyanophycae.compartment..In.the.following,.we.shall.compare. the.model.output.with.the.total.phytoplankton.abundance.

Figure 3.42 > Worthersee: Temporal dynamics of the different phytoplankton compartments recorded in Lake Woerthersee. Compartments are identified in the legend.

An. important. point. is. that. in. Woerthersee. high. algal. concentrations. (mainly. Cyanophycae). start. in. fall/winter. and. continue.for.several.months.(about.200.days),.at.variance.with. Figure 3.41 > Annecy: Upper panel: Comparison between the two-layer model phytoplankton concentration in the upper layer for faster winter turbulent exchange between the two what.happens.in.the.other.lakes..The.phytoplankton.dynamics. layers, μ0=0.2 day-1, and warmer summer and winter conditions (+3 °C) (solid line) and in.this.lake.is.completely.different.from.what.seen.in.the.other. current conditions (dashed line). Lower panel: the same for zooplankton concentra- tion. Concentrations are in μg-C/L. lakes.considered.in.this.study:.the.Cyanophycae.algae,.mainly. Burgundy. Red. Algae. (Planktothrix rubescens). stay. at. depth. during. the. summer. and. emerge. to. the. surface. in. early. winter. owing.to.the.vertical.circulation.and.mixing.of.the.lake,.see. http://www.kis.ktn.gv.at/188551_EN-Carinthian_Lakes- Seenseite.?seeid=46. Two .layer.model Owing.to.the.importance.of.the.vertical.mixing.in.this.lake,.it.is. necessary.to.consider.only.the.two.layer.model..In.the.following,. we.use.the.model.with.the.parameter.values.reported.in.Table. 2,.except.than: For.the.bottom.nutrient.reservoir,.we.use.N =50 μg -P L-1. 72 0 We. assume. a. delay. δ=0. between. the. seasonal. variation. of. Effects.of.temperature.and.environmental.changes temperature.and.light. Since. in. this. version. of. the. model. the. phytoplankton. growth. The.maximum.growth.rate.of.phytoplankton.is.fixed.at 1 d-1. rate. is. rather. insensitive. to. temperature,. an. increase. in. the. The. maximum. summer. temperature. in. the. surface. layer. is. maximum. summer. temperature. leads. to. almost. no. change. in. assumed.to.be.25.°C,.consistent.with.the.observations. the.model.behavior..An.increase.of.+3°C.in.winter.temperatures. The. half. saturation. constant. for. temperature. dependence. is. does. slightly. modify. the. model. behavior. Figure 3.45. shows. fixed.as θ = 1 °C,.that.is,.we.practically.assume.no.temperature. the. effects. of. increasing. temperatures. by. 3°C. and. increasing. dependence. in. the. growth. rate. of. the. algae.. This. is. the. main. the.winter.turbulent.exchange.rate.to.1/5 day-1.on.the.two-layer. difference.with.respect.to.the.other.phytoplankton.models.used. model. dynamics. for. Lake. Woerthersee.. Also. in. thie. case,. the. so.far. change.in.model.behavior.is.rather.minor. We.assume.a.faster.export.of.nutrient.from.the.surface.layer,.that. is, β = 0.1 day-1. Figure 3.43. shows. the. Phosphorus. concentration. for. Lake. Woerthersee,. comparing. the. two-layer. model. output. with. the. observed.data.. Figure 3.44. shows. the. Cyanophycae. concentration. from. data. and. from. the. two-layer. model.. A. good. agreement. between. data.and.model.output.is.obtained,.especially.in.the.timing.of. the.Cyanophycae.explosions..This.is.due.to.the.fact.that.at.the. beginning.of.winter.there.is.strong.vertical.mixing.in.the.lake.and. nutrients.are.brought.to.the.surface..However,.the.model.results. are.more.regular,.while.the.data.display.larger.variability. Interestingly,. these. results. show. the. wide. applicability. of. the. simple.two.layer.ecosystem.model.introduced.here.

Figure 3.45 > Woerthersee: Comparison between the two-layer model phytoplankton concentration in the upper layer for faster winter turbulent exchange between the two layers, -1, and warmer summer and winter conditions (+3°C) μ0=0.2 day (solid line) and current conditions (dashed line). Concentrations are in μg-C/L.

On. the. other. hand,. in. this. configuration. the. model. is. rather. sensitive. to. the. nutrient. export. rate. from. the. surface. layer... Figure 3.46. shows. the. case. for. a. slower. rate,. β = 0.05 day-1... In.this.case,.the.Cyanophycae.concentration.in.the.surface.layer. stays.larger.for.a.significantly.longer.time.

Figure 3.43 > Woerthersee: Total Phosphorus concentration in μg-P/L from measurements (solid points and connecting line) and from the two-layer model (dashed line).

Figure 3.46 > Woerthersee: Comparison between the two-layer model phytoplankton concentration in the upper layer for slower nutrient export from the upper layer, β=0.05 day-1 (solid line), and current conditions (dashed line). Concentrations are in μg-C.

Note.that.the.results.reported.here.should.be.taken.as.purely. indicative. and. they. refer. only. to. the. response. of. the. model. adopted.here..These.results.should.be.seen.more.as.a.test.on. the.parameter.sensitivity.of.the.model.than.a.projection.on.the. future.behavior.of.the.real.lake.ecosystem. Figure 3.44 > Woerthersee: Comparison between measured Cyanophycae concentration (solid points and connecting lines) and the phytoplankton concentration produced by the two-layer model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.2. 73 References

Andersen, T. and D. O. Hessen, 1991. Carbon, nitrogen, Steele, J.H. and W. Henderson, 1992. The role of predation in and phosphorus content of freshwater zooplankton, Limnol. plankton models. J. Plankton Res. 14, 152–172. Oceanogr. 36, 807-814. Straškrabová V., C. Callieri, P. Carrillo, L. Cruz-Pizarro, J. Fott, Bracco, A., A. Provenzale and I. Scheuring, 2000. Mesoscale P. Hartman, M. Macek, J. M. Medina-Sánchez, J. Nedoma vortices and the paradox of the plankton. Proc. R. Soc. B 267, and K. Šimek, 1999. Investigations on pelagic food webs in 1795-1800. mountain lakes - aims and methods. Journal of Limnology 58, 77-87. Bratbak, G. and I. Dundas, 1984. Bacterial Dry Matter Content and Biomass Estimations. Applied and Environmental Thébault, J.M. and S. Rabouille, 2003. Comparison between Microbiology, 48, 755-757. two mathematical formulations of the phytoplankton specific growth rate as a function of light and temperature, in two Koszalka, I., A. Bracco, C. Pasquero and A. Provenzale, simulation models (Aster&Yoyo). Ecological Modelling 163, . 2008. Plankton cycles disguised by turbulent advection. Theor. 145-151. Pop. Biology 72, 1–6. Touratier, F., J. G. Field and C. L. Moloney, 2001. . Kot, M. 2001. Elements of Mathematical Ecology, Cambridge A stoichiometric model relating growth substrate quality (C:N:P University Press. ratios) to N:P ratios in the products of heterotrophic release and Martin, A., K.J. Richards, A. Bracco and A. Provenzale, 2002. excretion, Ecological Modelling 139, 265-291. Patchy productivity in the open ocean. Global Biogeochem. Vrede, K., M. Heldal, S. Norland and G. Bratbak, 2002. Cycles 16, 10.1029/2001GB001449. Elemental composition (C, N, P) and cell volume of exponentially Pasquero, C., A. Bracco and A. Provenzale, 2005. Impact growing and nutrient-limited bacterioplankton, Applied and of the spatiotemporal variability of the nutrient flux on primary Environmental Microbiology, 68, 29652971. productivity in the ocean. J. Geophys. Res. 110, C07005. Wetzel, R.G., 2001. Limnology – Lake and River Ecosystems, Reynolds, C.S., 2006. Ecology of phytoplankton, Cambridge Elsevier Academic Press. University Press. Sarnelle, O. and A.E. Wilson, 2008. Type-III functional response in Daphnia, Ecology, 89, 1723–1732.

List of tables

Table 3.1...... 52 Table 3.3...... 58 Parameter values for the homogeneous model. Parameter values for the two-layer model.

Table 3.2...... 55 Alternative parameter values for the homogeneous model. The values differing from Table 1a have been indicated in bold.

List of figures

Figure 3.1...... xx Figure 3.2...... xx Model carbon concentrations for phytoplankton (solid line), Total Phosphorus concentration in μg-P/L from measurements at zooplankton (long-dashed line), planktivorous fish (short-dashed a depth of 10 meters in Lake Viverone (solid line) and from the line) and piscivorous fish (dotted line), for the parameter values homogeneous model (dashed line). reported in Table 1. All concentrations are expressed in μg-C/L.

74 Figure 3.3...... xx Figure 3.12...... xx Comparison. between. measured. chlorophycea. concentrations. Comparison.between.the.two-layer.model.dynamics.for.warmer. (solid.line),.assuming.cells.with.approximate.equivalent.radius. conditions.(solid.lines).and.model.dynamics.for.current.conditions. of.15 μm,.and.the.phytoplankton.concentration.produced.by.the. (dashed. lines).. Left. panel:. phytoplankton. concentrations. homogeneous.model.(dashed.line)..Concentrations.are.in μg-C/L. in. the. upper. layer;. right. panel:. zooplankton. concentration... assuming.a.carbon-to-biomass.ratio.R=0.2. All.concentrations.are.expressed.in.μg-C/L.

Figure 3.4...... xx Figure 3.13...... xx Comparison. between. measured. zooplankton. concentrations,. Comparison. between. the. two-layer. model. dynamics. for. including. Rotifera,. Copepoda. and. Cladocera. (solid. line),. and. faster.relaxation.to.the.bottom.nutrient.pool,.τ = 15 days.(solid. the.zooplankton.concentration.produced.by.the.homogeneous. line),.for.the.standard.value.of .the.relaxation.time,.τ = 30 days.. model.(dashed.line)..Concentrations.are.in.μg-C/L.assuming.a. (long-dashed.line).and.for.slower.relaxation,.τ = 60 days.(short- carbon-to-biomass.ratio.R=0.5 dashed.line)..All.concentrations.are.expressed.in.μg-C/L.

Figure 3.5...... xx Figure 3.14...... xx Total.Phosphorus.concentration.in.mg-P/L.from.measurements. Comparison.between.the.two-layer.model.dynamics.for.faster. at.a.depth.of.10.meters.in.Lake.Viverone.(solid.line).and.from. winter. turbulent. relaxation. between. the. two. layers,. τ. = 5 days the.homogeneous.model.with.the.alternative.set.of.parameter. (solid.line),.for.the.standard.value.of.the.relaxation.time,.τ = 30 days.. values.(dashed.line). (long-dashed.line).and.for.slower.relaxation,.τ = 60 days.(short- dashed.line)..All.concentrations.are.expressed.in.μg-C/L. Figure 3.6...... xx Comparison. between. measured. chlorophycea. concentrations. Figure 3.15...... xx (solid.line),.assuming.cells.with.approximate.equivalent.radius. Avigliana. Grande:. Total. phosphorus. concentration. in. μg-P/L. of. 15 μm,. and. the. phytoplankton. concentration. produced. by. from.measurements.at.a.depth.of.10.meters.(solid.points.and. the.homogeneous.model.with.the.alternative.set.of.parameter. connecting. line). and. from. the. homogeneous. model. (dashed. values.(dashed.line)..Concentrations.are.in.μg-C/L.assuming.a. line). carbon-to-biomass.ratio.R=0.2. Figure 3.16...... xx Figure 3.7...... xx Avigliana.Grande:.Measured.chlorophycea.concentrations.(solid. Left.panel:.Total.zooplankton.concentration.from.observations. points. and. connecting. line). and. phytoplankton. concentration. (solid.line).and.model.zoopankton.concentration.(dashed.line).. produced. by. the. homogeneous. model. (dashed. line).. Right. panel:. Cladoceran. concentration. from. observations. Concentrations. are. in. μg-C/L. using. a. carbon-to-biomass. ratio. (solid.line).and.model.zooplankton.concentration.(dashed.line).. R=0.2.and.assuming.cells.with.size.10 μm. Concentrations. are. in. mg-C/L. assuming. a. carbon-to-biomass. Figure 3.17...... xx ratio.R=0.5. Avigliana. Grande:. Total. phosphorus. concentration. in. μg-P/L. Figure 3.8...... xx from.measurements.at.a.depth.of.10.meters.(solid.points.and. Upper. panel:. Phytoplankton. in. the. upper. layer. (solid. connecting.line).and.from.the.two-layer.model.(dashed.line). curve),. zooplankton. (long-dashed. curve),. planktivorous. fish.. (short-dashed.curve).and.piscivorous.fish.(dotted.curve).for.the. Figure 3.18...... xx two-layer. lake. ecosystem. model.. Lower. panel:. Phytoplankton. Avigliana.Grande:.Measured.chlorophycea.concentrations.(solid. in.the.upper.layer.(solid.curve).and.in.the.lower.layer.(dashed. points. and. connecting. line). and. phytoplankton. concentration. curve).for.the.two-layer.lake.ecosystem.model. produced.by.the.two-layer.model.(dashed.line)..Concentrations. are. in. μg-C/L. using. a. carbon-to-biomass. ratio. R=0.2. and. Figure 3.9...... xx assuming.cells.with.size.10 μm. Total.Phosphorus.concentration.in.μg-P/L..from.measurements.in. lake.Viverone.(solid.line).and.from.the.two-layer.lake.ecosystem. Figure 3.19...... xx model.(dashed.line). Avigliana:. Upper. panel:. Comparison. between. the. two-layer. model.phytoplankton.concentration.in.the.upper.layer.for.faster. Figure 3.10...... xx winter.turbulent.exchange.between.the.two.layers,. -1,.. μ0=0.2 day Phytoplankton. concentration. from. measurements. in. lake. and. warmer. summer. and. winter. conditions. (+3°C). (solid. line). Viverone. (solid. line). and. upper-layer. phytoplankton. from. the. and.current.conditions.(dashed.line)..Lower.panel:.the.same.for. two-layer.lake.ecosystem.model.(dashed.line)..Concentrations. zooplankton.concentration..Concentrations.are.in.μg-C/L. are.in.μg-C/L.assuming.a.carbon-to-biomass.ratio.R=0.2. Figure 3.20...... xx Figure 3.11...... xx Sirio:. Total. phosphorus. concentration. in. μg-P/L. from. Upper.panel:.Total.zooplankton.concentration.from.observations. measurements. at. a. depth. of. 10. meters. (solid. points. and. (solid.line).and.model.zoopankton.concentration.(dashed.line).. connecting. line). and. from. the. homogeneous. model. (dashed. Lower. panel:. Cladoceran. concentration. from. observations. line). (solid.line).and.model.zooplankton.concentration.(dashed.line).. Concentrations. are. in. mg-C/L. assuming. a. carbon-to-biomass. ratio.R=0.5.

75 Figure 3.21...... xx Figure 3.30...... xx Sirio: Comparison between measured chlorophycea Caldonazzo: Comparison between measured chlorophycea concentrations (soild points and connecting lines) and the concentrations (solid points and connecting solid line) and the phytoplankton concentration produced by the homogeneous phytoplankton concentration produced by the homogeneous model (dashed line). Concentrations are in μg-C/L assuming a model (dashed line). Concentrations are in μg-C/L assuming a carbon-to-biomass ratio R=0.2. The concentration was derived carbon-to-biomass ratio R=0.2. from the number concentratation assuming an equivalent cell Figure 3.31...... xx radius of 10 μm. Caldonazzo: Total phosphorus concentration in μg-P/L from Figure 3.22...... xx measurements at a depth of 10 meters observed (solid points) Sirio: Total phosphorus concentration in μg-P/L from and from the two-layer model (dashed line) measurements at a depth of 10 meters (solid points and connecting line) and from the two-layer model (dashed line). Figure 3.32...... xx Caldonazzo: Comparison between measured chlorophycea Figure 3.23...... xx concentrations (solid points) and the phytoplankton Sirio: Comparison between measured chlorophycea concentration produced by the two-layer model (dashed line). concentrations (solid points and connecting lines) and the Concentrations are in μg-C/L assuming a carbon-to-biomass phytoplankton concentration produced by the homogeneous ratio R=0.2. model (dashed line). Concentrations are in μg-C/L assuming a Figure 3.33...... xx carbon-to-biomass ratio R=0.2. The concentration was derived from the number concentration assuming an equivalent cell Annecy: Temporal dynamics of the different phytoplankton radius of 10 μm. compartments recorded in Lake Annecy. Compartments are identified in the legend. Figure 3.24...... xx Figure 3.34...... xx Levico: Total phosphorus concentration in μg-P/L from measurements at a depth of 10 meters observed (solid points) Annecy: Total phosphorus concentration in μg-P/L from and from the homogeneous model (dashed line). measurements at a depth of 10 meters observed (solid points and connecting line) and from the homogeneous model (dashed Figure 3.25...... xx line). Levico: Comparison between measured chlorophycea concentrations (solid points) and the phytoplankton Figure 3.35...... xx concentration produced by the homogeneous model (dashed Annecy: Comparison between measured total phytoplankton line). Concentrations are in μg-C/L assuming a carbon-to- concentration (solid points and connecting line) and the biomass ratio R=0.2. The concentration was derived from the phytoplankton concentration produced by the homogeneous cell number concentration assuming an equivalent radius of . model (dashed line). Concentrations are in μg-C/L assuming a 10 μm. carbon-to-biomass ratio R=0.2.

Figure 3.26...... xx Figure 3.36...... xx SLevico: Total phosphorus concentration in μg-P/L from Annecy: Comparison between measured zooplankton measurements at a depth of 10 meters observed (solid points) concentrations, including Rotifera, Copepoda and Cladocera and from the two-layer model (dashed line). (solid points and connecting line), and the zooplankton concentration produced by the model (dashed line). Figure 3.27...... xx Concentrations are in μg-C/L assuming a carbon-to-biomass Levico: Comparison between measured chlorophycea ratio R=0.5. concentrations (solid points) and the phytoplankton concentration produced by the two-layer model (dashed line). Figure 3.37...... xx Concentrations are in μg-C/L assuming a carbon-to-biomass Annecy: Comparison between measured cladoceran ratio R=0.2. The concentration was derived from the cell number concentrations (solid points and connecting line), and the concentration assuming an equivalent radius of 10 μm. zooplankton concentration produced by the model (dashed line). Concentrations are in μg-C/L assuming a carbon-to- Figure 3.28...... xx biomass ratio R=0.5. Levico: Upper panel: Comparison between the two-layer model phytoplankton concentration in the upper layer for faster Figure 3.38...... xx winter turbulent exchange between the two layers, -1, . Annecy: Total phosphorus concentration in μ from μ0=0.2 day g-P/L and warmer summer and winter conditions (+3°C) (solid line) measurements at a depth of 10 meters (solid points and and current conditions (dashed line). Mid panel: the same for connecting line) and from the two-layer model (dashed line). zooplankton concentration. Lower panel: The same for the Figure 3.39...... xx planktivorous fish abundance. Concentrations are in μg-C/L. Annecy: Comparison between measured total phytoplankton Figure 3.29...... xx concentrations (solid points and connecting lines) and the Caldonazzo: Total phosphorus concentration in μg-P/L from phytoplankton concentration produced by the two-layer model measurements at a depth of 10 meters observed (solid points) (dashed line). Concentrations are in μg-C/L assuming a carbon- and from the homogeneous model (dashed line). to-biomass ratio R=0.2.

76 Figure 3.40...... xx Annecy:. Comparison. between. measured. total. zooplankton. concentration. (solid. points. and. connecting. line),. and. the. zooplankton. concentration. produced. by. the. two-layer. model. (dashed.line)..Concentrations.are.in.μg-C/L.assuming.a.carbon- to-biomass.ratio.R=0.5.

Figure 3.41...... xx Annecy:. Upper. panel:. Comparison. between. the. two-layer. model.phytoplankton.concentration.in.the.upper.layer.for.faster. winter.turbulent.exchange.between.the.two.layers,. -1,.. μ0=0.2 day and. warmer. summer. and. winter. conditions. (+3°C). (solid. line). and.current.conditions.(dashed.line)..Lower.panel:.the.same.for. zooplankton.concentration..Concentrations.are.in.μg-C/L.

Figure 3.42...... xx Worthersee:.Temporal.dynamics.of.the.different.phytoplankton. compartments. recorded. in. Lake. Woerthersee.. Compartments. are.identified.in.the.legend.

Figure 3.43...... xx Woerthersee:. Total. Phosphorus. concentration. in. μg-P/L. from. measurements.(solid.points.and.connecting.line).and.from.the. two-layer.model.(dashed.line).

Figure 3.44...... xx Woerthersee:. Comparison. between. measured. Cyanophycae. concentration. (solid. points. and. connecting. lines). and. the. phytoplankton. concentration. produced. by. the. two-layer. model.(dashed.line)..Concentrations.are.in.μg-C/L.assuming.a.. carbon-to-biomass.ratio R=0.2.

Figure 3.45...... xx Woerthersee:. Comparison. between. the. two-layer. model. phytoplankton.concentration.in.the.upper.layer.for.faster.winter. turbulent. exchange. between. the. two. layers,. -1,. and. μ0=0.2 day warmer. summer. and. winter. conditions. (+3°C). (solid. line). and. current.conditions.(dashed.line)..Concentrations.are.in.μg-C/L.

Figure 3.46...... xx Woerthersee:. Comparison. between. the. two-layer. model. phytoplankton. concentration. in. the. upper. layer. for. slower. nutrient.export.from.the.upper.layer,.β=0.05 day-1.(solid.line),.and. current.conditions.(dashed.line)..Concentrations.are.in.μg-C..

77 4. The hydrodynamic model

78 The present hydrodynamic climate impact lake simulation study (HCILS) uses a hydrodynamic model at three pilot sites to characterize the climate impact on the hydrodynamics of Alpine lakes. As part of the SILMAS work package »Alpine lakes running changes« the results of HCILS provide knowledge of current and possible future physical trends in temperature, circulation, stratification, and heat balance patterns.

Preface

Numerous. studies. have. proven. a. change. in. climate. since. concerns.the.circulation,.is:.How.does.a.characteristic.mixing. the. beginning. of. the. 19th. century.. This. change. is. coupled. to. behavior.of.a.lake.change.in.time.and.space?.Are.there.changes. industrial.and.technological.caused.greenhouse.gas.emissions. in.stratification.stability,.mixing.intensity,.and.mixing.duration?. significantly. and. causally.. Impact. studies,. that. concern. the. How.do.special.lake.types.react.to.this.(e.g..meromictic.lakes)? change’s.effects.to.water.resources,.are.just.as.numerous.(Bates,. Circulation. is. strongly. dependent. on. stratification. stability.. 2008)..Particularly,.lakes.are.sentinels,.indicators.and.regulators. Thus,. the. water. temperature. is. the. key. parameter. that. has. to. of.climate.change.(Williamson.et.al.,.2009)..Concerning.lake’s. be.analyzed.in.its.distribution.in.space.and.time.in.characteristic. physics,. one. can. see. a. strong. correlation. between. physical. depths.to.get.an.idea.of.temperature.changes.in.different.depths. processes’.change,.and.climate.change.. The. temperature,. in. turn,. depends. on. the. lakes. heat. budget.. There. is. a. need. for. knowledge. for. lake. management. about. a. The. main. focus. lies. on. heat. transport. processes. at. the. lakes. lake’s.physical. state.. Water. quality. respectively. the. ecological. surface.. This. will. also. concern. evaporation. as. the. connective. state.are.close.connected.to.physical.processes..Not.only.the. link.between.heat.and.water.budget. actual. state,. especially. future. properties. of. the. lake’s. physics. have. to. be. evaluated.. If. adaptation. strategies. are. developed,. This. study. is. divided. into. four. general. steps:. The. first. one. is. this. knowledge. is. a. necessary. input.. An. impact. study. is. the. the.choice.of.representative.alpine.lakes..Sufficient.data.must. appropriate.method.to.get.a.general.view.of.the.climate’s.role. be. available. for. these. lakes.. Additionally,. the. model’s. code. in. lakes. ecological,. chemical,. and. physical. state.. Against. this. is. fitted. to. alpine. surroundings. in. sense. of. implementation. of. background,. present. study. can. be. considered. as. an. impact. additional. processes.. A. hydrodynamic. model. (1-dimensional. study.that.analyses.the.effects.of.changing.mean.meteorological. vertical). is. provided. by. Deltares,. Netherlands.. In. the. second. conditions.onto.lake’s.physics..Hence,.it.is.the.objective.to.study. step,. the. measured. meteorological. data. are. preprocessed,. the.reaction.of.physical.processes.and.parameters. and.the.models.are.calibrated.and.validated..This.provides.the. groundwork.for.step.three:.Development.of.climate.scenarios.. The.intensity.of.circulation.is.one.of.these.processes..That’ s.of. Finally,.the.lake.physics.are.calculated.in.the.fourth.step.using. great. importance. for. nutrient. and. oxygen. transportation. into. the. climate. scenarios.. The. calculation. results. provide. the. the.deep.water.layers.of.a.lake..Most.of.the.larger.Alpine.lakes. necessary.data.to.answer.the.before.mentioned.questions. have. their. typical. circulation. characteristics,. e.g.. an. annual. mixing.in.the.transition.from.winter.to.spring..The.question,.that.

The hydrodynamic model

For.the.investigation.of.the.physics.in.an.Alpine.Lake,.a.physical. evanescent.in.comparison.to.the.vertical.gradients..Furthermore,. based. thermodynamic. model. was. applied. in. this. study.. It. is. pure. density. stratification. is. assumed. (hydrostatic. pressure. a. one-dimensional,. vertical. approach. (1dv).. The. model. was. assumption). developed.by.Deltares.(Netherlands)..It.solves.the.Navier.Stokes. equations.for.incompressible.fluids.under.the.shallow.water.and. Boussinesq.assumption..The..-..turbulence.model.is.used.in.the. model.setup. In. this. 1dv. model,. a. lake. is. considered. as. a. 1dimensional. column. that. is. divided. into. horizontal. layers (see Figure Figura 4.1)..The.volumes.and.boundary.layers.are.parameterized. by. the. bathymetry,. if. processes. like. momentum. exchange. or. constituent.transport.are.calculated..Hence,.one.cell.represents. the. mean. conditions. in. a. lake’s. horizontal. water. layer.. This. approach. presupposes. that. the. horizontal. gradients. are. Figura 4.1 > One dimensional, vertical model approach. 79 Adaption to alpine lakes

For. a. better. representation. the. model. is. adapted. to. Alpine. conditions. by. the. implementation. of. three. aspects. into. the. model.code..One.of.these.adaptions.is.an.implementation.of.a. variable.Secchi-depth..For.instance,.at.Lake.Constance.there.is. a.variation.in.a.range.of.15.meters..This.has.an.effect.onto.the. radiation.penetration.into.the.water..Thus,.it.has.an.effect.onto. the. epilimnion. depth. and. temperature.. The. implementation. of. these.processes.into.the.model.enables.the.consideration.of.a. default.yearly.course.of.the.Secchi-depth. The. surrounding. mountains,. which. reduce. the. incoming. solar. radiation.by.shading,.are.important,.too..This.was.also.refined.in. the.models.code.with.a.new.parameter.gipalt.(Figure 4.2). Figure 4.2 > Consideration of surrounding mountains by the new parameter gipalt. It.is.a.1-dimensional.model..Considering.this,.the.lakes.shape. has.an.effect.onto.the.wind.induced.impulse.transport.into.the. lakes.surface..An.additional.weighting.of.wind.speed.depending. on.its.direction.is.additionally.implemented..The.drag.coefficient... (see.e.g..Smith.&.Banke,.1975).is.modified.by.a.function.of.the. lakes.shape.and.orientation.(eq. 1 and Figure 4.3).

with..α. wind.direction.[0;360]° ..ϕ. angle.of.the.lake’s.orientation.[0;360]°

.. . the.lake’s.eccentricity.[0;1] Figure 4.3 > Modification of the drag coefficient. The.analysis.of.lakes.mixing.is.realized.by.applying.numerical. tracer..This. is. introduced. to. the. upper. model. layer. before. the. beginning.of.the.mixing.period..When.the.stagnation.starts.due. to.stratification,.the.distribution.of.the.tracer.is.analyzed.to.draw. some.conclusions.about.the.mixing.intensity.

The models parameterization and meteorological forcing

The. 1dv. model. is. a. physical-based. model.. That. means. that. Some.parameters.are.given.by.the.lake’s.geometry,.the.situation. all. processes. are. modeled. on. the. base. of. physical. laws,. and. around.the.lake,.or.by.measurements..These.parameters.differ. not. on. empirical. equations.. Nevertheless,. there. are. numerous. from.lake.to.lake..They.are.listed.in.Table 4.2. coefficients.. Some. of. these. parameters. are. case. dependent,. and.they.vary.by.the.modeled.situation..The.most.of.them.are. given.by.default.values,.which.are.prevailingly.and.sufficiently. Name Description Unit accurate..Others.have.to.be.calibrated..These.parameters.are. The.lake’s.bathymetry.(surface.as.. listed.in.Table 4.1. HYP [m²/s] a.function.of.depth)

SECHID Mean.annual.course.of.Secchi.depth [m²/s] Name Description Unit Mean.altitude.of.surrounding.(south). GIPALT VICBCK Background.eddy.viscosity [m²/s] mountain.tops [m]

DIFBCK Background.eddy.diffusivity [m²/s] EXZ Eccentricity:.width/length.of.a.lake [-]

OZMIDV Ozmidov.length.scale [m] DIRE Orientation.of.the.lake [-]

Stanton.number.for.convective.heat. STANTN [-] Table 4.2 > Parameters of the hydrodynamic model that are given by measurements. exchange

Dalton.number.for.evaporative.heat. DALTON [-] exchange

Table 4.1 > Calibrated parameters of the hydrodynamic model.

80 The.latter.mentioned.parameters.are.constant.or.they.vary.with. Name Description Unit a.constant.oscillation.(e.g..the.Secchi.depth)..The.meteorology. is.the.main.model.input,.and.it.is.not.constant..This.data.have.to. T Air.temperature [°C] be.given.as.time.series..The.necessary.climate.variables.can.be. Relative.air.humidity found.in.Table 4.3. rF [%]

WV Wind.velocity [m/s]

WDIR Wind.direction [0;360]°

CL Cloud.cover [%]

Table 4.3 > Meteorological variables.

Calibrating the model with an evolutionary algorithm

Table 1. shows. the. parameters. that. have. to. be. calibrated.. Appropriate. parameter. setups. must. be. found. for. each. lake.. This. guarantees. a. good. mathematical. representation. of. a. lake,. and. the. errors. i.e.. the. objective. function. are. minimized.. The.approach.that.is.used.in.this.study.is.copied.from.nature:. Evolution.. Evolutionary. algorithms. (EAs). are. powerful. tools. in. optimizing.multi.parameter.problems,.if.up.to.infinite.number.of. solutions.exist.in.a.multidimensional.parameter.space..They.are. used. in. science. and. engineering.. They. became. an. intensively. investigated. topic. in. the. recent. years,. what. can. be. seen. in. the.formidable.number.of.publications..For.an.overview.in.EA. application.see.Bäck.(1993). In. this. study. a. simple. EA. was. developed,. that. optimizes. the. model’s.parameter..The.principles.and.the.EA.setup.are.shown. in.Figure 4.4.

Figure 4.4 > The evolutionary algorithm (EA) that was developed in HCILS to optimize the parameter for the hydrodynamic model. The.algorithm.was.tested.with.different.numbers.of.parent.and. children.parameter.sets..Because.a.30.years.model.run.takes. approximately. 30. minutes. for. Lake. Constance. the. number. of. parameter.sets.was.the.limiting.factor..In.a.simple.test.the.EA. was. programmed. to. find. the. minimum. (x. and. y. position). in. a. Figure 4.5 > Simple test of the evolutionary algorithm (EA) with two parameters. The individuals «sine.landscape»..Figure.4.5.shows.the.results.of.this.test..The. aggregate successfully around the minimum in a «sine landscape» (a) after 10 generations (b). EA.worked.well.with.50.children.and.10.parent.individuals.

81 The three pilot sites

Three. pilot. sites. are. chosen:. Lago. di. Viverone. (Italy),. Lake. Wörthersee. (Austria),. and. Lake. Constance. (Austria,. Germany,. and. Switzerland). (Figure 4.6).. This. choice. was. made. by. the. consideration. of. data. availability,. representation. of. an. alpine. region,.and.viability.in.the.scheduled.project.timeframe. In. the. following. there. is. a. brief. description. of. each. lake. with. information.about.the.modeling-relevant.data.

Figure 4.6 > The three lakes in HCILS of the SILMAS project: Lake Constance, Lago di Viverone and Wörthersee.

Lake Constance: Data, parameterization, and calibration results

Lake. Constance. is. situated. at. the. northern. edge. of. the. Alps.. Climate data It. is. an. international. lake. at. the. borders. of. Austria,. Germany,. Mean.hourly.data.for.air.humidity,.air.temperature,.wind.speed,. and.Switzerland (see Figure 4.7)..In.HCILS.the.lake.is.modeled. wind.direction,.and.cloudiness.cover.the.period.1964-2007..This. without. the. »Untersee«.. Usually. the. latter. mentioned. lake. is. data. are. measurements. at. the. station. Konstanz. and. they. are. also. considered. as. Lake. Constance.. But. in. a. physical. view. it. assumed.as.representative.for.the.entire.lake. is. a. separate,. and. shallow. lake. with. an. approximately. 0.2m. deeper. lying. water. level.. Lake. Constance. is. an. oligotrophic,. Bathymetry monomictic.lake.that.is.pased.by.the.Rhine..The.most.important. characteristics.are.listed.in.Table 4.4. The.bathymetrie.of.Lake.Constance.is.shown.in Figure 4.8..There. are.three.characteristic.zones:.the.shallow.zone.(395.3.m+NN.–. ca..390.m+NN),.the.moderate.zone.(ca..390m+NN.-.220m+NN),. and. the. deep. and. steep. zone. (ca.. 220m+NN. -. 142.1m+NN).. Data.source.is.the.database.of.the.Institute.for.Lake.Research. of.the.LUBW.

Figure 4.7 > Map of Lake Constance.

Lake data Value Mean.water.level 395.3.m.above.sea-level

Maximal.depth 254.m Figure 4.8 > Bathymetrie of Lake Constance. Mean.depth 101.m Surface.area 472.33.km² Secchi depth Water.volume 47.6.km³ As.it.was.mentioned.in.chapter.4.2.2.2,.the.model.was.adapted. to.alpine.lakes.(among.others).by.implementing.a.variable.secchi. Table 4.4 > Characteristics of Lake Constance (the South-Eastern big part «Obersee»). depth..Figure 4.9.and.Table 4.5.show.why:.At.Lake.Constance. there.is.a.typical.annual.course.of.secchi.depth..In.this.boxplot. it.can.be.seen,.that.in.the.warm.seasons.there.is.a.mean.secchi. depth. of. around. 3. to. 7. m.. In. the. cold. seasons. this. value. lies. around.10m..In.Table.5.the.detailed.statistics.are.listed.

82 Parameter Value

Stanton.number 0.001 Dalton.number 0.0013 Background.viscosity 0.00001 Background.diffusivity 0.0000016

Table 4.6 > Calibrated Parameters for Lake Constance.

Figure 4.9 > Boxplot of the Secchi depth at Lake Constance in the period 1974-2010. Source of Secchi depth time series: Data base of the Institute for Lake Research (LUBW).

Month Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec

Max 14.9 18.0 16.5 13.5 7.4 14.0 7.5 8.0 8.0 10.5 11.0 16.0

Mean 10.67 11.50 11.17 6.99 3.77 6.61 4.40 4.05 4.83 6.49 8.17 9.74

Min 6.0 6.5 1.7 2.5 1.6 1.5 1.7 2.0 2.5 4.0 4.0 5.2

Table 4.5 > Mean monthly Secchi depth, and maximal and minimal measured values of Secchi depth in Lake Constance in the period 1974-2010. Data source: Data base of the Institute for Lake Research (LUBW).

Water temperature data There. are. long. time. series. of. measured. water. temperature. profiles.at.the.deepest.point.of.the.lake..These.measurements.are. conducted.by.the.Institute.for.Lake.Research.in.a.regular.interval. of.approximately.one.month..For.calibration.and.validation.the. Figure 4.11 > Deviation statistics of measured minus calculated water temperatures in each data.are.available.in.the.period.1964-2006..Figure 4.10.shows. month within the calibration at Lake Constance. the.mean.temperatures.in.the.measured.depths.

Figure 4.10 > The mean measured monthly water temperatures at Lake Constance in the period 1964-2006. The observation position is Fischbach-Uttwil (the lake’s deepest point).

Calibration Results The. calibration. with. the. evolutionary. algorithm. results. in. a. good. recalculation. of. mean. temperature. conditions. in. Lake. Constance.. Figure 4.11. shows. the. deviation. statistics. within. the. calibration. period.. It. can. be. seen,. that. the. recalculated. temperatures. are. scattered. with. a. low. variance. around. zero. degree.Celsius.deviation..In.August.to.October.the.differences. are. higher. because. of. a. not. exact. recalculated. thermocline. depth..This.little.shift.leads.to.higher.deviations.especially.in.the. range.of.higher.temperature.gradients..However,.the.quality.of. the.recalculated.water.temperatures.is.sufficient..In Table 4.6.the. calibrated.parameters.and.their.best.solutions.are.listed. .

83 Lago di Viverone: Data, parameterization, and calibration results

The.monomictic.lake.Lago.di.Viverone.is.located.at.the.borders. Secchi depth of.the.Piemont.provinces.Biella,.Torino,.and.Vercelli..In.Table.4.7. The.assumptions.for.the.Secchi.depth.in.the.model.setup.are. the.main.characteristics.are.listed. based.on.measurements.that.were.made.in.the.years.2001.and. 2002.(Regione.Piemonte,.2004)..There.is.no.clear.annual.course. of.the.Secchi.depth..The.mean.is.about.5.m..That’s.why.in.the. model.a.constant.Secchi.depth.of.5m.was.choosen. . Water temperature Data For. the. present. study,. water. temperature. measurements. are. available. in. the. period. 2001-2009.. The. interval. between. the. measurements.is.approximately.one.month..Figure.4.14.shows. the.mean,.measured,.and.monthly.temperatures.in.this.period..It. shows.for.instance,.that.July.is.the.warmest.month.with.around. 26.8°C.in.the.surface.water.layer..This.is.one.major.identifier.in. which.it.differs.from.lakes.in.the.northern.Alpine.region,.e.g..Lake. Figure 4.12 > Map of Lago di Viverone; Source: Regione Piemonte (2004); Figure 7-1. Constance.(epilimnic.temperature.maximum.in.August:.20.5°C).

Lake data Value

Mean.water.level 229.m.above.sea-level Maximal.depth 50.m Mean.depth 22.5.m Surface.area 5.72.km² Water.volume 128.77.Mm³ Figure 4.14 > Mean monthly measured water temperatures at Lago di Viverone in the period 2001-2009.

Table 4.7 > Characteristics of Lago di Viverone. Calibration Results Climate data Figure 4.15. shows. the. statistics. of. the. deviation. between. the. measured. and. modeled. water. temperatures. (the. model. was. After. filling. data. gaps,. and. after. choosing. the. periods. where. calibrated. by. the. evolutionary. algorithm). in. the. recalculation. all.necessary.climate.data.are.available.in.a.1h.time.resolution. of. the. measured. period. 2002-2009.. In. spring. there. is. an. there.are.eight.years.(2002-2009).of.complete,.representative,. underestimation. of. the. surface. water. temperature.. In. the. and. usable. data. for. Lago. di. Viverone.. Four. stations. around. stratification. period. the. depth. of. the. metalimnion. cannot. be. the.lake.were.used.to.get.this.data..These.stations.are.located. reproduced.exactly.by.the.model..This.leads.to.higher.deviations. at. Piverone. (at. the. banks. of. Anzasco),. Borgofranco. d’Ivera,. in. this. range.. The. main. reason,. why. these. results. are. of. less. Massazza,.and.Lago.di.Candia..Air.temperature,.humidity,.and. quality.as.at.Lake.Constance.is.the.forcing.input.of.cloud.cover.. wind. data. are. provided. by. the. station. at. Piverone.. Radiation. Latter.mentioned.data.are.estimated.from.radiation.data.around. data.are.combined.information.of.the.other.three.stations. the.lake..In.Table 4.8.the.calibrated.parameters.are.listed. Bathymetry During.the.calibration.procedure.the.start.values.for.the.Stanton. and. Dalton. number. (0.0013). turned. out. to. be. sufficient.. The. The. lake’s. Bathymetry. can. be. seen. in. Figure. 4.12. and. 4.13.. background. viscosity. and. the. background. diffusivity. are. very. There.is.an.almost.smooth.curve.of.A[m²]=f(depth[m]).with.no. low. exceptional.bottom.shape..The.data.in.Figure 4.13.were.derived. from Figure 7-1. in. Regione. Piemonte. (2004). by. planimeter. measurements. Lake data Value Stanton.number 0.0013 Dalton.number 0.0013 Background.viscosity 1??10-7 Background.diffusivity 1??10-7 Water.volume 128.77.Mm³

Table 4.8 > Parameters for Lago di Viverone.

Figure 4.13 > The Bathymetrie of Lago di Viverone.

84 Figure 4.15 > Deviation statistics of measured minus calculated water temperatures in each month within the calibration at Lake Constance.

Wörthersee: Data, parameterization, and calibration results

The. Austrian. lake. Wörthersee. circulates. once. a. year.. It. is. Bathymetry described.as.a.meromictic.lake.in.literature.(Findenegg,.1933).. The. distance. in. a. straight. line. between. the. western. and. the. Recent. studies. showed. that. mixing. occurred. in. several. years. eastern.shores.is.about.16.3.km..The.width.is.in.the.range.of. within.the.period.1970-2002.(Schulz.et.al.,.2005)..Hence,.Lake. 1-1.5km.. There. is. a. deep. western. basin,. an. eastern. basin,. Wörthersee.is.considered.as.a.facultative.meromictic.lake..The. and.a.zone.between.these.two.basins..Each.of.these.basins.is. lakes.characteristics.are.listed.in Table 4.9. characterized.by.a.typical.orientation.of.the.lakes.shape..Figure. . 4.18.shows.the.bathymetry.of.the.entire.lake,.and.of.the.western. basin.

Figure 4.16 > Map of Lake Wörthersee; Source: Google Maps.

Figure 4.17 > The western basin of Lake Wörthersee. Lake data Value Mean.water.level 439.mNN Maximal.depth 85.2.m Mean.depth 41.9.m Surface.area 19.39.km² Water.volume 816.4.Mm³

Table 4.9 > Characteristics of Lake Wörthersee.

Climate data The.climate.station.in.Pörtschach,.a.village.at.the.banks.of.the. Wörthersee,.provided.data.of.air.humidity,.air.temperature,.wind. Figure 4.18 > The Bathymetry of Lake Wörthersee (entire lake and the western basin). velocity,. wind. direction. in. a. period. of. 01.08.1990-31.07.2010. with.a.time.series.interval.of.one.hour..The.position.of.the.station. . is.14°10’E,.46°28’N..The.measurements.of.cloud.cover.at.7:00h,. 14:00h,.and.19:00h.each.day.were.interpolated.linearly.onto.a. 1-hour-resolution.. The. 80. gaps. (longest. gap:. 118. hours). were. filled.by.a.regression.function.between.station.Pörtschach.and. the.station.Feldkirch..Station.Feldkirch.is.situated.around.10.km. NNW.from.Pörtschach.at.14°06’E.and.46°43’N.. 85 Secchi depth The assumption on the Secchi depth that was implemented in the model setup are based on measurements that were made in the period 1996-2010 by the Karinthian Institut for Lake Research (KIS) at the measure point Wörthersee Saag in the western basin. There is no clear annual course of the Secchi depth. These measurements were interpolated onto a daily resolution. Figure 4.20 > Measured mean monthly water temperatures at Lake Wörthersee in the With this daily values the mean of 3.95m was calculated. If the period 1930-2009. mean monthly values are considered (see Figure 4.19), there is a tendency to clearer water in summer. But in the measured period there was a decreasing secchi depth despite the fact that the water quality has improved since the 1970’s because of restoration actions. It is assumed, that the changing composition of algae to small species like Bacillariophyceae, Crysophyceae, and Chlorophyceae are responsible for this change (Schulz et al., 2005). However, this instationarity in the annual course of the secchi depth lead to the consideration of the above mentioned mean secchi depth of 3.95 m. In the later calibration procedure it turned out, that 3.5 m leads to better results.

Water temperature Data Water temperature measurements in the western basin are available in the period 1930-2010 for the present study. These are considered representative for the whole lake. Because of the availability of climate data at station Feldkirchen. To fill gaps in the time series of station Pörtschach it was necessary to look onto the period 1996-2009.

Figure 4.21 > Deviation statistics of measured minus calculated water temperatures in each month within the calibration at Lake Wörthersee.

Figure 4.19 > Mean monthly secchi depth in Lake Wörthersee (Saag) in the period 1996-2010.

Calibration Results Figure 4.20 shows the statistics of the deviation between the measured and modeled water temperatures at Lake Wörthersee after the calibration with the evolutionary algorithm. Because this lake has a very long shape, the question appeared, whether this could influence the goodness of recalculation in the one- dimensional model approach. To answer this, the western basin (see Figure 4.17) was calibrated separately. The results show just a little difference to the entire lake model. Comparing the measured and modeled mean monthly water temperatures, the sum of squared residuals for the whole lake is 236.8°C², and for Figure 4.22 > Deviation statistics of measured minus calculated water temperatures in each the western basin it is 273.4°C². Reasons for this little difference month within the calibration at the western basin of Lake Wörthersee. are the different solutions for the best parameter set when calibrating the two models separately. This test showes, that the 1dv model can be applied for the whole lake. The model for the whole lake Wörthersee worked a little bit better in this test.

86 Scenario development

The.study’s.third.step.is.the.development.of.climate.scenarios. General. circulation. models. (GCMs). are. forced. by. these. for. Lake. Constance,. Lake. Wörthersee,. and. Lago. di. Viverone.. scenarios..They.calculate.the.climate.under.these.scenarios.on. These.scenarios.allow.the.evaluation.of.the.lakes.or.rather.the. a.grid,.that.is.to.coarse.for.regional.use..That’s.why.there.are. reaction. of. their. mathematical. representation. if. the. climate. several.technics.to.downscale.this.GCM.information.onto.a.finer. conditions. are. changing.. This. is. a. way. of. handling. the. future. grid..In.general.there.are.two.ways:.dynamical.downscaling.with. if. the. uncertainties. of. “what. will. come”. are. very. high.. The. nested.regional.climate.models.that.uses.the.GCM.information. unknown.future.development.can’t.be.foreseen..But.if.a.bundle. as. forcing. and. boundary. condition,. whereas. the. statistic. of.scenarios.are.used,.were.each.of.them.is.very.unlikely,.they. approach.uses.statistical.correlations.between.local.and.global. capture.a.wide.range.of.possibilities..But.there.is.the.very.likely. climate.patterns. future.in.this.range.of.possible.future..However,.there.is.always. Actually.these.approaches.fail.to.reproduce.the.regional.climate. the.little.but.present.probability.of.natural.disasters.like.extreme. in. the. alpine. region. in. a. general. sufficient. and. reliable. quality. volcano.eruption,.for.instance.the.Pinatubo.eruption.in.the.year. (e.g..Smiatek.et.al..2009;.Fuentes.and.Heimann,.1996)..High- 1991.(e.g..Minnis.et.al.,.1993),.or.impact.of.meteorites..These. quality.forcing.data.are.crucial.for.impact.studies,.according.the. events.are.not.within.the.most.climate.scenarios..The.common. old.modeler’s.wisdom:.Garbage.in,.garbage.out. scenarios. are. based. on. greenhouse. gas. emissions,. technical. and. social. development,. as. well. as. economic. structures. and. That’s. why. another. way. was. chosen. in. this. study..It. is. based. their. global. patterns.. The. philosophy. in. sense. of. life. style. of. on. measured. time. series. of. climate. variables. at. a. lake.. In. so. the.future.human.beings.in.using.their.planet.will.also.play.an. called.“what.if”-scenarios.these.time.series.are.manipulated.in.a. important.role.in.the.complicated.and.complex.game.of.climate. way,.that.the.future.is.a.statistical.manipulation.of.the.past..For. forcing.variables. instance.one.manipulation.is.a.shift.in.the.variance..The.result.is.a. detailed.analysis.of.the.lakes.behavior.if.certain.climate.variables. The. IPCC. developed. the. SRES-Scenarios. (Nakicenovic. et. al.,. are.changed.onto.another.level..If.such.a.variable.is.changed.by. 2000).. These. scenarios. are. »…alternative. images. of. how. the. small.steps,.one.can.get.information.about.the.lakes.sensitivity.. future. might. unfold. and. are. an. appropriate. tool. with. which. One.major.advantage.is.that.the.results.are.decoupled.from.the. to. analyze. how. driving. forces. may. influence. future. emission. GCM.output.and.from.the.long.chain.of.errors.in.the.downscaling. outcomes. and. to. assess. the. associated. uncertainties.«. procedure. (Nakicenovic. et. al.,. 2000).. There. are. 40. SRES-Scenarios. that. cover.several.storylines,.grouped.in.the.families.A1,.A2,.B1,.and. The.“what.if”-.scenarios.in.the.present.study.are.described.in. B2..Most.impact.studies.use.illustrative.scenarios.out.of.each. the.following.for.each.lake.in.detail. scenario.family.because.of.technical.issues.like.calculation.time.

Lake Constance

The.scenarios.are.based.on.the.climate.data.of.the.years.1964- 2007.that.were.used.to.calibrate.the.model.

Air temperature The. characteristic. values. of. the. eleven. temperature. scenarios. are. listed. in. Table 4.10.. Scenario. Nr.. 1. is. a. continuation. of. Number 1 2 3 4 5 6 7 8 9 10 11 the. significant. linear. trend. in. the. temperature. record. that. was. Mean temperature [°C] * 10.6 11.6 12.6 13.6 14.6 15.6 9.6 8.6 7.6 6.6 observed.since.the.mid.of.the.year.1980.(see Figure 4.23)..This. Relative to 0 1 2 3 4 5 -1 -2 -3 -4 break.point.was.calculated.with.a.segmented.trend.analyses.of. measurement [°C] the.low-pass.filtered.times.series.of.air.temperature..There.are. Min 6.0 6.5 1.7 2.5 1.6 1.5 1.7 2.0 2.5 4.0 4.0

two.segments:.Between.January.1964.and.June.1980.there.is.a. Table 4.10 > Characteristic values of temperature scenarios at Lake Constance. temperature.increase.of.0.13°C.per.decade..After.this.date.there. * Linear continuation of linear trend (0.59°C per decade). is.an.air.temperature.increase.of.0.59°C.per.decade. For. the. other. temperature. scenarios. the. trends. in. the. air. temperature. time. series. was. removed. by. subtraction. of. the. low-pass. filter. (filter. window:. 10. years;. see. the. blue. line. in.. Figure 4.23).. The. resulting. temperature. fluctuations. finally. were.added.to.constant.mean.values.(second.row.in.Table.10).. Because.scenarios.no..2-11.are.based.on.this.detrending.their. lengths.depend.on.the.length.of.the.low-pass.filtered.time.series. Figure 4.23 > Segmented trend analysis of the low-pass filtered hourly air temperature at Lake of.33.3.years..So,.the.scenario.length.is.33.3.years..The.variation. Constance in the period 01.01.1964 - 31.12.2007. around. the. mean. is. the. information. between. 30th. December. 1968.and.2nd.April.2002.

87 Air humidity No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 There.are.16.scenarios.for.the.air.humidity.(rF).at. Mean rF [°C] * 79 80 81 82 83 85 87 89 78 77 76 75 73 71 69 Lake.Constance.(see Table 4.11)..Scenario.Nr.1. Relative to measured [°C] * 0 +1 +2 +3 +4 +6 +8 +10 -1 -2 -3 -4 -6 -8 -10 is.a.continuation.of.the.trend.in.the.amplitudes.of. the.air.humidity’s.annual.course.that.was.found. Table 4.11 > Characteristic values of air humidity scenarios at Lake Constance. * Continuation of the trend in the amplitude in annual rF fluctuation. in. the. low-pass. filtered,. and. measured. time. series.. Figure 4.24 shows. the. measurements. with. the. centered. moving. average. filtered. time.series.(rFLP)..Here,.the.window.was.365. days,.and.the.weights.were.normal.distributed. (standard. deviation. =. 4a).. Figure 4.25. shows. the. maxima. and. minima. in. the. rFLP..There. is. a. decreasing. trend. of. the. minimum,. and. an. increasing.trend.of.the.maximum. To.generate.scenario.Nr.1.at.first.rFLP.(the.black. Figure 4.24 > Time series of the relative air humidity and its low-pass filter in the period 1964-2007. The filter is a moving oscillating. line. in. Figure 4.24). was. subtracted. average with an normal distributed weighting in a 365d window (sd = 4 years)Constance in the period from. the. measured. one. to. get. the. pure. rF. 01.01.1964 – 31.12.2007 noise.(rFn).without.oscillation.and.trend..rFLP. was.stretched.in.a.second.step.parallel.to.the. axis. of. ordinates. in. the. way,. that. the. starting. amplitude.was.equal.to.the.ending.amplitude.of. the.original.rFLP..Finally,.rFn.was.added.to.the. scratched.rFLP. The. other. scenarios. (Nr.. 2-16). are. simple.

modifications. of. the. mean. by. adding. or. Figure 4.25 > Trends in the annual maxima and minima of the low-pass filtered time series of relative air humidity in the subtracting.an.offset.to.each.element.of.the.rF. period 1964-2007.The filter is a moving average with an normal distributed weighting in a 365d window (sd = 4 years) Constance in the period 01.01.1964 – 31.12.2007. time.series..It.must.be.noticed.that.this.offset. is. not. equal. to. the. values. in. the. third. row. of. Table 11..Because.if.increasing.the.mean.value. of.rF.the.resulting.values.>.100.are.set.to.100... No. 0 1 2 3 4 5 6 7 8 9 10

So,.this.iteratively.found.offset.is.higher. Mean [m/s] 0 0.35 0.85 1.37 2.35 2.85 3.35 3.85 4.35 4.85 1.85 No. Offset [m/s] -1.5 -1.0 -0.5 0.5 1.0 1.5 2.0 2.5 3.0 0 Wind speed wind There.are.10.scenarios.for.wind.speed.at.Lake. Table 4.12 > Characteristic values of wind speed scenarios at Lake Constance. Constance..In.the.measured.time.series.there. is. no. remarkable. trend.. In. the. last. decade. there.is.a.little.increase.of.mean.wind.velocity.. (see Figure 4.26).. But. it. is. not. a. significant. climatic.long-term.behavior..So,.the.scenarios. are.simple.shifts.in.the.mean.(see Table 4.12),. except.scenario.No..0.with.no.wind.at.all..

Cloud cover Figure 4.26 > Low-pass filter of the wind speed measurements in the period 1968-2002 (filter window = 10 years; Twelve.scenarios. describe. the. radiation. input. Filter type: moving average with normal distributed weighting). The original time series covers the period by.different.means.of.cloud.cover.[%cc]..They. 1964-2007. are.all.shifts.of.each.value..But.these.shifts.are. not. equal. to. the. changes. in. the. mean. (Table 4.13). because. values. >. 100%cc. and. <0%cc. Number 1 2 3 4 5 6 7 8 9 11 11 12 are.cut.to.100%cc.resp..0%cc..Based.on.the. Change in the mean 1 2 3 4 8 16 -1 -2 -3 -4 -8 -16 measurements.the.scenarios.cover.a.period.of. [% cloud cover]

43.years. Table 4.13 > Characteristic values of cloud cover scenarios at Lake Constance.

88 Lago di Viverone Number 1 2 3 4 5 6 7 8 9 10 11

The. scenarios. are. based. on. the. climate. data. Change [°C] +1 +2 +3 +4 +5 -1 -2 -3 -4 -5 0 of. the. years. 2002-2009,. that. were. used. to. Mean Tair[°C] 14.02 15.02 16.02 17.02 18.02 12.02 11.02 10.02 9.02 8.02 13.02 calibrate.the.model..This.period.of.eight.years.is. Table 4.14 > Characteristic values of air temperature scenarios at Lago di Viverone. too.short.to.determine.a.trend.which.indicates. a.change.in.climate.conditions..That’s.why.all. of.the.scenarios.for.Lago.di.Viverone.are.shifts. of.the.measured.time.series..All.scenarios.and. measurements.are.in.hourly.intervals.

Air temperature (Tair) Possible. future. mean. temperature. conditions. are.covered.by.eleven.air.temperature.scenarios. Figure 4.27 > Measured time series of air temperature at Lago di Viverone in the period 2002-2009. The range of the scenarios is indicated by the moving average-filtered time series of maximum and minimum shifts. (Table 4.14)..Scenario.No..11.is.the.measured. Tair. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Air humidity (rF) Change [%rF] +1 +2 +3 +4 +6 +8 +10 -1 -2 -3 -4 -6 -8 -10 0 Fifteen.scenarios.are.used.to.evaluate.the.lakes. Mean sensitivity. on. changes. in. relative. air. humidity. [%rF] 72.3 73.3 74.3 75.3 77.3 79.3 81.3 70.3 69.3 68.3 67.3 65.3 63.3 61.3 71.3 (Table 4.15).. The. mean. of. the. measured. time. series.(scenario.No..15).is.changed.in.a.range. Table 4.15 > Characteristic values of air temperature scenarios at Lago di Viverone. of.+-10%rF (see Figure 4.28)..

Wind velocity (Vw) Ten. scenarios. are. used. to. evaluate. the. lakes. sensitivity. on. changes. in. mean. wind. velocity. (Table 4.16).. The. mean. of. the. measured. time. series.(scenario.No..10).is.changed.in.a.range. of.-0.75.m/s.to.+3.0.m/s (see Figure 4.29)..In. Figure 4.28 > Measured time series of relative air humidity at Lago di Viverone in the period 2002-2009. The range scenario.9.there.is.approximately.no.wind. of the scenarios is indicated by the moving average filtered time series of maximum and minimum shifts.

Global Radiation (RG) No. 1 2 3 4 5 6 7 8 9 10

The. mean. global. radiation. is. varied. between. Change [m/s] +0.5 +1.0 +1.5 +2.0 +2.5 +3.0 -0.25 -0.5 -0.75 0 +50Wm-2. and. -50Wm-2. relative. to. the. Mean [m/s] 1.27 1.77 2.27 2.77 3.27 3.77 0.52 0.27 0.02 0.77 measurements. in. the. years. 2002-2009.. (see Figure 4.30)..There.are.11.scenarios.(see Table 4.16 > Characteristic values of wind velocity scenarios at Lago di Viverone. Table 4.17).. Scenario. No.. 11. is. the. measured. time.series.(reference.scenario).

Figure 4.29 > Measured time series of wind velocity at Lago di Viverone in the period 2002-2009. The range of the scenarios is indicated by the moving average filtered time series of maximum and minimum shifts.

No. 1 2 3 4 5 6 7 8 9 10 11

Change [m/s] +10 +20 +30 +40 +50 -10 -20 -30 -40 -50 0

Mean [m/s] 169.6 179.6 189.6 199.6 209.6 149.6 139.6 129.6 119.6 109.6 159.6

Table 4.17 > Characteristic values of global radiation scenarios at Lago di Viverone.

Figure 4.30 > Measured time series of global radiation at Lago di Viverone in the period 2002-2009 (hourly values). The89 range of the scenarios is indicated by the moving average filtered time series of maximum and minimum. shifts. Lake Wörthersee Number 1 2 3 4 5 6 7 8 9 10 11

The. scenarios. are. based. on. the. climate. data. Change [°C] +1 +2 +3 +4 +5 -1 -2 -3 -4 -5 0 of.14.years.(hourly.intervals).that.were.used.to. Mean Tair[°C] 10.44 11.44 12.44 13.44 14.44 8.44 7.44 6.44 5.44 4.44 9.44 calibrate.the.model..The.original.data.start.at.1st. Table 4.18 > Characteristic values of air temperature scenarios at Lake Wörthersee. July.1996.and.end.at.31st.July.2010..The.time. series. were. shortened. to. the. period. 01st. Jan. 1996.-.31st.Dec.2009.because.of.calculation.of. complete.years.for.mean.annual.statistics..This. period.is.too.short.to.determine.a.trend,.which. indicates.a.change.in.climate.conditions..That’s. why.all.of.the.scenarios.for.Lake.Wörthersee.are. shifts.of.the.measured.time.series..All.scenarios. and.measurements.are.in.hourly.intervals. Figure 4.31 > Measured time series of air temperature at Lake Wörthersee in the period 1996-2009. The range of the scenarios is indicated by the moving average filtered time series of maximum and minimum shifts. Air temperature Scenarios.of.possible.future.mean.temperature. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 conditions. are. covered. by. eleven. air. Change +1 +2 +3 +4 +6 +8 +10 -1 -2 -3 -4 -6 -8 -10 0 temperature. scenarios. (Table 4.18).. Scenario. [%rF] No..11.is.the.measured.Tair. Mean [%rF] 80 81 82 83 85 87 89 78 77 76 75 73 71 69 79

Air humidity Table 4.19 > Characteristic values of air temperature scenarios at Lake Wörthersee. Fifteen.scenarios.are.used.to.evaluate.the.lake’s. sensitivity. on. changes. in. relative. air. humidity. (Table 4.19).. The. mean. of. the. measured. time. series.(scenario.No..15).is.changed.in.a.range. of.+-10%rF (see Figure 4.32)..

Wind speed

Eleven.scenarios.are.used.to.evaluate.the.lakes. Figure 4.32 > Measured time series of relative air humidity at Lake Wörthersee in the period 2002-2009. The range sensitivity. on. changes. in. mean. wind. velocity. of the scenarios is indicated by the moving average filtered time series of maximum and minimum shifts. (Table 4.20).. The. mean. of. the. measured. time. series.(scenario.No..11).is.changed.in.a.range. No. 1 2 3 4 5 6 7 8 9 10 of.-0.75.m/s.to.+3.0.m/s.(see Figure 4.33)..In. Change [m/s] +0.5 +1.0 +1.5 +2.0 +2.5 +3.0 -0.25 -0.5 -0.75 0 scenario.9.there.is.approximately.no.wind. Mean [m/s] 1.27 1.77 2.27 2.77 3.27 3.77 0.52 0.27 0.02 0.77 Cloud cover Table 4.20 > Characteristic values of wind velocity scenarios at Lake Wörthersee. Twelve. scenarios. are. used. to. evaluate. the. lakes. sensitivity. on. changes. in. cloud. cover (Table 4.21).. The. mean. of. the. measured. time. series.(scenario.No..12).is.changed.in.a.range. of.52.8%cc.to.76.8%cc.(see Figure 4.34).

Figure 4.33 > Measured time series of wind velocity at Lake Wörthersee in the period 1996-2009. The range of the scenarios is indicated by the moving average filtered time series of maximum and minimum shifts

No. 1 2 3 4 5 6 7 8 9 10 11 12 Change [%cc] +1 +2 +3 +4 +8 +16 -1 -2 -3 -4 -8 +-0 Mean [%cc] 61.8 62.8 63.8 64.8 68.8 76.8 59.8 58.8 57.8 56.8 52.8 60.8

Table 4.21 > Characteristic values of cloud cover scenarios at Lake Wörthersee.

90 Figure 4.34 > Measured time series of cloud cover at Lake Wörthersee in the period 1996-2009. The range of the scenarios is indicated by the moving average filtered time series of maximum and minimum shifts.shifts. Tracer calculations

All. lakes. in. this. study. are. monomictic. lakes. that. circulate. Wahl. (2007). developed. a. mixing. index. to. evaluate. the. mixing. in. winter/spring.. It. is. one. aim. of. the. model. simulations. to. intensity.of.Lake.Constance..Equation.3.shows.the.definition.of. investigate. this. circulation. under. different. climate. conditions.. this.indicator. Because. the. intensity. of. mixing. is. crucial. for. the. oxygen. and. .. nutrient. concentrations. in. deep. water. layers. it. is. important. to. know.their.possible.future.developments. Pb.visuel (3) The. mixing. intensity. is. estimated. in. this. study. by. comparing. the. concentration. of. a. numerical. tracer. in. deep. and. surface. layers:. In. the. autumn. (1st. October, see Figure 4.35). the. tracer. is.introduced.into.the.upper.water.layer..Because.of.mixing,.the. If. Z. exceeds. 0.8. for. orthophosphate,. it. is. an. indicator. for. a. tracer.is.distributed.uniformly.in.the.epilimnion,.whose.thickness. good.vertical.mixing.(Wahl,.2007)..Now.it.was.tested,.whether. is. growing. until. the. beginning. of. complete. mixing.. At. 1st. of. mixing. is. reproduced. accurate. with. the. 1dv-model.. Therefore. May.it.is.very.likely.that.the.lake.has.mixed.completely..This.is. good. mixing. was. defined. in. the. results. of. the. 1dv-model. as. the.date.when.mixing.intensity.is.calculated.by.comparing.the. tracer.concentrations.higher.0.01.[-].in.a.depth.>.150m..Then,. tracer.concentrations.above.and.below.a.comparison.level..At. the. recalculated. goodness. of. mixing. in. the. years. 1970-2002. the.following.30th.September.the.tracer.is.removed.completely. of. Lake. Constance. were. compared. with. the. orthophosphate. to.start.a.new.tracer.evaluation.at.the.1st.October..The.mixing. measurements.of.Wahl.(2007). Figure 4.37.shows.the.comparison. intensity.is.described.by.a.mixing.index.xxxx...At.Lake.Constance. of.reproduced.good.or.bad.mixing.intensity.in.both.cases..It.can. it.is.calculated.by.Equation.2. be.seen,.that.the.model.reproduces.sufficiently.the.time.series. of.good.mixing. basse.déf. (2)

basse.déf. is.the.tracer.concentration.. below.100m.depth.

basse.déf. is.the.tracer.concentration.. Figure 4.37 > Good spring mixing defined by Wahl (2007) at Lake Constance derived from ortho- above.100m.depth. phosphate measurements, and reproduced by the 1dv model (tracer calculations). 1= good mixing, 0 = In the model calculations good mixing is defined as tracer concentrations > 0.01 [-] below 150m depth at 1st May.

basse.déf. At.Lake.Wörthersee.and.Lago.di.Viverone.the.level.for.. calculation.is.at.40m.depth.

Figure 4.35 > Tracer calculations at Lake Constance in the period 1970-2002. Good mixing is defined as tracer concentrations > 0.01 [-] below 150m water depth (dashed white line).

Figure 4.36 > An example for a year with good mixing followed by a year with less intensive mixing at Lake Constance. The white dashed lines indicate the relevant dates and depths for mixing index, and tracer calculations.

91 Results and Discussion

Results for Lake Constance

Results of air temperature scenario calculations at Lake Higher air humidity warms the surface water temperatures: Constance. An increase of 10%rF relative to the reference scenario Higher mean temperatures lead to less mixing indexes. E.g. (measurements) increases the stratified water by around 1°C in an increase in air temperature of about 2°C results in a halved the summer months (Figure 4.48). The relationship between air mixing index (Figure 4.38). humidity is approximately linear. The maximum mixing intensity is achieved if temperature This effect can be seen also in the bottom water temperatures decreases by -3°C relative to the reference measurement . (Figure 4.49). But at the bottom the temperature changes are (Figure 4.38). not that sensitive. The rF variation between +10%rF and -10%rF If temperature increases, there is a higher amount of evaporation relative to the reference scenario leads to a bottom water (E). Pb visuel (Figure 4.39) temperature variation between 4.25°C (+10%rF relative to the This increase in E is dominantly an effect of higher evaporation reference) and 4.0°C (-10%rF relative to the reference). in summer. While in summer less evaporation cause a lower Looking at the heat balance, there is an effect of only a few . -2 amount of evaporation than in the reference period, in winter W/m (Figure 4.50). The heat loss in winter is lower if air humidity there is an increase in E if the mean air temperature decreases is low. This is coupled to the low evaporation in these months (Figure 4.40). and scenarios. In each month the mean surface water temperatures change No changes in the position of the thermocline appear approximately in the same way as the mean air temperature (Figure 4.51). The variations in the winter months can be ignored (Figure 4.41). because the discussion about the thermocline is focused onto If the mean air temperature increases, the bottom water the stratified period (see also the discussion of thermocline temperature increases slightly. But a decrease of the mean air position calculation if air temperature conditions are changed). temperature reduces the bottom water temperature until the Results of wind velocity scenario calculations at Lake Constance maximum density at 4°C is reached (Tair change of ca. -1°C There is the highest sensitivity of mixing index if the mean wind relative to the reference data). velocity (VW) is changed in a range of -1m/s to +1m/s relative Between February and March there is a shift from heat loss to to measured VW (Figure 4.52). The maximum mixing index is heat profit (Figure 4.43). Contrariwise, between September and achieved if VW is equal or higher than +1.5m/s relative to the October the heat budget turns from positive to negative values. measurements in the period 1964-2007. Due to lower mean temperature, the latter mentioned budget The wind has a high influence onto the amount of evaporation (E) turn is shifted to the period between August and September. (Figure 4.53). An increase of 3m/s relative to the reference period Overall, the change in the monthly heat balance is small if mean increases E by 132 mm. But evaporation depends not only to VW air temperature changes (see the ordinate range in Figure 4.43). because of e.g. free convection: If there is no wind (scenario 0), . If the mean air temperature is changed, the position of the the mean annual E is around 436 mm at Lake Constance. thermocline does not change (Figure 4.44). There is no distinct This dependence can also be found in the mean monthly E temperature profile in winter. The changes that can be seen (Figure 4.53). Some special behavior appears in April. Against in the period November to April are caused by the algorithm the above mentioned connection, there is a declining E if the that searches the position of the highest temperature gradient. wind velocity exceeds an offset of +1m/s. One possible reason In those cases there is no characteristic metalimnion. In. is the additional influence of water temperature onto E. If the Figure 4.44 the stratification period is relevant (May-October). wind velocity increases, the surface water layer cools down. Results of air humidity scenario calculations at Lake Constance E declines because of colder water. In April there is possibly The scenario calculations show a decreasing mixing intensity if interplay between wind velocity and water temperature that air humidity rises (Figure 4.45). Low mean air humidity is coupled leads to a different influence of this variables in high and low with high mixing intensities. The sensitivity of mixing is slightly VW ranges. higher in low air humidity change ranges. Generally, there is a warming of the surface water if VW declines The higher the water vapor concentration in the air, the less is (Figure 4.55). If there is no wind the maximum temperature is the amount of mean annual evaporation (Figure 4.46). The mean 22.1°C in August. An extreme increase of mean VW to +3m/s annual amount of evaporation is 547.1 mm in the reference relative to measurements in the period 1964-2007 in August period 1968-2002. The variation of mean air humidity in the cools the water to 18.2°C. range of -10/+10%rF leads to a evaporation range of 578.6 mm The bottom water temperature around 4°C does not change if to 525.3 mm. This relationship is approximately linear. mean VW exceeds the mean within the measured period (Figure Comparing the mean monthly evaporation under varying air 4.56). But it increases up to 5°C if VW is lower, or if it is zero. humidity conditions, it can be concluded that the mean annual The heat fluxes react more sensitive to a change of mean VW behavior reflects the monthly behavior (Figure 4.47). than of rF (Figure 4.57). The change from negative to positive In the rF trend scenario with increasing air humidity amplitudes heat fluxes in February and March is not affected by the variation there is a strong drop of winter evaporation, whereas in the of VW. The change from positive to negative heat fluxes is in the period February - September this scenario has no remarkable months August and September, if the VW is low. Otherwise it is effect on the amount of evaporation. shifted to the months September and October. 92 During. the. stratification. period. the. thermocline. depth. varies. January.there.is.a.reversed.relationship:.The.more.clouds,.the. in.a.range.of.6.-.14m,.with.6m.in.May.and.14m.in.September. higher.is.E..This.effect.is.explained.by.cloud.as.a.heat.source.in. (Figure 4.58).. Higher. VW. +3m/s. (relative. to. reference). enlarge. winter.(long.wave.radiation)..In.summer,.when.clouds.have.an.E. the.epilimnion.up.to.a.range.of.9-17m.in.this.period..Figure 4.58. reduction.effect,.the.short.wave.radiation.is.dominant. shows.also.that.no.wind.or.a.wind.that.is.lower.than.the.reference. The.more.clouds,.the.colder.is.the.surface.water (Figure 4.62).. scenario.has.no.effect.onto.the.depth.of.the.thermocline..There. This.effect.is.higher.in.summer,.where.e.g..in.July.the.variation.of. is. an. epilimnion. depth. in. a. range. of. 6-14. in. the. period. May- cc.between.-16%cc.and.+16%cc.relative.to.the.measurements. September.without.wind..The.radiation.penetration.is.responsible. changes.the.surface.water.temperature.in.the.range.of.17.7°C. for.the.epilimnion.formation.in.the.cases.with.no/less.wind. to.20.7°C. The. bottom. water. temperature. cools. down. if. more. clouds. Results. of. cloud. cover. (cc). scenario. calculations. at. Lake. cover.the.sky..But.this.effect.is.very.little:.The.range.of.water. Constance temperature.change.is.between.4.2°C.and.4.0°C (Figure 4.63). The.mixing.intensity.depends.on.the.stratification.stability.that. The. above. mentioned. process. of. cloud. cover. as. heat. source. is. higher. if. high. radiation. input. increases. the. surface. water. in. winter. can. be. found. in. the. reaction. of. heat. budget. under. temperature.. Figure 4.59. shows. this. fact. at. Lake. Constance:. changing.cloud.cover.conditions:.While.in.the.months.of.negative. High.cc.lead.to.less.energy.input.into.the.surface.water.layers.. heat.budget.(and.also.in.the.transition.times).higher.cc.reduces. Thus,.the.mixing.index.rises.because.less.energy.is.necessary.to. the.heat.loss,.it.reduces.the.heat.input.in.summer.(Figure 4.64). circulate.the.water.against.buoyancy.effects. Both.effects.are.dominant.in.different.cc.ranges.in.February,.and. The. less. energy. input. (=higher. cc),. the. less. is. the. amount. of. they.produce.a.hyperbolic.kind.of.relationship. evaporation (Figure 4.60).. While. E. is. around. 553. mm. (1965- The. impact. of. radiation. onto. the. depth. of. thermocline. can. 2007).if.the.lake.is.forced.by.measured.cc,.it.changes.in.a.range. be. seen. in. Figure 4.65.. If. cc. is. changed. to. -16%cc. (relative. of. 328-716mm.if. cc. is. varied. between. +16%cc. and. -16%cc. to. reference). the. thermocline. is. up. to. 3m. deeper.. Higher. cc. relative.to.the.measurements. (+16%cc. relative. to. reference). reduces. the. thickness. of. the. Looking.at.the.mean.monthly.evaporation.amount.this.behavior. epilimnion.. Little. variations. between. -4%cc. and. +4%cc. don’t. can. be. found. in. the. months. February-October.. In. November- have.an.effect.

Results at Lago di Viverone

Results. of. air. temperature. scenario. calculations. at. Lago. di. reduces.it.as.well..The.maximum.heat.profit.in.February.around. Viverone a.change.of.0°C.(relative.to.the.period.2002-2009).is.shifted.in. The. mixing. index. at. Lago. di. Viverone. changes. only. (it. goes. March.to .a.Tair-change.of.-2.5°C.in.March.and.-5°C.in.April..In. down). if. the. air. temperature. rises. (Figure 4.66).. Lower. mean. the. period. May-July. there. is. a. clear. connection. between. low. temperatures. have. no. effect.. The. highest. gradient. of. mixing. heat.input.and.high.mean.air.temperatures..In.the.transition.from. index.as.a.function.of.mean.Tair.is.around.+2.5°C.relative.to.the. a.positive.to.a.negative.heat.balance.in.August/September.this. measurements.in.the.period.2002-2009. connection.is.turned.into.minimum.heat.loss.at.high.mean.Tair. The. mean. annual. evaporation. amount. within. the. years. 2002- Figure 4.72. shows. no. change. of. the. thermocline. position. in. 2009.was.585.3.mm..The.change.of.the.amount.of.evaporation. the. stratification. period. if. Tair.changes.. Some. exceptions. are. neglected. because. of. the. errors. in. the. thermocline. position. is.dE/dTair.≅.19mm/°C (Figure 4.67) in.the.range.of.-3°C.to.+5°C. relative.to.the.period.2002-2009..dE/dTair.is.about.7.4.mm/°C. definition.algorithm.in.winter. between.changes.of.-3.to.-5. Results.of.air.humidity.scenario.calculations.at.Lago.di.Viverone High. summer. evaporation. causes. this. effect. (Figure 4.68).. In. A.change.in.the.mean.air.humidity.has.no.effect.onto.the.mixing. could.months.(especially.in.December).dE/dTair.is.approximately. intensity.at.Lago.di.Viverone (Figure 4.73). zero.or.vary.in.a.range.of.a.1-2.mm..Considering.the.model.errors,. Contrary. to. the. expectation. of. less. evaporation. (E). if. there. is. there.should.not.be.much.emphasis.onto.this.little.variations. higher. air. humidity,. the. model. results. show. an. increasing. E. There.is.a.strong.linear.correlation.between.Tair.and.the.surface. (Figure 4.74).. Looking. at. the. monthly. resolution. (Figure 4.75),. water. temperature.. E.g.. dTsurface. water/dTair. is. around. this.can.be.traced.back.to.the.months.May-September..Between. 0.83°C/°C.in.August.and.1.0°C/°C.in.January. November.and.January.we.see.the.expected.high.E.at.less.mean. The.bottom.water.temperature.reaction.onto.changes.in.mean. r F. . At. the. other. transition. months. February,. March,. April,. and. Tair.can.be.divided.into.two.sections:.dTair.=.-5°C.to.+1°C.and. October.there.seems.to.be.some.processes.that.superimposes. dTair.=.+1°C.to.+5°C.(Figure 4.70)..In.the.first.range.there.is.no. each.other. or.just.a.little.reaction.of.Tbottom.water.onto.changes.in.Tair.. Higher.air.humidity.warms.up.the.surface.water.temperatures:. Increasing. Tair.into. the. second. range. a. significant. reaction. in. An. increase. of. 10%rF. relative. to. the. reference.. scenario Tbottom. water. can. be. seen.. But. compared. with. the. surface. (measurements).increases.the.stratified.water.in.by.around.1.2°C. water. temperatures. this. reaction. is. little. (Tbottom. water. is. in.the.summer.months (Figure 4.76)..The.relationship.between. between.4.0°C.and.4.6°C). air.humidity.is.approximately.linear. While.the.special.behavior.of.the.heat.balance.at.Lake.Constance. This.effect.can.be.seen.also.in.the.bottom.water.temperatures. can. be. explained. by. atmospheric. long. radiation,. it. is. quite. (Figure 4.77).. But. here. the. temperature. changes. are. not. that. different.at.the.much.smaller.Lago.di.Viverone.(Figure 4.71):.From. sensitive..The.rF.variation.between.+10%rF.and.-10%rF.relative. January.to.February.there.is.a.transition.from.heat.loss.to.heat. to.the.reference.scenario.leads.to.a.bottom.water.temperature. profit..An.increasing.mean.air.temperature.in.February.reduces. variation. between. 4.14°C. (+10%rF. relative. to. the. reference;. the.positive.heat.balance..A.reduction.of.mean.air.temperature.

93 December) and 4.0°C (-10%rF relative to the reference). During the stratification period the thermocline depth varies in Concerning heat balance, there is an effect of only a few . a range of 6 – 11m, with 6m in May and 11m in October (Figure W/m-2 e(Figur 4.78). In general, the heat loss in winter reacts 4.86). Higher VW of +3m/s relative to reference enlarge this more sensitive than the heat-benefit in summer. range up to 10-14m. Figure 4.86 shows also that no wind or a No changes in the position of the thermocline appear (Figure wind that is lower than the reference scenario has no effect onto 4.79). Some little variations can be ignored because they the depth of the thermocline. Without wind there is an epilimnion are artifacts of some difficulties in the thermocline position depth in a range of 6-11 in the period May-October. The radiation calculation algorithm. penetration is responsible for the epilimnion formation in the cases with no/less wind. Results of wind velocity scenario calculations at Lago di Viverone There is the highest sensitivity of mixing index if the mean wind Results of global radiation (RG) scenario calculations at Lago di velocity (VW) is changed in a range of -0.75m/s to 0/s relative Viverone to measured VW (Figure 4.80). The maximum mixing index is The mixing intensity reacts on a change of RG in a range that is achieved if VW is equal or higher than the measurements in the higher than +10W/m² relative to the measurements (Figure 4.87). period 1964-2007. This is the result of higher stratification stability if high radiation The mean annual amount of evaporation (E) reacts sensitive to a input increases the surface water temperature. Thus, increasing change of mean wind speed (Figure 4.81). An increase of 3m/s radiation causes a lower water density in the epilimnion. relative to the reference period increases E from 585 to 732 mm. The less energy input due to radiation, the less is the amount But evaporation depends not only to VW because of e.g. free of evaporation (Figure 4.60). In this case the mean annual convection: If there is no wind (scenario 0), the mean annual E is evaporation amount (E) is approximately linear coupled with the around 502mm at Lago di Viverone. shortwave radiation input. While E is around 585 mm (2002-2009) This dependence can also be found in the mean monthly E if the lake is forced by measured global radiation, it changes into (Figure 4.82). Some special behavior appears in April. E declines a range of 226-945mm if RG is varied between -50W/m² and if the wind velocity exceeds an offset of +2m/s. One possible +50W/m² relative to the measurements. reason is (as it was discussed for Lake Constance in the latter If the mean annual amount of evaporation is resolved to the text) the additional influence of water temperature onto E. If the monthly E, this relationship between short wave radiation input wind velocity increases, the surface water layer cools down. and evaporation can be found in each month (Figure 4.89). This E declines because of colder water. In April there is possibly effect is dominated by evaporation in the warm season because an interplay between wind velocity and water temperature that of up to ten times higher evaporation amounts in summer. leads to different influence of these variables in high and low The higher the mean RG, the warmer is the surface water . VW ranges. (Figure 4.90). E.g. in July the variation of RG between -50W/m² From May to December there is a warming of the surface water and +50W/m² relative to the measurements changes the surface if VW declines (Figure 4.83). In these months this connection water temperature in the range of 22.38°C to 28.85°C. is valid for the whole variation range. If there is no wind, the The bottom water temperature cools down as well if less maximum temperature is 28.2°C in July. An extreme increase of radiation energy is absorbed by the water. But this effect is mean VW to +3m/s (relative to measurements) cools the water very little and limited at the density maximum of 4°C. The range down to 23.6°C in August. From January to April, when the of water temperature change is between 4.5°C and 4.0°C . water temperature is around or below 4°C, there are ranges of (Figure 4.91). wind velocity changes where the surface water is proportional to High input radiation causes a positive heat balance. This is only wind velocity changes. One possible explanation is: The warmer partially true for Lago di Viverone: If RG is up to 50W/m² higher water from the deep water layers is transported to the colder than in the measurements, higher heat budget is observed in the surface. The amount of this heat transport is proportional to the model calculations (Figure 4.92). But if RG is reduced down to wind velocity. In summer the heat transport is directed into the -50W/m² relative to the measurements, the mean monthly heat opposite direction. budget rises in the warm season, too (in the model calculations). This effect can be seen in the bottom water temperatures as Perhaps, the reason for this unexpected result can be found in the well (Figure 4.84): In winter more heat is transported to the circumstance that in all RG-scenarios the other input variables surface if VW is higher. In summer the heat flux is directed to the are constant, or, to be more precise, the air temperatures are hypolimnion if wind velocity is increasing. equal to the measurements in the period 2002-2009. The heat fluxes at the lake surface react more sensitive to a The importance of radiation in the development and shape of change of mean VW than of rF (Figure 4.85). The change from the epilimnion can be seen in Figure 4.93. If RG is changed to negative to positive heat fluxes in January and February is not -50W/m² (relative to the reference period), the thermocline, e.g. affected by the variation of VW. In August and October the in July, is up to 2m deeper than in the measured period. transition from positive to negative heat fluxes is not affected by changing wind velocities.

Results at Lake Wörthersee

Results of air temperature scenario calculations at Lake +1°C relative to the measurements in the period 1996-2009 Wörthersee (Figure 4.94). Lower mean temperatures have no effect onto the The mixing index at Lake Wörthersee changes only (it goes circulation in winter/spring. The highest gradient of mixing index down) if the mean air temperature exceeds approximately IndM=f(Tair) is around +3.5°C relative to the measurements.

94 The. mean. annual. evaporation. amount. within. the. years. 1996- At.the.bottom.of.a.lake.no.clear.signal.of.an.rF-change.is.detected. 2009.was.540.mm..The.change.of.the.amount.of.evaporation.is. (Figure 4.105)..The.values.lie.within.the.range.of.4-4.05°C. dE/dTair.→.16mm/°C.(Figure.4.95).in.the.range.of.-5°C.to.+5°C. Concerning.heat.balance,.there.is.an.effect.of.only.a.few.W/m-2. relative.to.the.period.1996-2009. (Figure 4.106)..In.general,.the.heat.loss.in.winter.is.more.sensitive. High. summer. evaporation. causes. this. effect. (Figure 4.96).. In. than.heat.benefit.in.summer. could.months.(especially.in.December).dE/dTair.is.roughly.not. No. changes. in. the. position. of. the. thermocline. appear constant.over.the.range.of.the.temperature.changes..In.some. (Figure 4.107)..The.little.variations.in.November.can.be.ignored. months. this. gradient. is. negative.. Considering. the. model. and. because.they.are.artifacts.of.some.difficulties.in.the.thermocline. data.errors.and.the.relative.little.evaporation,.there.should.not. position.calculation.algorithm. be.much.emphasis.onto.this.issue. Results. of. wind. velocity. (VW). scenario. calculations. at. Lake. There.is.a.strong.correlation.between.Tair.and.the.surface.water. Wörthersee temperature..E.g..dTsurface.water/dTair.is.around.0.83°C/°C.in. August.and.0.94°C/°C.in.January..In.the.warm.season.Tsurface. There.is.the.highest.sensitivity.of.mixing.index.if.the.mean.VW. water=f(dTair).is.approximately.linear..From.January.to.May.there. is.changed.in.a.range.of.-1m/s.to.0m/s.relative.to.measured.VW. are.different.sections.within.this.function (Figure 4.97). (Figure 4.108)..The.maximum.mixing.index.is.achieved.if.VW.is. The.bottom.water.temperature.reaction.onto.changes.in.mean. equal.or.higher.than.the.measurements.in.the.period.1996-2009. Tair.can.be.divided.into.two.sections:.dTair.=.-5°C.to.+3°C.and. The. mean. annual. amount. of. evaporation. (E). reacts. sensitive. dTair.=.+3°C.to.+5°C (Figure 4.70)..In.the.first.range.there.is.no. to.a.change.of.mean.wind.speed (Figure 4.81)..An.increase.of. or.just.a.little.reaction.of.Tbottom.water.onto.changes.in.Tair..If. 3m/s.relative.to.the.reference.period.increases.E.from.540mm. Tair.is. increased. into. the. second. range,. a. significant. reaction. to.617mm..But.evaporation.depends.not.only.to.VW.because. in.Tbottom.water.can.be.seen..But.compared.with.the .surface. of.e.g..free.convection:.If.there.is.no.wind,.the.mean.annual.E.is. water. temperatures,. this. reaction. is. little. (Tbottom. water. is. around.504mm.at.Lake.Wörthersee. between.4.0°C.and.4.3°C). This. dependence. can. also. be. found. in. the. mean. monthly. E. The. behavior. of. heat. balance. is. similar. to. the. that. at. Lago. di. (Figure 4.110)..Some.special.behavior.appears.in.February-April.. Viverone.(compare.Figure 4.99.and.Figure 4.71):.From.January. In.some.variation.ranges.there.is.a .declining.of.E..One.possible. to. February. there. is. a. transition. from. heat. loss. to. heat. profit.. reason.is.(as.it.was.discussed.for.Lake.Constance.in.the.latter. An. increasing. mean. air. temperature. in. February. reduces. the. text).the.additional.influence.of.water.temperature.onto.E..If.the. positive.heat.balance.as.well.a.reduction.of.mean.air.temperature. wind.velocity.increases.then.the.surface.water.layer.cools.down.. does..The.maximum.heat.profit.in.February.around.a.change.of. E. declines. because. of. colder. water..In. February-April. there. is. 1°C.(relative.to.the.period.1996-2009).is.shifted.in.March.to.a. possibly.interplay.between.wind.velocity.and.water.temperature. Tair-change. of. -2°C. in. March. and. -3°C. in. April.. In. the. period. that.leads.to.different.influence.of.this.variables.in.high.and.low. May-July.there.is.a.clear.connection.between.low.heat.input.and. VW.ranges. high.mean.air.temperatures..In.the.transition.from.a.positive.to. There. is. a. warming. of. the. surface. water. if. VW. declines.. a.negative.heat.balance.in.August/September.it.is.turned.into. (Figure 4.111)..If.there.is.no.wind.the.maximum.temperature.is. minimum.heat.loss.at.high.mean.Tair. 25.4°C.in.August..An.extreme.increase.of.mean.VW.to.+3m/s. Figure. 4.100. shows. no. change. of. the. thermocline. position. in. (relative.to.measurements).cools.the.water.down.to.21.7°C.in. the. stratification. period. if. Tair.changes.. Some. exceptions. are. this.month..From.January.to.March,.when.the.water.temperature. neglected.because.of.errors.in.the.thermocline.position.definition. is.below.4°C,.there.are.ranges.of.wind.velocity.changes.in.which. algorithm.in.winter. the.surface.water.is.proportional.to.wind.velocity.changes..One. possible. explanation:. the. warmer. water. from. the. deep. water. Results.of.air.humidity.scenario.calculations.at.Lake.Wörthersee layers.is.transported.to.the.colder.surface..The.amount.of.this. A.change.in.the.mean.air.humidity.has.negligible.effects.onto.the. heat.transport.is.proportional.to.the.wind.velocity. mixing.intensity.(Figure 4.101;.see.the.scale.of.y-axis). This. effect. can. be. seen. in. the. bottom. water. temperatures. as. There.is.a.less.evaporation.(E).if.the.mean.air.humidity.is.higher. well (Figure 4.112):.In.winter.more.heat.is.transported.from.the. than.in.the.reference.period.1996-2009 (Figure 4.102)..But.the. bottom.to.the.surface.if.VW.is.higher. change.is.very.little:.e.g..a.change.of.+10%rF.leads.to.an.increase. The.heat.fluxes.at.the.lake’s.surface.react.more.sensitive.to.a. of.about.3.48.mm/a..If.the.mean.rF.is.changed.to.smaller.values,. change. of. mean. VW. than. on. changes. in. rF. (compare. Figure the. E. is. increased. likewise.. Looking. at. the. monthly. resolution. 4.106.with.Figure 4.113)..The.change.from.negative.to.positive. (Figure 4.103).one.can.see.that.these.two.sections.in.the.function. heat. fluxes. in. January. and. February. is. not. affected. by. the. E=f(cange.of.%).are.a.combination.of.different.E.behaviors.in. variation. of. VW.. In. August. and. October. the. transition. from. the.periods.February-July.and.August-January..Between.August. positive.to.negative.heat.fluxes.is.not.affected.by.changing.wind. and.January.we.see.the.expected.high.E.at.less.mean.r F. .In.the. velocities,.too. other.months.some.processes.seems.to.be.which.superimposes. During. the. beginning. of. the. stratification. period. (March- each.other..Regarding.the.little.sensitivity,.the.little.changes.of.E,. June). and. the. end. of. this. period. (September-November). the. and.the.model.uncertainties,.the.conclusion.is,.that.there.is.no/ thermocline.depth.varies.in.a.range.of.0.–.17m (Figure 4.114) little.change.of.mean.annual.and.monthly.E.if.the.air.humidity. if. VW. is. changed.. The. higher. the. wind. velocity,.the. deeper. is. changes. the. thermocline. position.. In. the. months. of. warmest. surface. Higher.air.humidity.warms.up.the.surface.water.temperatures:. water. temperatures. (July. and. August). there. is. no. reaction. on. An. increase. of. 10%rF. relative. to. the. reference. scenario. changing. wind. velocities. because. of. stratification. stabilities. (measurements). increases. the. surface. water. by. around. 1°C. which. inhibit. wind. induced. deeper. surface. mixing.. In. January. in. the. summer. months. (Figure 4.104).. The. relationship. is. there.is.no.stratification..Hence,.the.plot.for.January.shows.not. approximately.linear. a. thermocline. position. but. even. it. shows. the. position. of. the. highest.temperature.gradient.

95 Results of cloud cover (cc) scenario calculations at Lake The lower the mean cc, the warmer is the surface water Wörthersee (Figure 4.118). E.g. in August the variation of cc between -8%cc The mixing intensity reacts on a change of cc in a range which is and +16%cc relative to the measurements changes the surface below -2%cc relative to the measurements (Figure 4.115). This water temperature in the range of 24.65°C to 22.24°C. is the result of higher and longer stratification stability if high The bottom water temperature cools down as well if less radiation input increases the surface water temperature. radiation energy is absorbed by the water (Figure 4.119). But The less energy input due to clouds, the less is the amount this effect is very little and these changes are in a range of a few of evaporation (Figure 4.116). In this case, the mean annual 0.01°K around 4°C. evaporation amount (E) is coupled approximately linear, and Low cc (high input radiation) increases the positive heat inverse to cc. While E is around 540 mm (1996-2009) if the lake budget in the period February-July (Figure 4.120). Within the is forced by measured cc, it changes into a range of 376mm- period September-January there is a reciprocal reaction: high 610mm if cc is varied within +16%cc and -8%cc relative to the cc reduces the negative heat budget. In the transition month measurements. August (positive to negative heat budget) both processes are If the mean annual amount of evaporation is resolved to the combined. monthly E, this relationship between cc and evaporation can Concerning the mean monthly thermocline position (tcp), there be found in almost each month (Figure 4.117). In November is no reaction if cc is changed in a range of -8%cc to +16%cc and December it is reciprocal: The more clouds, the higher in the period January-April (Figure 4.121). In the other months is E. Beside the fact of small evaporation and sensitivity one (stratification period) there is an indication for lower tcp if there conjecture is that in winter cloud emitted long wave radiation is are more clouds. But this change is within a narrow range of an energy source for evaporation. 1-2m.

Conclusion

To assess the reaction of lakes in the alpine region three lakes are different reactions if air temperature, air humidity, cloud cover, modeled with a one dimensional, hydrodynamic model. Several and wind velocity are changed. But the direction of sensitivity scenarios for air temperature, air humidity, cloud cover/global is always the same: The higher the temperatures are, the radiation, and wind velocity are used to identify a lakes sensitive worse is the mixing intensity. High relative air humidity causes behavior under changing climate conditions. The scenarios are worse mixing, too. Higher wind velocities enhance the mixing based on measured time series. Changes of statistical properties processes in winter/spring. (e.g. mean value) lead to a realistic forcing that is independent These and more results show the lakes very sensitive processes. from regional climate projections. Small variations in the climate conditions have remarkable The calculations show differences between the small lakes and effects. The results of HCILS can be used to assess a lake actual the big Lake Constance. Especially the mixing intensity shows state, and they show what can be expected in the future.

96 Appendix to the results of HCILS: Figures Results of air temperature scenario calculations at Lake Constance

Table 4.38 > Mixing index of Lake Constance Table 4.39 > Mean annual amount of if the mean air temperature varies evaporation of Lake Constance between -4 and +5°C relative if the mean air temperature varies to the detrended air temperatures between -4 and +5°C relative to in the period 1968-2002. the detrended air temperatures in the period 1968-2002. Table 4.40 > Mean monthly amount of evaporation at Lake Constance if the mean air temperature varies between -4 and +5°C relative to the detrended air temperatures in the period 1968-2002.

Table 4.41 > Mean monthly surface water temperature at Lake Constance if the mean Table 4.42 > Mean monthly bottom water temperature at Lake Constance if the mean air temperature varies between -4 and +5°C relative to the detrended air air temperature varies between -4 and +5°C relative to the detrended air temperatures in the period 1968-2002. temperatures in the period 1968-2002.

Table 4.43 > Mean monthly heat budget at Lake Constance if the mean air temperature varies Table 4.44 > Mean monthly position of thermocline at Lake Constance if the mean air 97 between -4 and +5°C relative to the detrended air temperatures in the period temperature varies between -4 and +5°C relative to the detrended air 1968-2002. Negative values are heat losses to the atmosphere, positive values temperatures in the period 1968-2002. The values here are the depths describe a heat input into the water. below water surface of the mean monthly maximum temperature gradient. Results of air humidity scenario calculations at Lake Constance

Table 4.45 > Mixing index of Lake Constance Table 4.46 > Mean annual amount of if the mean air humidity varies evaporation of Lake Constance between -10 and +10%rF relative if the mean air humidity varies to the air humidity in the period between -10 and +10%rF relative 1968-2002. to the air humidity in the period 1968-2002. Table 4.47 > Mean monthly amount of evaporation at Lake Constance if the mean air humidity varies between -10 and +10°C relative to the air humidity in the period 1968-2002.

Table 4.48 > Mean monthly surface water temperature at Lake Constance if the mean air Table 4.49 > Mean monthly bottom water temperature at Lake Constance if the mean air humidity varies between -10 and +10°C relative to the detrended air humidity varies between -10 and +10%rF relative to the air humidity in the temperatures in the period 1968-2002. period 1968-2002.

Table 4.50 > Mean monthly heat budget at Lake Constance if the mean air humidity varies Table 4.51 > Mean monthly position of thermocline at Lake Constance if the mean air between -10 and +10°C relative to the air humidity in the period 1968-2002. humidity varies between -4 and +5°C relative to the air humidity in the period Negative values are heat losses to the atmosphere, positive values describe 1968-2002. The values here are the depths below water surface of the mean a heat input into the water. monthly maximum temperature gradient.

98 Results of wind velocity scenario calculations at Lake Constance

Table 4.52 > Mixing index of Lake Constance Table 4.53 > Mean annual amount of if the mean wind velocity varies evaporation of Lake Constance between -1.85 and +3m/s relative if the mean wind velocity varies to the wind velocity in the period between -1.85 and +3m/s relative 1964-2007. to the wind velocity in the period 1964-2007. Table 4.54 > Mean monthly amount of evaporation at Lake Constance if the mean wind velocity varies between -1.85 and +3m/s relative to the wind velocity in the period 1964-2007.

Table 4.55 > Mean monthly surface water temperature at Lake Constance if the mean wind Table 4.56 > Mean monthly bottom water temperature at Lake Constance if the mean wind velocity varies between -1.85 and +3m/s relative to the wind velocity in the velocity varies between -1.85 and +3m/s relative to the wind velocity in the period 1964-2007. period 1964-2007.

Table 4.57 > Mean monthly heat budget at Lake Constance if the mean wind velocity varies Table 4.58 > Mean monthly position of thermocline at Lake Constance if the mean wind between -1.85 and +3m/s relative to the wind velocity in the period 1964-2007. velocity varies between -1.85 and +3m/s relative to the wind velocity in the period 1964-2007. The values here are the depths below water surface of the mean monthly maximum temperature gradient.

99 Results of cloud cover scenario calculations at Lake Constance

Table 4.59 > Mixing index of Lake Constance if Table 4.60 > Mean annual amount of evaporation the mean cloud cover (cc) varies of Lake Constance if the mean cloud between -16 and +16%cc relative cover (cc) varies between -16 and to the mean cloud cover in the +16%cc relative to the mean cloud period 1965-2007. cover in the period 1965-2007.

Table 4.61 > Mean monthly amount of evaporation at Lake Constance if the mean cloud cover (cc) varies between -16 and +16%cc relative to the mean cloud cover in the period 1965-2007.

Table 4.62 > Mean monthly surface water temperature at Lake Constance if the mean cloud Table 4.63 > Mean monthly bottom water temperature at Lake Constance if the mean cloud cover (cc) varies between -16 and +16%cc relative to the mean cloud cover in cover (cc) varies between -16 and +16%cc relative to the mean cloud cover the period 1965-2007. in the period 1965-2007.

Table 4.64 > Mean monthly heat budget at Lake Constance if the mean cloud cover (cc) varies Table 4.65 > Mean monthly position of thermocline at Lake Constance if the mean cloud between -16 and +16%cc relative to the mean cloud cover in the period 1965- cover (cc) varies between -16 and +16%cc relative to the mean cloud cover in 2007. Negative values are heat losses to the atmosphere, positive values describe the period 1965-2007.The values here are the depths below water surface of a heat input into the water. the mean monthly maximum temperature gradient.

100 Results of air temperature scenario calculations at Lago di Viverone

Table 4.66 > Mixing index of Lago di Viverone Table 4.67 > Mean annual amount of evapora- if the mean air temperature varies tion of Lago di Viverone if the mean between -5 and +5°C relative to air temperature varies between -5 the detrended air temperatures in and +5°C relative to the detrended the period 2002-2009. air temperatures in the period 2002-2009. Table 4.68 > Mean monthly amount of evaporation at Lago di Viverone if the mean air temperature varies between -5 and +5°C relative to the detrended air temperatures in the period 2002-2009.

Table 4.69 > Mean monthly surface water temperature at Lago di Viverone if the mean air Table 4.70 > Mean monthly bottom water temperature at Lago di Viverone if the mean air temperature varies between -5 and +5°C relative to the detrended air temperature varies between -5 and +5°C relative to the detrended air temperatures in the period 2002-2009. temperatures in the period 2002-2009.

Table 4.71 > Mean monthly heat budget at Lago di Viverone if the mean air temperature varies Table 4.72 > Mean monthly position of thermocline at Lago di Viverone if the mean air between -5 and +5°C relative to the detrended air temperatures in the period temperature varies between -5 and +5°C relative to the detrended air 2002-2009. Negative values are heat losses to the atmosphere, positive values temperatures in the period 2002-2009. The values here are the depths below describe a heat input into the water. water surface of the mean monthly maximum temperature gradient.

101 Results of air humidity scenario calculations at Lago di Viverone

Table 4.73 > Mixing index of Lago di Viverone Table 4.74 > Mean annual amount of evaporation if the mean air humidity varies of Lago di Viverone if the mean air between -10 and +10%rF relative humidity varies between -10 and to the air humidity in the period +10%rF relative to the air humidity in 2002-2009. the period 2002-2009.

Table 4.75 > Mean monthly amount of evaporation at Lago di Viverone if the mean air humidity varies between -10 and +10°C relative to the air humidity in the period 2002-2009.

Table 4.76 > Mean monthly surface water temperature at Lago di Viverone if the mean air Table 4.77 > Mean monthly bottom water temperature at Lago di Viverone if the mean air humidity varies between -10 and +10°C relative to the detrended air humidity varies between -10 and +10%rF relative to the air humidity in the temperatures in the period 2002-2009. period 2002-2009.

Table 4.78 > Mean monthly heat budget at Lago di Viverone if the mean air humidity varies Table 4.79 > Mean monthly position of thermocline at Lago di Viverone if the mean air between -10 and +10°C relative to the air humidity in the period 2002-2009. humidity varies between -10 and +10°C relative to the air humidity in the Negative values are heat losses to the atmosphere, positive values describe period 2002-2009. The values here are the depths below water surface a heat input into the water. of the mean monthly maximum temperature gradient.

102 Results of wind velocity scenario calculations at Lago di Viverone

Table 4.80 > Mixing index of Lago di Viverone Table 4.81 > Mean annual amount of if the mean wind velocity varies evaporation of Lake Constance between -0.75 and +3m/s relative if the mean wind velocity varies to the wind velocity in the period between -0.75 and +3m/s relative 2002-2009. to the wind velocity in the period 2002-2009. Table 4.82 > Mean monthly amount of evaporation at Lago di Viverone if the mean wind velocity varies between -0.75 and +3m/s relative to the wind velocity in the period 2002-2009.

Table 4.83 > Mean monthly surface water temperature at Lago di Viverone if the mean wind Table 4.84 > Mean monthly bottom water temperature at Lago di Viverone if the mean wind velocity varies between -0.75 and +3m/s relative to the wind velocity in the velocity varies between -0.75 and +3m/s relative to the wind velocity in the period 2002-2009. period 2002-2009.

Table 4.85 > Mean monthly heat budget at Lago di Viverone if the mean wind velocity varies Table 4.86 > Mean monthly position of thermocline at Lago di Viverone if the mean wind between -0.75 and +3m/s relative to the wind velocity in the period 2002-2009. velocity varies between -0.75 and +3m/s relative to the wind velocity in the period 2002-2009. The values here are the depths below water surface of the mean monthly maximum temperature gradient.

103 Results of global radiation (RG) scenario calculations at Lago di Viverone

Table 4.87 > Mixing index of Lago di Viverone Table 4.88 > Mean annual amount of evaporation if the mean RG varies between of Lago di Viverone if the mean RG -50 and +50Wm-2 relative to the varies between -50 and +50Wm-2 mean cloud cover in the period relative to the mean RG in the period 2002-2009. 2002-2009.

Table 4.89 > Mean monthly amount of evaporation at Lago di Viverone if the mean RG varies between -50 and +50Wm-2 relative to the mean RG in the period 2002-2009.

Table 4.90 > Mean monthly surface water temperature at Lago di Viverone if the mean Table 4.91 > Mean monthly bottom water temperature at Lago di Viverone if the mean RG varies between -50 and +50Wm-2 relative to the mean RG in the period RG varies between -50 and +50Wm-2 relative to the mean RG in the period 2002-2009. 2002-2009.

Table 4.92 > Mean monthly heat budget at Lago di Viverone if the mean RG varies between Table 4.93 > Mean monthly position of thermocline at Lago di Viverone if the mean RG -50 and +50Wm-2 relative to the mean RG in the period 2002-2009. Negative varies between -50 and +50Wm-2 relative to the mean cloud cover in the values are heat losses to the atmosphere, positive values describe a heat input period 2002-2009.The values here are the depths below water surface of the into the water. mean monthly maximum temperature gradient.

104 Results of air temperature scenario calculations at Lake Wörthersee

Table 4.94 > Mixing index of Lake Wörthersee Table 4.95 > Mean annual amount of if the mean air temperature varies evaporation of Lake Wörthersee between -5 and +5°C relative to if the mean air temperature varies the air temperatures in the period between -5 and +5°C relative to 1996-2009. the air temperatures in the period 1996-2009. Table 4.96 > Mean monthly amount of evaporation at Lake Wörthersee if the mean air temperature varies between -5 and +5°C relative to the air temperatures in the period 1996-2009.

Table 4.97 > Mean monthly surface water temperature at Lake Wörthersee if the mean air Table 4.98 > Mean monthly bottom water temperature at Lake Wörthersee if the mean air temperature varies between -5 and +5°C relative to the air temperatures in the temperature varies between -5 and +5°C relative to the air temperatures in the period 1996-2009. period 1996-2009.

Table 4.98 > Mean monthly heat budget at Lake Wörthersee if the mean air temperature varies Table 4.100 > Mean monthly position of thermocline at Lake Wörthersee if the mean air between -5 and +5°C relative to the air temperatures in the period 1996-2009. temperature varies between -5 and +5°C relative to the air temperatures in Negative values are heat losses to the atmosphere, positive values describe a the period 1996-2009. The values here are the depths below water surface of heat input into the water. the mean monthly maximum temperature gradient.

105 Results of air humidity scenario calculations at Lake Wörthersee

Table 4.101 > Mixing index of Lake Wörthersee Table 4.102 > Mean annual amount of if the mean air humidity varies evaporation of Lake Wörthersee between -10 and +10%rF relative if the mean air humidity varies to the air humidity in the period between -10 and +10%rF relative 1996-2009. to the air humidity in the period 1996-2009. Table 4.103 > Mean monthly amount of evaporation at Lake Wörthersee if the mean air humidity varies between -10 and +10°C relative to the air humidity in the period 1996-2009.

Table 4.104 > Mean monthly surface water temperature at Lake Wörthersee if the mean Table 4.105 > Mean monthly bottom water temperature at Lake Wörthersee if the mean air air humidity varies between -10 and +10°C relative to the air humidity in the humidity varies between -10 and +10%rF relative to the air humidity in the period 1996-2009. period 2002-2009.

Table 4.106 > Mean monthly heat budget at Lake Wörthersee if the mean air humidity varies Table 4.107 > Mean monthly position of thermocline at Lake Wörthersee if the mean air between -10 and +10°C relative to the air humidity in the period 1996-2009. humidity varies between -10 and +10°C relative to the air humidity in the Negative values are heat losses to the atmosphere, positive values describe a period 1996-2009. The values here are the depths below water surface of the heat input into the water. mean monthly maximum temperature gradient.

106 Results of wind velocity scenario calculations at Lake Wörthersee

Table 4.108 > Mixing index of Lake Wörthersee Table 4.109 > Mean annual amount of if the mean wind velocity varies evaporation of Lake Wörthersee between -1 and +3m/s relative if the mean wind velocity varies to the wind velocity in the period between -1 and +3m/s relative 1996-2009. to the wind velocity in the period 1996-2009. Table 4.110 > Mean monthly amount of evaporation at Lake Wörthersee if the mean wind velocity varies between -1 and +3m/s relative to the wind velocity in the period 1996-2009.

Table 4.111 > Mean monthly surface water temperature at Lake Wörthersee if the mean Table 4.112 > Mean monthly bottom water temperature at Lake Wörthersee if the mean wind velocity varies between -1 and +3m/s relative to the wind velocity in the wind velocity varies between -1 and +3m/s relative to the wind velocity in the period 1996-2009. period 1996-2009.

Table 4.113 > Mean monthly heat budget at Lake Wörthersee if the mean wind velocity varies Table 4.114 > Mean monthly position of thermocline Lake Wörthersee if the mean wind between -1 and +3m/s relative to the wind velocity in the period 1996-2009. velocity varies between -1 and +3m/s relative to the wind velocity in the period 1996-2009. The values here are the depths below water surface of the mean monthly maximum temperature gradient.

107 Results of cloud cover scenario calculations at Lake Wörthersee

Table 4.115 > Mixing index of Lake Wörthersee Table 4.116 > Mean annual amount of if the mean cloud cover (cc) varies evaporation of Lake Wörthersee between -8 and +16%cc relative if the mean cloud cover (cc) varies to the mean cloud cover in the between -8 and +16%cc relative to period 1996-2009. the mean cloud cover in the period 1996-2009. Table 4.117 > Mean monthly amount of evaporation at Lake Wörthersee if the mean cloud cover (cc) varies between -8 and +16%cc relative to the mean cloud cover in the period 1996-2009.

Table 4.118 > Mean monthly surface water temperature at Lake Wörthersee if the mean Table 4.119 > Mean monthly bottom water temperature at Lake Wörthersee if the mean cloud cover (cc) varies between -8 and +16%cc relative to the mean cloud cloud cover (cc) varies between -8 and +16%cc relative to the mean cloud cover in the period 1996-2009. cover in the period 1996-2009.

Table 4.120 > Mean monthly heat budget at Lake Wörthersee if the mean cloud cover (cc) Table 4.121 > Mean monthly position of thermocline at Lake Wörthersee if the mean cloud varies between -8 and +16%ccrelative to the mean cloud cover in the period cover (cc) varies between -8 and +16%cc relative to the mean cloud cover in 1996-2009. Negative values are heat losses to the atmosphere, positive values the period 1996-2009. The values here are the depths below water surface of describe a heat input into the water. the mean monthly maximum temperature gradient.

108 References

Bäck, T., 1993:. An. Overview. of. Evolutionary. Algorithms. for. P. Minnis, E. F. Harrison, L. L. Stowe, G. G. Gibson, F. M. Parameter. Optimization.. Evolutionary. Computation,. Spring. Denn, D. R. Doelling, W. L. Smith Jr., 1993:.Radiative.Climate. 1993,.Vol..1,.No..1,.Pages.1-23.. Forcing.by.the.Mount.Pinatubo.Eruption..Science.5.March.1993:. Vol..259.no..5100.pp..1411-1415. Bates, B.C., Z.W. Kundzewicz, S. Wu, J.P. Palutikof, Eds., 2008:. Climate. Change. and. Water.. Technical. Paper. of. the. Regione Piemonte, 2004:. Allegato. Tecnico,. PTA. –. Rev.. 01. Intergovernmental.Panel.on.Climate.Change,.IPCC.Secretariat,. Luglio.2004. Geneva,.210.pp. Schulz, L., and others, 2005: Der.Wörthersee,.Limnologische. Deltares, 2009:.Delft3D-FLOW,.Simulation.of.multi-dimensional. Langzeitentwicklung. des. Wörthersees. und. limnologische. hydrodynamic. flows. and. transport. phenomena,. including. Untersuchungen. des. Jahres. 1999. unter. besonderer. sediments.–.User.Manual..Version.3.13,.Published.and.printed. Berücksichtigung. der. Planktonbiocönosen.. Veröffentlichung. by.Deltares,.Netherlands. des.Kärntner.Institutes.für.Seenforschung. Findenegg, I., 1933:.Zur.Naturgeschichte.des.Wörthersees..–. Smiatek, G., H. Kunstmann, R. Knoche, A. Marx, 2009:. Carinthia.II,.Sonderheft. Precipitation. and. temperature. statistics. in. high-resolution. regional. climate. models:. Evaluation. for. the. European. Alps.. Fuentes, U., D. Heimann, 1996:. Verification. of. statistical- Journal. of. Geophysical. Research,. Vol.. 114,. D19107,. 16. PP.,. dynamical.downscaling.in.the.Alpine.region..Climate.Research,. 2009 Vol..7:.151-168,.1996. Smith, S. D., E. G. Banke, 1975:.Variation.of.the.sea.surface. Nakicenovic, N., J. Alcamo, G. Davis, B. d. Vries, J. Fenhann,S. drag.coefficient.with.wind.speed..Quart..J. .R..Met..Soc..(1975),. Gaffin, K. Gregory, A. Griibler, T.Y.Jung, T. Kram,E. Lebre La 101,.pp..665-673. Rovere, L. Michaelis, S. Mori, T. Morita,W. Pepper, H. Pitcher, L. Price, K. Riahi, A. Roehrl,H.-H. Rogner, A. Sankovski, M. Williamson, C.E., J.E. Saros, W.F. Vincent, J.P. Smol, 2009: Schlesinger, P. Shukla, S. Smith, R. Swart, S. van Rooijen, N. Lakes.and.reservoirs.as.sentinels,.integrators,.and.regulators.of. Victor, Z. Dadi, 2000:.Special.Report.on.Emissions.Scenarios,. climate.change..Limnol..Oceanogr.,.54(6,.part.2),.2009,.2273- A.Special.Report.of.Working.Group.III.of.the.Intergovernmental. 2282. Panel.on.Climate.Change..Cambrige.University.Press,.ISBN.0. . 521.80081.1.

List of figures

Figura 4.1...... 79 Figure 4.7...... 82 One.dimensional,.vertical.model.approach. Map.of.Lake.Constance.

Figure 4.2...... 80 Figure 4.8...... 82 Consideration.of.surrounding.mountains.by.the.new.parameter. Bathymetrie.of.Lake.Constance. gipalt. Figure 4.9...... 83 Figure 4.3...... 80 Boxplot. of. the. Secchi. depth. at. Lake. Constance. in. the. period. Modification.of.the.drag.coefficient. 1974-2010..Source.of.Secchi.depth.time.series:.Data.base.of. the.Institute.for.Lake.Research.(LUBW). Figure 4.4...... 81 The.evolutionary.algorithm.(EA).that.was.developed.in.HCILS.to. Figure 4.10...... 83 optimize.the.parameter.for.the.hydrodynamic.model. The. mean. measured. monthly. water. temperatures. at. Lake. Constance.in.the.period.1964-2006..The.observation.position.is. Figure 4.5...... 81 Fischbach-Uttwil.(the.lake’s.deepest.point). Simple. test. of. the. evolutionary. algorithm. (EA). with. two. parameters..The.individuals.aggregate.successfully.around.the. Figure 4.11...... 83 minimum.in.a.»sine.landscape«.(a).after.10.generations.(b). Deviation. statistics. of. measured. minus. calculated. water. temperatures. in. each. month. within. the. calibration. at. Lake. Figure 4.6...... 82 Constance. The.three.lakes.in.HCILS.of.the.SILMAS.project:.Lake.Constance,. Lago.di.Viverone.and.Wörthersee.

109 Figure 4.12...... 84 Figure 4.27...... 89 Map of Lago di Viverone; Source: Regione Piemonte (2004); Measured time series of air temperature at Lago di Viverone in Figure 7-1. the period 2002-2009. The range of the scenarios is indicated by the moving average-filtered time series of maximum and Figure 4.13...... 84 minimum shifts. The Bathymetrie of Lago di Viverone. Figure 4.28...... 89 Figure 4.14...... 84 Measured time series of relative air humidity at Lago di Viverone Mean monthly measured water temperatures at Lago di Viverone in the period 2002-2009. The range of the scenarios is indicated in the period 2001-2009. by the moving average filtered time series of maximum and minimum shifts. Figure 4.15...... 85 Deviation statistics of measured minus calculated water Figure 4.29...... 89 temperatures in each month within the calibration period at Measured time series of wind velocity at Lago di Viverone in the Lago di Viverone. period 2002-2009. The range of the scenarios is indicated by the moving average filtered time series of maximum and minimum Figure 4.16...... 85 shifts. Map of Lake Wörthersee; Source: Google Maps. Figure 4.30...... 89 Figure 4.17...... 85 Measured time series of global radiation at Lago di Viverone in The western basin of Lake Wörthersee. the period 2002-2009 (hourly values). The range of the scenarios is indicated by the moving average filtered time series of Figure 4.18...... 85 maximum and minimum shifts. The Bathymetry of Lake Wörthersee (entire lake and the western basin). Figure 4.31...... 90 Measured time series of air temperature at Lake Wörthersee in Figure 4.19...... 86 the period 1996-2009. The range of the scenarios is indicated Mean monthly secchi depth in Lake Wörthersee (Saag) in the by the moving average filtered time series of maximum and period 1996-2010. minimum shifts.

Figure 4.20...... 86 Figure 4.32...... 90 Measured mean monthly water temperatures at Lake Wörthersee Measured time series of relative air humidity at Lake Wörthersee in the period 1930-2009. in the period 2002-2009. The range of the scenarios is indicated by the moving average filtered time series of maximum and Figure 4.21...... 86 minimum shifts. Deviation statistics of measured minus calculated water temperatures in each month within the calibration at Lake Figure 4.33...... 90 Wörthersee. Measured time series of wind velocity at Lake Wörthersee in the period 1996-2009. The range of the scenarios is indicated by the Figure 4.22...... 86 moving average filtered time series of maximum and minimum Deviation statistics of measured minus calculated water shifts. temperatures in each month within the calibration at the western basin of Lake Wörthersee. Figure 4.34...... 90 Measured time series of cloud cover at Lake Wörthersee in the Figure 4.23...... 87 period 1996-2009. The range of the scenarios is indicated by the Segmented trend analysis of the low-pass filtered hourly air moving average filtered time series of maximum and minimum temperature at Lake Constance in the period 01.01.1964 – shifts. 31.12.2007. Figure 4.35...... 91 Figure 4.24...... 88 Tracer calculations at Lake Constance in the period 1970-2002. Time series of the relative air humidity and its low-pass filter in Good mixing is defined as tracer concentrations > 0.01 [-] below the period 1964-2007. The filter is a moving average with an 150m water depth (dashed white line). normal distributed weighting in a 365d window (sd = 4 years). Figure 4.36...... 91 Figure 4.25...... 88 An example for a year with good mixing followed by a year with Trends in the annual maxima and minima of the low-pass filtered less intensive mixing at Lake Constance. The white dashed lines time series of relative air humidity in the period 1964-2007. indicate the relevant dates and depths for mixing index, and tracer calculations. Figure 4.26...... 88 Low-pass filter of the wind speed measurements in the period Figure 4.37...... 91 1968-2002 (filter window = 10 years; Filter type: moving average Good spring mixing defined by Wahl (2007) at Lake Constance with normal distributed weighting). The original time series derived from orthophosphate measurements, and reproduced covers the period 1964-2007. by the 1dv model (tracer calculations). 1= good mixing, 0 = In the model calculations good mixing is defined as tracer 110 concentrations > 0.01 [-] below 150m depth at 1st May. Figure 4.38...... 92 Figure 4.50...... 92 Mixing. index. of. Lake. Constance. if. the. mean. air. temperature. Mean. monthly. heat. budget. at. Lake. Constance. if. the. mean. varies. between. -4. and. +5°C. relative. to. the. detrended. air. air. humidity. varies. between. -10. and. +10°C. relative. to. the. air. temperatures.in.the.period.1968-2002. humidity. in. the. period. 1968-2002.. Negative. values. are. heat. losses.to.the.atmosphere,.positive.values.describe.a.heat.input. Figure 4.39...... 92 into.the.water. Mean.annual.amount.of.evaporation.of.Lake.Constance.if.the. mean.air.temperature.varies.between.-4.and.+5°C.relative.to.the. Figure 4.51...... 92 detrended.air.temperatures.in.the.period.1968-2002. Mean.monthly.position.of.thermocline.at.Lake.Constance.if.the. mean. air. humidity. varies. between. -4. and. +5°C. relative. to. the. Figure 4.40...... 92 air.humidity.in.the.period.1968-2002..The.values.here.are.the. Mean.monthly.amount.of.evaporation.at.Lake.Constance.if.the. depths. below. water. surface. of. the. mean. monthly. maximum. mean.air.temperature.varies.between.-4.and.+5°C.relative.to.the. temperature.gradient. detrended.air.temperatures.in.the.period.1968-2002. Figure 4.52...... 92 Figure 4.41...... 92 Mixing.index.of.Lake.Constance.if.the.mean.wind.velocity.varies. Mean.monthly.surface.water.temperature.at.Lake.Constance.if. between. -1.85. and. +3m/s. relative. to. the. wind. velocity. in. the. the.mean.air.temperature.varies.between.-4.and.+5°C.relative.to. period.1964-2007. the.detrended.air.temperatures.in.the.period.1968-2002. Figure 4.53...... 92 Figure 4.42...... 92 Mean.annual.amount.of.evaporation.of.Lake.Constance.if.the. Mean.monthly.bottom.water.temperature.at.Lake.Constance.if. mean.wind.velocity.varies.between.-1.85.and.+3m/s.relative.to. the.mean.air.temperature.varies.between.-4.and.+5°C.relative.to. the.wind.velocity.in.the.period.1964-2007. the.detrended.air.temperatures.in.the.period.1968-2002. Figure 4.54...... 92 Figure 4.43...... 92 Mean.monthly.amount.of.evaporation.at.Lake.Constance.if.the. Mean.monthly.heat.budget.at.Lake.Constance.if.the.mean.air. mean.wind.velocity.varies.between.-1.85.and.+3m/s.relative.to. temperature.varies.between.-4.and.+5°C.relative.to.the.detrended. the.wind.velocity.in.the.period.1964-2007. air.temperatures.in.the.period.1968-2002..Negative.values.are. heat.losses.to.the.atmosphere,.positive.values.describe.a.heat. Figure 4.55...... 92 input.into.the.water. Mean.monthly.surface.water.temperature.at.Lake.Constance.if. the.mean.wind.velocity.varies.between.-1.85.and.+3m/s.relative. Figure 4.44...... 92 to.the.wind.velocity.in.the.period.1964-2007. Mean.monthly.position.of.thermocline.at.Lake.Constance.if.the. mean.air.temperature.varies.between.-4.and.+5°C.relative.to.the. Figure 4.56...... ?? detrended.air.temperatures.in.the.period.1968-2002..The.values. Mean.monthly.bottom.water.temperature.at.Lake.Constance.if. here.are.the.depths.below.water.surface.of.the.mean.monthly. the.mean.wind.velocity.varies.between.-1.85.and.+3m/s.relative. maximum.temperature.gradient. to.the.wind.velocity.in.the.period.1964-2007.

Figure 4.45...... 92 Figure 4.57...... 92 Mixing.index.of.Lake.Constance.if.the.mean.air.humidity.varies. Mean.monthly.heat.budget.at.Lake.Constance.if.the.mean.wind. between. -10. and. +10%rF. relative. to. the. air. humidity. in. the. velocity. varies. between. -1.85. and. +3m/s. relative. to. the. wind. period.1968-2002. velocity.in.the.period.1964-2007.

Figure 4.46...... 92 Figure 4.58...... 93 Mean.annual.amount.of.evaporation.of.Lake.Constance.if.the. Mean.monthly.position.of.thermocline.at.Lake.Constance.if.the. mean.air.humidity.varies.between.-10.and.+10%rF.relative.to. mean.wind.velocity.varies.between.-1.85.and.+3m/s.relative.to. the.air.humidity.in.the.period.1968-2002. the.wind.velocity.in.the.period.1964-2007..The.values.here.are. the.depths.below.water.surface.of.the.mean.monthly.maximum. Figure 4.47...... 92 temperature.gradient. Mean.monthly.amount.of.evaporation.at.Lake.Constance.if.the. mean.air.humidity.varies.between.-10.and.+10°C.relative.to.the. Figure 4.59...... 93 air.humidity.in.the.period.1968-2002. Mixing. index. of. Lake. Constance. if. the. mean. cloud. cover. (cc). varies. between. -16. and. +16%cc. relative. to. the. mean. cloud. Figure 4.48...... 92 cover.in.the.period.1965-2007. Mean.monthly.surface.water.temperature.at.Lake.Constance.if. the.mean.air.humidity.varies.between.-10.and.+10°C.relative.to. Figure 4.60...... 93 the.detrended.air.temperatures.in.the.period.1968-2002. Mean.annual.amount.of.evaporation.of.Lake.Constance.if.the. mean.cloud.cover.(cc).varies.between.-16.and.+16%cc.relative. Figure 4.49...... 92 to.the.mean.cloud.cover.in.the.period.1965-2007. Mean.monthly.bottom.water.temperature.at.Lake.Constance.if. the.mean.air.humidity.varies.between.-10.and.+10%rF.relative. to.the.air.humidity.in.the.period.1968-2002.

111 Figure 4.61...... xx Figure 4.72...... 93 Mean monthly amount of evaporation at Lake Constance if the MMean monthly position of thermocline at Lago di Viverone if mean cloud cover (cc) varies between -16 and +16%cc relative the mean air temperature varies between -5 and +5°C relative to the mean cloud cover in the period 1965-2007. to the detrended air temperatures in the period 2002-2009. The values here are the depths below water surface of the mean Figure 4.62...... 93 monthly maximum temperature gradient. Mean monthly surface water temperature at Lake Constance if the mean cloud cover (cc) varies between -16 and +16%cc Figure 4.73...... 93 relative to the mean cloud cover in the period 1965-2007. Mixing index of Lago di Viverone if the mean air humidity varies between -10 and +10%rF relative to the air humidity in the Figure 4.63...... 93 period 2002-2009. Mean monthly bottom water temperature at Lake Constance if the mean cloud cover (cc) varies between -16 and +16%cc Figure 4.74...... 93 relative to the mean cloud cover in the period 1965-2007. Mean annual amount of evaporation of Lago di Viverone if the mean air humidity varies between -10 and +10%rF relative to Figure 4.64...... 93 the air humidity in the period 2002-2009. Mean monthly heat budget at Lake Constance if the mean cloud cover (cc) varies between -16 and +16%cc relative to the mean Figure 4.75...... 93 cloud cover in the period 1965-2007. Negative values are heat Mean monthly amount of evaporation at Lago di Viverone if the losses to the atmosphere, positive values describe a heat input mean air humidity varies between -10 and +10°C relative to the into the water. air humidity in the period 2002-2009.

Figure 4.65...... 93 Figure 4.76...... 93 Mean monthly position of thermocline at Lake Constance if the Mean monthly surface water temperature at Lago di Viverone if mean cloud cover (cc) varies between -16 and +16%cc relative the mean air humidity varies between -10 and +10°C relative to to the mean cloud cover in the period 1965-2007.The values the detrended air temperatures in the period 2002-2009. here are the depths below water surface of the mean monthly maximum temperature gradient. Figure 4.77...... 93 Mean monthly bottom water temperature at Lago di Viverone if Figure 4.66...... 93 the mean air humidity varies between -10 and +10%rF relative Mixing index of Lago di Viverone if the mean air temperature to the air humidity in the period 2002-2009. varies between -5 and +5°C relative to the detrended air temperatures in the period 2002-2009. Figure 4.78...... 94 Mean monthly heat budget at Lago di Viverone if the mean Figure 4.67...... 93 air humidity varies between -10 and +10°C relative to the air Mean annual amount of evaporation of Lago di Viverone if the humidity in the period 2002-2009. Negative values are heat mean air temperature varies between -5 and +5°C relative to the losses to the atmosphere, positive values describe a heat input detrended air temperatures in the period 2002-2009. into the water.

Figure 4.68...... 93 Figure 4.79...... 94 Mean monthly amount of evaporation at Lago di Viverone if the Mean monthly position of thermocline at Lago di Viverone if the mean air temperature varies between -5 and +5°C relative to the mean air humidity varies between -10 and +10°C relative to the detrended air temperatures in the period 2002-2009. air humidity in the period 2002-2009. The values here are the depths below water surface of the mean monthly maximum Figure 4.69...... xx temperature gradient. Mean monthly surface water temperature at Lago di Viverone if the mean air temperature varies between -5 and +5°C relative to Figure 4.80...... 94 the detrended air temperatures in the period 2002-2009. Mixing index of Lago di Viverone if the mean wind velocity varies between -0.75 and +3m/s relative to the wind velocity in the Figure 4.70...... 93 period 2002-2009. Mean monthly bottom water temperature at Lago di Viverone if the mean air temperature varies between -5 and +5°C relative to Figure 4.81...... 94 the detrended air temperatures in the period 2002-2009. Mean annual amount of evaporation of Lake Constance if the mean wind velocity varies between -0.75 and +3m/s relative to Figure 4.71...... 93 the wind velocity in the period 2002-2009. Mean monthly heat budget at Lago di Viverone if the mean air temperature varies between -5 and +5°C relative to the detrended Figure 4.82...... 94 air temperatures in the period 2002-2009. Negative values are Mean monthly amount of evaporation at Lago di Viverone if the heat losses to the atmosphere, positive values describe a heat mean wind velocity varies between -0.75 and +3m/s relative to input into the water. the wind velocity in the period 2002-2009.

112 Figure 4.83...... 94 Figure 4.95...... 95 MMean.monthly.surface.water.temperature.at.Lago.di.Viverone.if. Mean.annual.amount.of.evaporation.of.Lake.Wörthersee.if.the. the.mean.wind.velocity.varies.between.-0.75.and.+3m/s.relative. mean.air.temperature.varies.between.-5.and.+5°C.relative.to.the. to.the.wind.velocity.in.the.period.2002-2009. air.temperatures.in.the.period.1996-2009.

Figure 4.84...... 94 Figure 4.96...... 95 Mean.monthly.bottom.water.temperature.at.Lago.di.Viverone.if. Mean.monthly.amount.of.evaporation.at.Lake.Wörthersee.if.the. the.mean.wind.velocity.varies.between.-0.75.and.+3m/s.relative. mean.air.temperature.varies.between.-5.and.+5°C.relative.to.the. to.the.wind.velocity.in.the.period.2002-2009. air.temperatures.in.the.period.1996-2009.

Figure 4.85...... 94 Figure 4.97...... 95 Mean.monthly.heat.budget.at.Lago.di.Viverone.if.the.mean.wind. Mean.monthly.surface.water.temperature.at.Lake.Wörthersee.if. velocity. varies. between. -0.75. and. +3m/s. relative. to. the. wind. the.mean.air.temperature.varies.between.-5.and.+5°C.relative.to. velocity.in.the.period.2002-2009. the.air.temperatures.in.the.period.1996-2009.

Figure 4.86...... 94 Figure 4.98...... xx Mean.monthly.position.of.thermocline.at.Lago.di.Viverone.if.the. Mean.monthly.bottom.water.temperature.at.Lake.Wörthersee.if. mean.wind.velocity.varies.between.-0.75.and.+3m/s.relative.to. the.mean.air.temperature.varies.between.-5.and.+5°C.relative.to. the.wind.velocity.in.the.period.2002-2009..The.values.here.are. the.air.temperatures.in.the.period.1996-2009. the.depths.below.water.surface.of.the.mean.monthly.maximum. temperature.gradient. Figure 4.99...... 95 Mean. monthly. heat. budget. at. Lake. Wörthersee. if. the. mean. Figure 4.87...... 94 air.temperature.varies.between.-5.and.+5°C.relative.to.the.air. Mixing.index.of.Lago.di.Viverone.if.the.mean.RG.varies.between. temperatures.in.the.period.1996-2009..Negative.values.are.heat. -50.and.+50Wm-2.relative.to.the.mean.cloud.cover.in.the.period. losses.to.the.atmosphere,.positive.values.describe.a.heat.input. 2002-2009. into.the.water.

Figure 4.88...... 94 Figure 4.100...... 95 Mean.annual.amount.of.evaporation.of.Lago.di.Viverone.if.the. Mean.monthly.position.of.thermocline.at.Lake.Wörthersee.if.the. mean.RG.varies.between.-50.and.+50Wm-2.relative.to.the.mean. mean.air.temperature.varies.between.-5.and.+5°C.relative.to.the. RG.in.the.period.2002-2009. air.temperatures.in.the.period.1996-2009..The.values.here.are. the.depths.below.water.surface.of.the.mean.monthly.maximum. Figure 4.89...... 94 temperature.gradient. Mean.monthly.amount.of.evaporation.at.Lago.di.Viverone.if.the. mean.RG.varies.between.-50.and.+50Wm-2.relative.to.the.mean. Figure 4.101...... 95 RG.in.the.period.2002-2009. Mixing.index.of.Lake.Wörthersee.if.the.mean.air.humidity.varies. between. -10. and. +10%rF. relative. to. the. air. humidity. in. the. Figure 4.90...... 94 period.1996-2009. Mean.monthly.surface.water.temperature.at.Lago.di.Viverone.if. the.mean.RG.varies.between.-50.and.+50Wm-2.relative.to.the. Figure 4.102...... 95 mean.RG.in.the.period.2002-2009. Mean.annual.amount.of.evaporation.of.Lake.Wörthersee.if.the. mean.air.humidity.varies.between.-10.and.+10%rF.relative.to. Figure 4.91...... 94 the.air.humidity.in.the.period.1996-2009. Mean.monthly.bottom.water.temperature.at.Lago.di.Viverone.if. the.mean.RG.varies.between.-50.and.+50Wm-2.relative.to.the. Figure 4.103...... 95 mean.RG.in.the.period.2002-2009. Mean.monthly.amount.of.evaporation.at.Lake.Wörthersee.if.the. mean.air.humidity.varies.between.-10.and.+10°C.relative.to.the. Figure 4.92...... 94 air.humidity.in.the.period.1996-2009. Mean.monthly.heat.budget.at.Lago.di.Viverone.if.the.mean.RG. varies. between. -50. and. +50Wm-2. relative. to. the. mean. RG. in. Figure 4.104...... 95 the. period. 2002-2009.. Negative. values. are. heat. losses. to. the. Mean.monthly.surface.water.temperature.at.Lake.Wörthersee.if. atmosphere,.positive.values.describe.a.heat.input.into.the.water. the.mean.air.humidity.varies.between.-10.and.+10°C.relative.to. the.air.humidity.in.the.period.1996-2009. Figure 4.93...... 94 Mean. monthly. position. of. thermocline. at. Lago. di. Viverone. if. Figure 4.105...... 95 the.mean.RG.varies.between.-50.and.+50Wm-2.relative.to.the. Mean.monthly.bottom.water.temperature.at.Lake.Wörthersee.if. mean.cloud.cover.in.the.period.2002-2009.The.values.here.are. the.mean.air.humidity.varies.between.-10.and.+10%rF.relative. the.depths.below.water.surface.of.the.mean.monthly.maximum. to.the.air.humidity.in.the.period.2002-2009. temperature.gradient.

Figure 4.94...... 94 Mixing. index. of. Lake. Wörthersee. if. the. mean. air. temperature. varies.between.-5.and.+5°C.relative.to.the.air.temperatures.in. the.period.1996-2009. 113 Figure 4.106...... 95 Figure 4.115...... 96 Mean monthly heat budget at Lake Wörthersee if the mean Mixing index of Lake Wörthersee if the mean cloud cover (cc) air humidity varies between -10 and +10°C relative to the air varies between -8 and +16%cc relative to the mean cloud cover humidity in the period 1996-2009. Negative values are heat in the period 1996-2009. losses to the atmosphere, positive values describe a heat input into the water. Figure 4.116...... 96 Mean annual amount of evaporation of Lake Wörthersee if the Figure 4.107...... 95 mean cloud cover (cc) varies between -8 and +16%cc relative to Mean monthly position of thermocline at Lake Wörthersee if the mean cloud cover in the period 1996-2009. the mean air humidity varies between -10 and +10°C relative to the air humidity in the period 1996-2009. The values here are Figure 4.117...... 96 the depths below water surface of the mean monthly maximum Mean monthly amount of evaporation at Lake Wörthersee if the temperature gradient. mean cloud cover (cc) varies between -8 and +16%cc relative to the mean cloud cover in the period 1996-2009. Figure 4.108...... 95 Mixing index of Lake Wörthersee if the mean wind velocity varies Figure 4.118...... 96 between -1 and +3m/s relative to the wind velocity in the period Mean monthly surface water temperature at Lake Wörthersee 1996-2009. if the mean cloud cover (cc) varies between -8 and +16%cc relative to the mean cloud cover in the period 1996-2009. Figure 4.109...... 95 Mean annual amount of evaporation of Lake Wörthersee if the Figure 4.119...... 96 mean wind velocity varies between -1 and +3m/s relative to the Mean monthly bottom water temperature at Lake Wörthersee wind velocity in the period 1996-2009. if the mean cloud cover (cc) varies between -8 and +16%cc relative to the mean cloud cover in the period 1996-2009. Figure 4.110...... 95 Mean monthly amount of evaporation at Lake Wörthersee if the Figure 4.120...... 96 mean wind velocity varies between -1 and +3m/s relative to the Mean monthly heat budget at Lake Wörthersee if the mean wind velocity in the period 1996-2009. cloud cover (cc) varies between -8 and +16%ccrelative to the mean cloud cover in the period 1996-2009. Negative values are Figure 4.111...... 95 heat losses to the atmosphere, positive values describe a heat Mean monthly surface water temperature at Lake Wörthersee if input into the water. the mean wind velocity varies between -1 and +3m/s relative to the wind velocity in the period 1996-2009. Figure 4.121...... 96 Mean monthly position of thermocline at Lake Wörthersee if the Figure 4.112...... 95 mean cloud cover (cc) varies between -8 and +16%cc relative Mean monthly bottom water temperature at Lake Wörthersee if to the mean cloud cover in the period 1996-2009. The values the mean wind velocity varies between -1 and +3m/s relative to here are the depths below water surface of the mean monthly the wind velocity in the period 1996-2009. maximum temperature gradient.

Figure 4.113...... 95 Mean monthly heat budget at Lake Wörthersee if the mean wind velocity varies between -1 and +3m/s relative to the wind velocity in the period 1996-2009.

Figure 4.114...... 95 Mean monthly position of thermocline Lake Wörthersee if the mean wind velocity varies between -1 and +3m/s relative to the wind velocity in the period 1996-2009. The values here are the depths below water surface of the mean monthly maximum temperature gradient.

114 List of tables

Table 4.1...... 80 Table 4.13...... 88 Calibrated.parameters.of.the.hydrodynamic.model. Characteristic. values. of. cloud. cover. scenarios. at. Lake. Constance. Table 4.2...... 80 Parameters. of. the. hydrodynamic. model. that. are. given. by. Table 4.14...... 89 measurements. Characteristic. values. of. air. temperature. scenarios. at. Lago. di. Viverone. Table 4.3...... 81 Meteorological.variables. Table 4.15...... 89 Characteristic. values. of. air. temperature. scenarios. at. Lago. di. Table 4.4...... 82 Viverone. Characteristics.of.Lake.Constance.(the.South-Eastern.big.part. »Obersee«). Table 4.16...... 89 Characteristic. values. of. wind. velocity. scenarios. at. Lago. di. Table 4.5...... 83 Viverone. Mean.monthly.Secchi.depth,.and.maximal.and.minimal.measured. values.of.Secchi.depth.in.Lake.Constance.in.the.period.1974- Table 4.17...... 89 2010..Data.source:.Data.base.of.the.Institute.for.Lake.Research. Characteristic. values. of. global. radiation. scenarios. at. Lago. di. (LUBW). Viverone.

Table 4.6...... 83 Table 4.18...... 90 Calibrated.Parameters.for.Lake.Constance. Characteristic. values. of. air. temperature. scenarios. at. Lake. Wörthersee. Table 4.7...... 85 Characteristics.of.Lago.di.Viverone.. Table 4.19...... 90 Characteristic. values. of. air. temperature. scenarios. at. Lake. Table 4.8...... 85 Wörthersee. Parameters.for.Lago.di.Viverone Table 4.20...... 90 Table 4.9...... 86 Characteristic. values. of. wind. velocity. scenarios. at. Lake. Characteristics.of.Lake.Wörthersee. Wörthersee.

Table 4.10...... 87 Table 4.21...... 90 Characteristic. values. of. temperature. scenarios. at. Lake. Characteristic. values. of. cloud. cover. scenarios. at. Lake. Constance. Wörthersee..

Table 4.11...... 88 Characteristic. values. of. air. humidity. scenarios. at. Lake. Constance.

Table 4.12...... 88 Characteristic. values. of. wind. speed. scenarios. at. Lake. Constance.

115