Overview of current drought research and its relevance to the challenges we face
Kevin Hiscock School of Environmental Sciences University of East Anglia, Norwich E-mail: [email protected] Outline of presentation
Coupled climate change and groundwater modelling (local scale)
Climate change impacts on groundwater quality
Integrated modelling for multi-objective decision-making
Global-scale modelling of climate change impacts on groundwater
Further research recommendations Impacts of increasing GHG concentrations on the natural hydrological cycle emphasising changes in hydrogeological conditions Multi-model mean changes in: (a) precipitation (mm/day), (b) soil moisture content (%), (c) runoff (mm/day) and (d) evaporation (mm/day)
(proxy indicator for Δrecharge)
To indicate consistency in the sign of change, regions are stippled where at least 80% of models agree on the sign of the mean change. Changes are annual means for the medium, A1B scenario ‘greenhouse gas’ emissions scenario for the period 2080-2099 relative to 1980-1999. Soil moisture and runoff changes are shown at land points with valid data from at least 10 models. (Collins et al. 2007 IPCC AR4 WG1) Climate change impacts on groundwater-fed wetlands TL88/008 Ring Mere
N 38 NORTH NORFOLK COAST 36
BROADLAND THE WASH PRIORY MEADOWS 34 ALDERFEN BROAD DERSINGHAM BOG LUDHAM-POTTER HEIGHAM MARSHES ROYDON COMMON 32 SHALLAM DYKE MARSHES, THURNE DERBY FENS BURE BROADS & MARSHES UPPER THURNE BROADS & MARSHES 30 ANT BROADS & MARSHES OUSE WASHES DUCAN'S MARSH, CLAXTON CANTLEY MARSHES BADLEY MOOR 28 YARE BROADS & MARSHES R Wissey PASHFORD POORS FEN HALVERGATE MARSHES CHETTISHAM MEADOW REDGRAVE & LOPHAM FENS 26 CHIPPENHAM FEN MINSMERE & MARSHES R LittleCAM WASHES Ouse LANGMERE GYPSY CAMP FOWLMERE 24 RINGMERE MEADOWS AOD) level (m Water DEVIL’S PUNCH BOWL WICKEN FEN SNAILWELL POOR'S FEN 22 NAILWELL MEADOWS
ALDE-ORE ESTUARY
1979 1980 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1993 1995 1996 1998 1999 2000 1981 1992 1994 1997 0 20 km 1978
Coltishall Sites of Special Scientific Interest Comparison of Chalk groundwater levels at observation
Ramsar Sites borehole TL88/008 and water levels at Ringmere (Environment Agency data)
Water level thresholds of wetland plants typical of East Anglia (after Newbold and Mountford, 1997). Negative numbers indicate water levels below ground level
Dry Wet Species Common name water level water level (cm) (cm) Carex nigra Common sedge -65 0 Schoenus nigricans Black bog-rush -20 0 Carex elata Tufted sedge -30 40 Langmere (August 2007) Phragmites australis Common reed -100 50 Carex rostrata Bottle sedge -15 100 River boundary (River Wissey) N Groundwater recharge
12 km Abstraction well Layer 1 (sand) Potential evapotranspiration Layer 2 (clay) Layer 3 (chalk)
Water table Mere
River boundary (Little Ouse River) Constant head boundary 0 0 8 km
Conceptual groundwater model for a wetland fed by an Initialunconfined hydraulic Chalk parameters Calibrated hydraulic values aquifer in East Anglia Kx Ky Kz Kx Ky Kz Specific Specific (m d-1) (m d-1) (m d-1) (m d-1) (m d-1) (m d-1) storage yield Chalk 8.0 8.0 0.8 7.55 7.55 0.77 0.005 0.05 Sand 3.0 3.0 3.0 3.0 3.0 3.0 0.02 0.25 Clay 1.5 1.5 0.005 1.5 1.5 0.015 0.001 0.01 Herrera-Pantoja et al. Ecohydrology (In Press) Baseline 1961-1991 No data 2020s 2050s 2080s 350
300 )
) 250
1 1
- -
200
150
Hxr (mm year year (mm (mm Hxr Hxr 100
50
0
1960 1960 1965 1965 1970 1970 1975 1975 1980 1980 1985 1985 1990 1990 1995 1995 2000 2000 2005 2005 2010 2010 2015 2015 2020 2020 2025 2025 2030 2030 2035 2035 2040 2040 2045 2045 2050 2050 2055 2055 2060 2060 2065 2065 2070 2070 2075 2075 2080 2080 2085 2085 2090 2090 2095 2095 2100 2100
Annual potential groundwater recharge (Hxr) values for the baseline period (1961–1990) and three time periods for a ‘high’ gas emissions scenario (2020s, 2050s and 2080s) for northern East Anglia. The horizontal line shows the mean annual value for the baseline period
Herrera-Pantoja et al. Ecohydrology (In Press) Water level Baseline mean 30 cm below gl 50 cm below gl 1 Standard deviation 2 Standard deviation 37 a) 35
33
31
29 Water level level Water(mAOD)
27
1962 1963 1964 1966 1967 1969 1970 1972 1973 1974 1976 1977 1979 1980 1982 1983 1984 1986 1987 1989 1990 1961 1965 1968 1971 1975 1978 1981 1985 1988
Water level Baseline mean 30 cm below gl b) 50 cm below gl 1 Standard deviation 2 Standard deviation 37
35
33
31
29 Water level level Water(mAOD)
27
2011 2015 2016 2017 2018 2019 2020 2024 2025 2026 2027 2028 2029 2033 2034 2035 2036 2037 2038 2012 2013 2014 2021 2022 2023 2030 2031 2032 2039 2040
Water level Baseline mean 30 cm below gl 50 cm below gl 1 Standard deviation 2 Standard deviation Water level Baseline mean 30 cm below gl 50 cm below gl 1 Standard deviation 2 Standard deviation c) 37 37 a) 35 35 33 33
31 31
29 29
Water level level Water(mAOD) Water level level Water(mAOD)
27 27
1962 1963 1964 1966 1967 1969 1970 1972 1973 1974 1976 1977 1979 1980 1982 1983 1984 1986 1987 1989 1990 1961 1965 1968 1971 1975 1978 1981 1985 1988
2041 2042 2043 2047 2048 2049 2050 2051 2052 2056 2057 2058 2059 2060 2061 2065 2066 2067 2068 2069 2070 2044 2045 2046 2053 2054 2055 2062 2063 2064
Water level Baseline mean 30 cm below gl Water level Baseline mean 30 cm below gl 50 cm below gl 1 Standard deviation 2 Standard deviation b) d) 50 cm below gl 1 Standard deviation 2 Standard deviation 37 37
35 35 33 33 31
29 31 Water level level Water(mAOD)
27
Water level (mAOD) 29
2011 2015 2016 2017 2018 2019 2020 2024 2025 2026 2027 2028 2029 2033 2034 2035 2036 2037 2038 2013 2014 2021 2022 2023 2030 2031 2032 2039 2040 2012 27
Water level Baseline mean 30 cm below gl
2071 2072 2073 2077 2078 2079 2080 2081 2082 2086 2087 2088 2089 2090 2091 2095 2096 2097 2098 2099 2100 2075 2076 2083 2084 2085 2092 2093 2094 50 cm below gl 1 Standard deviation 2 Standard deviation 2074 c) 37 35
33 Time series of water levels for a wetland fed by an unconfined Chalk aquifer 31 for the baseline period 1961–1990 and for the 2020s, 2050s and 2080s 29 Water level level Water(mAOD) periods of the ‘high’ gas emissions scenario 27
2041 2042 2043 2047 2048 2049 2050 2051 2052 2056 2057 2058 2059 2060 2061 2065 2066 2067 2068 2069 2070 2044 2045 2046 2053 2054 2055 2062 2063 2064 TheWater series level presentBaseline meanthe baseline30 cm below mean, gl one and two standard deviations about d) 50 cm below gl 1 Standard deviation 2 Standard deviation 37 the mean above ground level (gl) and two dry water level thresholds 35
33
31 Herrera-Pantoja et al. Ecohydrology (In Press)
Water level (mAOD) 29
27
2071 2072 2073 2077 2078 2079 2080 2081 2082 2086 2087 2088 2089 2090 2091 2095 2096 2097 2098 2099 2100 2074 2075 2076 2083 2084 2085 2092 2093 2094 Simulated water levels in a wetland fed by an unconfined Chalk aquifer during the baseline period (1961–1990) and the 2020s, 2050s and 2080s future periods of the ‘high’ gas emissions scenario and their likely impacts on groundwater-fed wetland communities
Mean Minimum Duration of Maximum Duration of Likely impact on water table water table low water table water table hight water wetland communities (m AOD ) (m AOD) (months) (m AOD) table (months)
Baseline 30.74 28.68 12 33.85 3 2020s 31.06 27.98 24 33.76 2 Changes towards swamp-like stands 2050s 30.79 29.35 9 36.85 6 Recovery of rare and local plant species 2080s 29.84 27.91 61 32.03 0 Loss of wetlands communities
Herrera-Pantoja et al. Ecohydrology (In Press) Over-abstraction in the High Plains Aquifer
In 1990, 2.2 million people were supplied by groundwater from the High Plains Aquifer with total public supply abstractions of 1.26 million m3 per day In general, water levels in this important aquifer and dropping (>30 m in some areas) Strong Correlations: PDO, precip., gw. Courtesy J. Gurdak, USGS
Pacific Decadal Oscillation (PDO, every 10-25 years) appears to have strongest correlations with groundwater level fluctuations and is a dominant control on climate varying recharge to this aquifer Mobilisation of chemical reservoir by climate Courtesy J.Gurdak, UGSG
The occurrence of these types of events could GCM projections of High Plains region show an further mobilise the large chemical reservoirs increased frequency of intense precipitation events downward to the water table, leading to further decline of groundwater quality
Integrated modelling Natural Social, environment Economic and Environmental Research (SEER) Project
Multi-objective land use decision Water making Water environmentLand use environment Integrated Natural modelling environment
Climate change
Policy Market
Farm decisions & income (£)
Water Land use environment
Economic value of water (£) If land use in part drives water quality what drives land use?
Set aside rate Output prices Soils
NVZ, ESA, Parks, etc. Input costs Temperature
Milk quota Technology Rainfall Data and analysis Agricultural Census data for every 2km grid square of GB from 1969 plus 50,000 farm years of Farm Business Survey data:
• Agricultural land use hectares (wheat, barley, grass, etc.); • Livestock numbers (dairy, beef, sheep, etc) • Time trends (response times, new crops, etc.)
We then add • Environmental & climatic data (rainfall, temperature, etc.) • Policy determinants (CAP reform etc.) • Input and output prices for the period
Resulting models tested by comparing predictions with actual land use Validation: Actual versus predicted tests
Cereals
Temporary grassland Climate change impacts
Rainfall: 2004 - 2040
Temperature: 2004 - 2040 Predicted climate change impacts on land use
Dairy (Δcows/2km sq) < -100 -100 to -20 -20 to 20 20 to 80 80 to 200
Oilseed rape (Δha/2km sq)
< -30 -30 to -12 -12 to 12 12 to 30 30 to 70
2004-2020 2004-2040 2004-2060 Holding all else constant* - what is the impact of climate change on farm incomes by 2050?
UKCIP low emissionsUKCIP scenario high emissions scenario
*this, of course, wont happen as new crops will develop in response to climate change. Nevertheless, results show an interesting spatial pattern Land use Landchange use & change water quality
Nitrate leaching per month
Modelling land use change as a result of: • climate change; • new policy;
• Modelling worldIntegrated market the modelling: impactsshifts; of land use change on river • etc.Linking land use with diffusewater water quality pollution and ecosystems services Also estimating resultant- and how farm water incomes policies like WFD forces land use to change Schematic of water storage compartments (boxes) and flows (arrows) within each 0.5o grid cell of the WaterGAP Global Hydrology Model (WGHM version 2.1h). Water use estimates for each source in each grid cell computed with GWSWUSE
Qb – outflow from groundwater to surface water; controlled by an outflow -1 coefficient, kg, set globally at 0.01 day Döll et al. Journal of Geodynamics (In Press) Global water use during the period 1998-2002, including groundwater fractions
Impact of human water use on seasonal amplitude (SA) of total water storage (TWS).
(a) SA computed as the grid-cell specific value of maximum mean monthly TWS minus minimum mean monthly TWS, averaged over 1998-2002, taking account of water withdrawals, in mm.
(b) Change of SA with water withdrawals relative to SA without withdrawals, in percent of SA without water withdrawals (positive values indicate that water withdrawals increase SAs of TWS)
Döll et al. Journal of Geodynamics (In Press) Scatter plots of recession coefficient, kbf vs. various catchment parameters and WHYMAP (2010) hydrogeological classes
kbf t Qt Qoe
Qo - discharge at the start of baseflow recession
Qt - discharge at later time, t kbf - aquifer or recession coefficient
Q data from the Global Runoff Database from the GRDC
Peña-Arancibia et al. Hydrology and Earth System Sciences (2010) Pan-tropical map of baseflow recession coefficient using the exponential regression equation and mean annual rainfall (MAR)
Peña-Arancibia et al. Hydrology and Earth System Sciences (2010) Climate impacts modelling
ERMITAGE: Enhancing Robustness and Model Integration for the Assessment of Global Environmental Change
The ERMITAGE project aims to link several key component models into a common framework in order to better understand how management of land, water and the Earth’s climate system can best be understood. Key component models represent: the climate system (MAGICC6, GENIE; ClimGEN); climate change and land use change impacts upon water resources, agricultural and ecological systems (LPJmL); the agricultural/agro-economic system (MAgPIE); and the world economy and energy technologies (TIAM, REMIND, GEMINI-E3) Results of pattern-scaled climate scenario data (provided by ClimGEN) used to drive the LPJmL land use model, providing a global-scale impact scenario assessment
RCP2.6 RCP4.5
Simulated impacts shown as ensemble median changes for each of four RCPs and each of 18 GCMs, showing 30-yr average changes (2071–2100) relative to the observed period 1971–2000
Note that impacts tend to strongly increase for the higher RCPs, suggesting significant decreases in runoff (as a proxy for water availability) in many regions
RCP6.0 RCP8.5 Representative concentration pathways (RCPs) Ensemble median changes in simulated average annual (sub)surface runoff
(mm) Climate Change Effects on Groundwater Resources: A Global Synthesis of Findings and Recommendations
Treidel, H, Martin-Bordes, M. Gurdak, J. (eds). IAH/CRC Press, 2012
Recommendations for further research
1. Integrate climate change and variability to improve conceptual hydrological models
2. Institute a comprehensive strategy to monitor global groundwater resources
3. Give greater attention to groundwater quality
4. Study mechanisms of snowmelt runoff and recharge
5. Continue interdisciplinary and multidisciplinary collaboration
6. Make groundwater research usable by (ground)water managers