Energy Fluxes in an Ice Cave of Sporadic Permafrost in the Swiss Jura Mountains – Concept and First Observational Results

Home , Jura Mountains

Energy fluxes in an ice cave of sporadic permafrost in the Swiss Jura mountains – concept and first observational results

M. Luetscher Swiss Institute for Speleology and Karstology (SISKA), La Chaux-de-Fonds, Switzerland Glaciology and Geomorphodynamics Group, Geography Department, University of Zurich, Switzerland P.-Y. Jeannin Swiss Institute for Speleology and Karstology (SISKA), La Chaux-de-Fonds, Switzerland W. Haeberli Glaciology and Geomorphodynamics Group, Geography Department, University of Zurich, Switzerland

ABSTRACT: Low-altitude perennial ice occurrences are observed in ice caves of the Jura mountains. Resulting from a fragile climatological equilibrium, the presence of ice is subject to significant variations. The paper aims at presenting first observations on distribution in time and space of heat transfer processes taking place between air, ice and rock, and their influence on ice formation and conservation of ice. Correlation with rock and air temperatures allows to present a simplified model of heat transfer processes with the surrounding limestone. In a first approach, only conductive heat exchanges are considered. In order to calibrate the model, observations are carried out about the ice mass of “La Glacière de Monlési” (Boveresse, Swiss Jura) at an altitude of 1135 m. Continuous temperature measurements permit a better definition of its thermodynamical characteristics and document the role played by water infiltration and convective heat transfers by air circulation around the ice filling.

1 INTRODUCTION of subsurface climate was stimulated by the work of Trombe (1952). During the last fifty years, several 1.1 Previous studies authors completed important studies on this particular subject (e.g. Cigna 1968, Andrieux 1969, Eraso 1978). Known for a long time by local people, ice caves have Encouraged by application on prehistorical sites, the been of interest to European naturalists since the sec- research on cave temperatures and air circulation ond part of the XVIIIth century. Various hypotheses, considerably improved the data acquisition and their more or less credible, were established on the genesis interpretation (e.g. Andrieux 1983). of such fillings (e.g. Kyrle 1923), among them a first Badino (1995) and Lismonde (2002) proposed the real scientific work can be mentioned: Thury (1861) first synthesis handbooks on cave climatology. How- made a detailed description of three ice caves in the Jura ever, the special problem of the physics of ice caves Mountains at different seasons. This author first dis- was still rather marginally considered. Our research played the importance of subsurface climate for the aims at filling the gap by applying heat transfer and fluid perennial preservation of ice in a cave. dynamics modelling to the study of ice caves. The pres- Throughout the last century several authors initi- ent note describes the first ideas developed to this pur- ated larger studies on some particular ice caves. Long- pose. Ice caves of the Jura Mountains are used for term measurements were started in Scarisoara Ice observations. Cave/Romania (e.g. Silvestru 1999, Racovita 1994, Viehmann 1976) and several meteorological measurements were carried out in the Dachstein massif/Austria 1.2 Definition and nomenclature (e.g. Mais 1999, Pavuza & Mais 1999) and Dobsinska Ice Cave/Slovakia (e.g. Lalkoviè 1995). Luetscher & Jeannin (2001) outlined the considerable These numerous studies brought a lot of field obser- confusion existing today in the nomenclature about ice vations and permitted a better understanding of differ- caves. In order to facilitate the understanding of the ent phenomena occurring in ice caves. However, only following text, ice caves are defined as: few of the presented ideas concerning the formation of “Natural karstic cavities presenting a perennial ice cave ice are based on the physics of heat transfers in filling.” such cavities. Such ice caves have been recognized for a long time Notions of static and dynamic caves were already as special features of perennially frozen ground, some- proposed by Thury (1861), but a real physical approach times occurring at very low altitudes (Haeberli 1978).

691 The wide definition used here is independent of any type of morphological cave description and different ice origins.

2 SITE DESCRIPTION

2.1 Ice caves in the Jura Mountains

The Jura Mountains form a ϳ400 km wide arc between the Savoie (France) in the South and the Black Forest (Germany) in the North. Their total surface is estimated at about 14,000 km2, and their highest peaks reach about 1700 m a.s.l. Mainly composed of karstic ter- rains, this massif includes over 10,000 known natural cavities. The altitude of the 0°C-isotherm of mean annual air temperature, is estimated in this region at approximately 2250 m a.s.l. Far below this level, the Jura Mountains include about 50 ice caves, located between 1000 and 1500 m a.s.l., with an ice filling volume varying between a few cubic meters up to more 3 than 5000 m . Figure 1. Plan view and vertical cross sections of the “Glacière de Monlési” – adapted from Luetscher & Wenger 2.2 The Monlési ice cave 2002. The position of the minilogger is mentioned by an *.

Located in the Swiss Jura Mountains (Boveresse/ An ice core sample could furnish further informa- Ne, 6°354/46°5618, 1135 m), the “Glacière de tion on the age of the ice. At the same time, isotopic Monlési” is the largest Ice Cave in the Jura. First analysis may provide additional information on cli- descriptions of its important ice filling were provided matic conditions in the Jura Mountains. by Browne (1865), while further studies were initiated by Stettler & Monard (Stettler & Monard 1960, Stettler 1971). Three entrance shafts allow to reach a large room of 3 ENERGY BALANCE AND HEAT approximately 25 m 45 m at a depth of 20 m below EXCHANGE ESTIMATION surface. Partially filled with an important mass of congelation ice, the lowest point of the cave can be 3.1 Definition of the system reached at 33 m through a large “rimaye” located on the south-western part of the “glacier”. Luetscher & An ice cave can be described as a system controlled Wenger (2002) estimated the total ice volume to be by the surface climatological conditions and those pre- about 6000 m3. vailing in the surrounding karstic massif. Unfortunately, the thickness of the ice mass cannot This system is defined as an ice volume filling a ver- be measured precisely. The presence of numerous cryo- tical shaft. The ice column is in contact with the sur- clastic sediments (e.g. Pancza 1992) and wood inclu- rounding limestone on its lower part and along the sions in the ice make the use of traditional geophysical vertical walls. The upper ice surface is in contact with methods (e.g. GPR) difficult. In the same way, steam the cavity air (fig. 2). As a first approach, necessarily a drilling could not yet reach the bottom of the total bit simplistic, we will only consider a two-dimensional ice mass. system. Up to now only few descriptions were prepared on the ice itself. The access provided by the “rimaye” enables 3.1.1 Basic assumptions observation of a 15 m ice thickness. Stratification The filling is expected to be essentially constituted of within the ice seems to indicate a maximum age of 100 congelation ice (e.g. Shumskii 1964). It is assumed that years for the lower levels of the observed sequence. This the cave acts as a cold air trap, i.e. air circulations due age is confirmed so far by the dating of a tile included to the possible presence of several entrances are neg- in the ice mass. Due to the young age of the material, lected in this first approach. dendrochronological correlation and dating could not In order to understand the thermal processes related be carried out on the numerous wood inclusions. to the thermodynamic behaviour of ice caves, energy

692 water and air temperatures (e.g. Badino 1995, Jeannin et al. 1997). The distance to the cave at which this fixed temperature boundary has to be set is not known so far. Depending on the fractures surrounding the cave, it may range between some meters to tens of meters.

3.2 Estimation and discussion of the respective heat fluxes

3.2.1 Ground heat flux In order to estimate the role of ground heat fluxes, steady-state, one-dimensional, conductive heat transfers are considered. To this purpose, the surrounding limestone is expected to be a homogeneous body (2.2 Wm1 K1) without any water infiltration. Heat fluxes due to conductive energy exchanges through Figure 2. Schematic representation of the studied system. the surrounding limestone can be expressed with the Energy transfers are represented by the different arrows. As a following equation: first assumption, the model is considered to be a cold air trap. TT L: distance between the cave and a constant rock temperature. Sl m (1) grL

5 where g ground heat flux [W]; S exchange sur- 2 face [m ]; lr thermal conductivity of limestone 0 1 1 [WK m ]; Tm rock temperature at the fixed temperature boundary; T temperature at ice-rock inter- -5 face; L distance [m]. Assuming a (constant) local temperature boundary of 5°C (the ice-rock interface is assumed to be at melt- -10

Temperature [ º C] ing point, i.e. 0°C) it is possible to estimate a plausible distance for “L” ranging between 3–20 m. Below this Air temperature -15 value, ice melt would probably be too important for a 07 08 09 10 11 12 01 02 03 04 05 06 07 long time conservation of the filling. On the other hand, Month higher values than 20 m would mean that conductive Figure 3. Air temperatures acquired during a one year heat exchanges through the surrounding limestone cycle in the Monlési ice cave present a typical cold air trap hardly affect the ice and would enable ice to grow up signature. close to the top of the shaft. Sensible heat fluxes characterise the heat storage in exchanges between the filling and its surroundings must the ice and surrounding limestone as related to their be determined. temperature changes. The cooling of the cave (air, but especially rockwalls and ice) during the winter season 3.1.2 Boundary conditions will therefore contribute to the total inertia of the Air temperatures in the cave constitute the upper bound- system. ary condition of the defined system. Due to melting, these temperatures cannot rise above 0°C and the ice 3.2.2 Fluxes at ice-air interface surface (i.e. cave air) has a well defined thermal max- Due to the geometry of the cave entrance and the veg- imum. As air exchanges within the cave are restricted etation around it, direct effects from radiative heat to winter time, minima are defined by air temperature fluxes can often be neglected. Hoelzle et al. (1993) during cold periods. Figure 3 presents a typical tem- already provided evidence for the possible presence of perature record acquired in an ice cave which acts as a low-altitude ice occurrences under similar conditions. cold-air trap. More important could be the role of latent heat Lower boundary conditions are characterised by a fluxes induced by air circulation which will largely constant rock temperature at a certain depth. It can be favour phase changes of water (sublimation, evapora- easily demonstrated that, due to the presence of a high tion and condensation). Besides their influence on heat water flux in karstic terrain, geothermal heat flux exchanges, condensation processes could possibly play is negligible and rock temperatures are controlled by a non-negligible role for the ice origin (Bock 1913).

693 Convective heat exchange with surface air have not Where S ground heat flux, M heat flux due been approached yet. Their importance is, however, to surface melt or freezing water, H sensible heat essential in the final energy balance of the system and flux, LE latent heat flux, R radiative heat flux, is at the origin of the cold air present in the cave. A anthropogenic heat flux. Two major processes control this energy balance, i.e. 3.2.3 Freezing and melting fluxes the presence of ice: water infiltration and air exchanges. Freezing, essentially affecting water infiltrated from The importance for the genesis and conservation of the surface or more rarely meltwater from the ice, is the ice filling from water infiltration is evident. Each usually restricted to the winter season. As freezing is an cave must have its own infiltration optimum depend- exothermal process, the released energy contributes ing on geometric characteristics of the conduits with to the warming of the cave. Negative temperatures will, their spatial frequency and fluxes involved. therefore, be rapidly equilibrated and very low temper- Too little infiltration prevents ice to form because atures will not persist for long periods as far as infil- of the absence of water, too much infiltration brings tration provides enough water. Convective exchanges too much heat into the cave and ice melts. A cav- with outside air induces warm air flowing out and being ity connected to fissures and conduits leading deeper replaced by colder air. This contributes to the cooling into the karst will “breath” a lot as a function depend- of the massif (Lismonde 2002). ing on inside/outside meteorology. High heat fluxes are If the ice filling is assumed to be in equilibrium, thus transferred through the cave and ice cannot occur at melting (mostly during summer) must compensate the low elevation. A cave closed at its bottom will act as a total ice produced in the cave during winter. Plausible cold trap and represents good conditions for ice to stay. values for the yearly ice deposit are in an order of The penetration depth of the fractures near the cave 2 100–500 kg/m . The energy necessary to melt such a vol- walls is important: depending on the length scale of the ume of ice corresponds to a heat flux varying between air and water transportation processes involved, the 2 1–5 W/m over the entire year. The less the ice volume system boundary might correspond to a fixed temper- is, the more the latent heat will constitute the essential ature bringing a high heat flux into the cave. factor influencing the thermal inertia of the cave. In order to better quantify the importance of each Water infiltration is necessary for the formation and parameter, field observations are carried out in several presence of ice but, also for the supply of an important ice caves of the Jura Mountains with one of them, the part of the energy used to melt the ice filling. With Monlési Ice Cave, being completely instrumented. annual precipitations of some 1500 mm and a temper- A better understanding of ice conservation processes ature difference of 5°C between the in- and output of should thereby become possible. the water-infiltration system, this energy supply can be estimated at about 1 W/m2. Snow accumulation constitutes an important cold reservoir and largely contributes to the total inertia of 4 FIELD OBSERVATIONS the filling. 4.1 First temperature records 3.2.4 Human impact Human impact can under certain circumstances play As a first step of the observational programme, air an important role on ice conservation/disappearance. temperatures were acquired on an annual cycle with Besides the energy supply resulting from each visitor, “UTL-1 Miniloggers” developed by the University of visits influence the natural state of the cave by enhanc- Bern/CH (Hoelzle et al. 1999). A mean annual air ing the exchange of air (turbulences) with the surface temperature of 0.9°C could be established during and, hence, by disturbing a rather static environment. the first annual cycle recorded (07.01–06.02) in the Ice exploitation was frequent in certain caves dur- Monlési ice cave (Fig. 3). The minilogger was located ing the 19th century and, could also have considerably in the rimaye of the northern part of the cave. disturbed the equilibrium permitting the conservation The collected data document the importance of the of ice. However, some ice caves seem to have com- thermal inertia induced by the melting of the ice filling. pletely recovered. The presence of ice prevents temperature to rise above 0°C during the whole summer season. 3.3 Energy balance of the proposed model The absence of significant water infiltrations during winter induced an important cooling of the sub- The energy balance of the defined system can now be surface air. expressed as following: Boreholes temperatures in the ice are measured since march 2002 with platine thermistors (Pt 100) at 1year ∫ ()SMHLERA 0 (2) different depths in the ice filling (0.1 m;1m; t0

694 2,5 m; 4m;5m;8 m). Due to many impurities REFERENCES of the ice (stones and wood), measurements at depths below 8 m are not yet possible. As a consequence, the Andrieux, C. 1969. Contribution à l’étude du climat des recorded temperatures are still influenced by seasonal cavités naturelles des massifs karstiques. Thèse variations. Long-term measurements are expected to d’Université, Bordeaux: 239 p. provide a better understanding on the inertia resulting Andrieux, C. 1983. Etude des circulations d’air dans la grotte de Niaux- Conséquences. Karstologia, (1): 19–24. from sensible heat flux. First values acquired during Badino, G. 1995. Fisica del clima sotterraneo, Memorie March 2002 already indicate temperatures close to the dell’istituto italiano di speleologia, 7 serie II: 136 p. melting point. Bock, H. 1913. Matematisch-physikalische Untersuchung In the same way, rock temperature are recorded at der Eishöhlen und Windröhren. in: Die Höhlen im different depths in small drill holes. These measure- Dachstein, Graz: 102–144. ments improve our knowledge on water circulations Browne, G.F. 1865. Ice caves of France and Switzerland, affecting the surroundings of the cave and help to better Longmans, Green, and co., London: 315 p. define the importance of the rock heat flux. Cigna, A. 1968. Air circulations in caves. Proceedings of the 4th International Congress of Speleology, Llubljana: 43–45. 4.2 Water infiltration Eraso, A. 1978. Vision termodinamica de los fenomenos karsticos. Espeleolosie, 21: 11–16. More than ten temporary infiltration points (from less Haeberli, W. 1978. Special aspects of high mountain per- than 0.1 l/min up to more than 10 l/min) have already mafrost methodology and zonation in the Alps. Abstract volume, the Third International Conference on Per- been recognised in the cave. The importance of each mafrost, 1: 379–384. of them must be quantified in order to determine the Hoelzle, M., Haeberli, W., Keller, F. 1993. Application of volume of ice produced and the melting occasioned BTS-measurements for modelling mountain permafrost by the energy supply. Frequency and flow will play distribution. Sixth International Conference on Per- a major role in the final mass and energy balance of mafrost, 1: 272–277. the filling. Hoelzle, M., Wegmann, M., Krummenacher, B., 1999. Miniature Temperature Datalogger for mapping and monitoring of permafrost in high mountain areas: first 5 CONCLUSION AND PERSPECTIVES experience from Swiss Alps. Permafrost and periglacial processes, 10: 113–124. Jeannin, P.-Y., Liedl, R., Sauter, M. 1997. Some concepts Consideration of conductive heat transfer processes in about heat transfer in karstic systems. Proceedings of the ice filling of an ice cave helps with the design of an the 12th International Congress of Speleology, vol. 1: observational programme. The strong inertia provided 195–198. by the presence of ice points to marked effects of auto- Kyrle, G. 1923. Grundriss der theoretischen Speläologie. regulation of the underground air temperature by stor- Druck der Oesterreichischen Staatsdruckerei, Wien: age of “cold” in ice and rockwalls. 353 p. Comparison with long-term field measurements Lalkoviè, M. 1995. On the problems of ice fillings in the indicates the a priori importance of water infiltration as Dobsina ice cave. Acta carsologica, XXIV: 312–322. related to conductive heat transfers as well as for the Lismonde, B. 2002. Aérologie des systèmes karstiques. final mass balance of the ice. An identification of the CDS Isère: 362 p. Luetscher, M. & Wenger, R. 2002. Nouveau levé respective role of each parameter should allow a better topographique de la glacière de Monlési. Cavernes, knowledge of heat exchanges in an ice cave. The under- 1–2002: 9–15. standing of the climatological conditions determining Luetscher, M., Jeannin, P.-Y. 2001. Les glacières du Jura: ice preservation should be considerably improved by synthèse des connaissances et directions de recherche, the combination of field observations and numerical Actes du 11è Congrès National de Spéléologie, Genève modelling. 2001: 119–124. Mais, K. 1999. Untersuchungen des Höhlenklimas in der Dachstein-Rieseneishöhle von 1910–1962, Die Höhle, ACKNOWLEDGEMENTS Heft 3: 118–140. Pancza, A. 1992. La gélivation des parois rocheuses dans une glacière du Jura Neuchâtelois. Permafrost and The authors would like to thank Prof. M. Beniston periglacial processes, vol. 3: 49–54. (UNIFR) for his participation in this project and Pavuza, R., Mais, K., 1999. Aktuelle höhlenklimatische F. Bourret for the help brought in the acquisition of Aspekte der Dachstein-Rieseneishöhle. Die Höhle, Heft the field data. This work is part of a larger study on ice 3: 126–140. caves in the Jura Mountains supported by the Swiss Racovita, G. 1994. Bilan climatique de la grotte glacière de National Found No. 21-63764.00. Scarisoara (Monts du Bihor, Roumanie) dressé sur dix

695 années d’observation, Trav. Inst. Spéol. Emile Racovitza, Stettler, R. 1971. La Glacière de Monlési (Boveresse, NE). XXXIII: 107–158. Actes du 4è congrès national de spéléologie: 138–149. Shumskii, P.A. 1964. Principles of structural Glaciology. Thury, M. 1861. Etude des Glacières naturelles. Archives des Dover Publications, Inc.: 497 p. sciences de la bibliothèque universelle, Genève: 59 p. Silvestru, E. 1999. Perennial ice in caves in temperate climate Trombe, F. 1952. Traité de spéléologie. Paris, Payot: 376 p. and its significance. Theoretical and applied karstology, Viehmann, J. 1976. Dix ans de recherches périodiques dans vol. 11–12: 83–94. une grotte de glace (la grotte de Scarisoara, Roumanie). Stettler R., Monard M. 1960. La Glacière de Monlési. Actes du 6ème Congrès Int. de Spéléologie T. III: Cavernes no. 1: 1–10. 323–327.

696