Invited key note address The cryosphere: an early indicator and global player Hartmut Grassl, Max Planck Institute for Meteorology Ice at or below the surface of the planet Earth is an important part of the climate system. The solid phase of water has two unique characteristics which make it both an early indicator of climate change and a global player. First, if warmed to the melting point at O T , higher air temperatures and/ or higher long-wave back radiation just increase the melting rate but not - as with all other surfaces - the temperature, which stays at 0°C. Small ice caps and mountain glaciers thus become early indicators of a changed climate. Second. if seawater is cooled to the freezing point at about - 1.8"C. the sea ice formation process ejects salt causing the denser water to sink, thereby filling the global ocean interior with very cold water. The location where most of this deep convection occurs is strongly dependent on the freshwater balance and thus on the average salinity of ocean basins. Present ocean configuration and ocean topography, as well as precipitation distribution, make the northern North Atlantic more saline than any other high latitude ocean part and thus the site with most of this deep water formation. Sea ice formation is therefore of high significance for the European climate. Since it drives the near surface warm North Atlantic current northward off the European coast in compensation for southward deep water flow in the western Atlantic, north- western Europe is warmer by about 4°C than the same latitudes on the eastern Pacific coast of America. Since I restrict myself to climate-related aspects of the cryosphere, I will not speak about reactions of the biosphere and high latitude populations to global change, but I will include cryospheric processes outside the polar latitudes. A few facts: the cryosphere 0 is the largest freshwater reservoir: 98% of all freshwater is fixed in the Antarctic and Green- land ice sheets as well as in the other small ice caps and mountain glaciers; 0 contains the brightest natural surface: fresh powder snow reaches an albedo of about 90% as compared to about 15 to 20% for grassland; 0 is responsible for a positive feedback: melting of high albedo surfaces (snow and ice) increases absorption of solar radiation which in turn increases temperature and accelerates melting of nearby snow and ice surfaces; 0 emits least heat to space: the low surface temperatures of Greenland and Antarctica cause the lowest emissions to space both in winter and summer for their respective hemispheres; 0 insulates the ocean from heat loss: sea ice reduces all three heat losses of a surface (long- wave net radiation, sensible heat flux, latent heat flux); the latter two to a small fraction of the values for an open ocean area; 0 fixes large amounts of carbon dioxide ( C 0 2 ) and methane (CH4): organic material in permafrost is to a large part emitted as CO? and CH4 after the melting of permafrost; 0 changes by more than a factor of two within an annual cycle: for example, Northern Hemi- sphere snow cover reaches 46.5.106 km' i n late January and is lowered to 3.9.10hkm2 in September; 0 creates the cold interior of the global ocean: Antarctic bottom water and North Atlantic deep water formed by deep convection are the main water masses of the global ocean, with tem- peratures mostly below about 3°C even in tropical latitudes. Some recent findings This selection of examples tries to avoid overlap with other presentations in this volume. Therefore, the following three examples will be briefly presented: remote sensing of mountain glacier retreat; stability investigations of the thermohaline Grassl 1999: Polar Research 18(2), 119-125 119 t ' 0 - n e O . $ -lo c M . c v . P - 4 0 - m - '; -20 2 -30 .e Y - 2 -50- 4- -70 Fig. I. Glacier retreat in the Austrian Alps: ( a ) relative change in area in 1992 versus area i n 1985 as a function of glacier area; (b) change in area with exposition (from Paul 1997. with perrnigsion of author). i + t 4 . 8 % : x 1 2 5 A $ 0 - 8 x 4 O O ; + * - " $ B o n - x ; : $ * A X - 0 X V 0 Stubai Nollh 0 Stuba South + Otaal North v Omal East x Zillerlal *average v A Otaal West t o + ' ' ' StubaiNorth StubaiSouth - OtaalNollh - Otaal West - OrnalEast overturning in the Atlantic; and selection of a sea ice rheology code best suited for global climate models. Alpine glacier retreat: Small valley glaciers react comparably rapidly to temperature and precipita- tion changes. Their retreat depends strongly on their length. mean slope, ratio of accumulation to entire surface area and orientation. Landsat images with sufficiently high spatial resolution exist since 1973 for a trend analysis. As one of my students (Paul 1997) has shown, for many glaciers in the eastern Alps, the 19 year period from 1973 to 1992 was one of strong retreat with clear dependencies on orientation and size of the glaciers. As Fig. 1 demonstrates, even the 1985 to 1992 period gave clear signals. The main reason for the retreat in this area with no significant change of annual precipitation during the recent decades was a rather rapid warming. Thermohaline circulation changes in the Atlantic: At present the meridional overturning of the Atlantic drives the so-called global ocean con- veyor belt. Palaeoclimatic evidence points to rapid changes of the deep water formation rate in the high latitude Atlantic, the core region of the conveyor belt. As also shown by modelling studies (e.g. Rahmstorf 1996) the stability dia- gram (see Fig. 2 ) of the conveyor contains a large 120 The cryosphere: an early indicator and global player -0.1 0 0.1 Freshwater forcing (Sv) F i g . 2. North Atlantic deep water (NADW) formation rate ;IS a function of additional freshwater input into the Atlantic north of 30 S (modified from Rahmstorf 1999, with permission of the author). Solid lines show stehle equilibrium states: unstable states are dotted. Note the strong hysteresis as well a h the different mechanisms responsible for the cessation ot deep water formation. Transition mechanisms include: a ) local convective instability; h) polar halocline catastrophe: c) advective spin-down: and d) start-up of convection and NADW formation. hysteresis. The addition of freshwater north of 30's in the Atlantic - only 0.1 Sverdrup (1 Sv = 1 0 h m 3 s - ' ) is sufficient - stops deep water formation, which will only start again if all the additional freshwater input is removed. The cryosphere (for example, due to additional melt- water from decaying ice sheets) has exerted this Aug Feb Aug Feb Aug Feb Aug Feh Aug 1990 1991 1992 1993 1994 A O D F A J Month (4 24c g 35 25 .2 20 2 15 E 5 C 30 TJ x c 10 x o 5 1 2 3 4 5 6 7 8 9 1 0 1 1 12 Month Fig. 3. (a) First multi-year time series of sea ice draft i n Fram Strait: ( h ) the annual cycle; and (c) the maximum draft during different months (T. Vinje and N. Nordlund, pers. cornm.). Ice thickness = draft x 1.136. Ice export per year = ca. 2.8 x 10'km'. influence in the past, but a change in precipitation a n d o r evaporation over the Atlantic would also suffice. Much more research is needed to find out T u h k 1. Similarities and contrasts among polar regions. NH and SH indicate whether the intersection of a cryospheric element w i t h a climate process is specific to the Northern or the Southern Heniisphere, or occurs i n both. Climate processed cryospheric elements Ice Glaciers. Frozen rivers. FroLen Ice sheet\ \helve\ ice caps Sea ice Snow laker ground Sen level change Deep water tormatioii Surtace energy halance Albedo Latent and wnsible heat OLe.iniL \urt.ice buoyancy f l u x Precipitation Melting/ahldtion freezing/accumulation lceherg calving Soil moisture (run-off) Radiative processes (through changed atmospheric composition) NH trend NH SH. catastrophic SH NH, SH SH NH SH NH NH SH NH, SH NH.SH NH.SH NH N H , S H N H , S H NH NH NH, SH N H , S H NH.SH NH NH NH NH NH Grass1 1999: Polar Research 18(2). 119-125 121 TahL 2. The state of observational data for the Northern (NH) and Southern (SH) hemispheres, and international research programmeslgroupslinstitutions that have contributed or will contribute to the data. D- indicates absence of necessary data; D+ indicates availability of some data; Do indicates lack of information on the data required. Climate processed Ice ' Glaciers. Frozen rivers, Frozen cryospheric elements Ice sheets shelves ice caps Sea ice Snow lakes ground Sea level change D- NH: DO/- MAGICS MAGICS ISMASS SH: D- WAISS Deep water formation D- FRISP Surface energy balance. Albedo NH: Do SH: D + GEWEX Latent and sensible heat D- FRISP Oceanic surface buoyancy flux. Precipitation Melting/ablation. freezing/accumulation D- D- NH: DO/- MAGICS FRISP MAGICS ISMASS SH: D- WAIS Iceberg calving Do/+ DO/+ Soil moisture (run-off) Radiative processes (through changed atmospheric composition) NH: D + ACSYS CLIVAR SH D- iAnZone ASPECt D + D+ Do ACSYS GEWEX GEWEX N H . D o D+ D+ ACSYS ECMWF ECMWF SHEBA NCEP NCEP ASPECt WCRP SH. D- D- Snowdnft D + Do NH.ACSYS NH. SH. ASPECt ACSYS WCRP SH: iAnZone ASPECt Do D o GEWEX GEWEX Do ECMWF D o GEWEX D- IGAC whether the global warming by an enhanced greenhouse effect would bring us nearer to such a rapid change. Sea ice dynamics in climate models: The coupling of ocean - sea ice and the atmosphere as well as of soil, vegetation and atmosphere is among the major challenges of climate modelling. Sea ice dynamics is an important part in this coupling. The Sea Ice Modelling Panel (SIOMP) of the Arctic Climate System Study (ACSYS) of the World Climate Research Programme (WCRP) conducted sea ice dynamics model intercom- parisons and found that only the viscous-plastic model could reproduce the observed sea ice thickness distribution in the Arctic (Lemke et al. 1997). Global climate models therefore should adopt this code for a more realistic representation of sea ice in global coupled climate models. The Arctic Climate System Study The different projects of WCRP are normally global since both the atmosphere and the ocean circulate globally. Only the Arctic Climate System Study (ACSYS) has a regional focus, i.e. i t 122 T h e cryosphere: an early indicator and global player Tuble 3. The state of modelling for the Northern ( N H ) and Southern ( S H ) hemispheres. and international research programmed groups/institutions that have contributed or will contribute to modelling efforts. M + indicates that some modelling activities are underway; M- indicates insufficient modelling efforts; Mo indicates lack of information on sufficient modelling activities. Climate processes/ Ice Glaciers. Frozen rivers, Frozen cryospheric elements Ice sheets shelves ice caps Sea ice Snow lakes ground Sea level change Mo/+ NH Mo EISMlNT MAGICS SH M- Deep water formation M + M + FRISP ACSYS i AnZone i AnZone EISMINT ice-ocean coupling Surface energy habdnce Albedo Mo M + Mo EISMINT ACSYS WGNE: lAnZone Albedo LX = a (T. surface type) Latent and sensible heat M+ M+ Mo FNSP ACSYS WGNE, EISMINT iAnZone subgnd i AnZonr parameter- ization (Land) Oceanic surface buoyancy flux. Precipitation Mo Melting/ahlation, Mo/+ M+ NH:Mo M + Mol+ freezingiaccumulation EISMINT FRISP MAGlCS ACSYS ACSYS EISMINT SH: M- IAnZone tAnZone iAnZonr Iceberg CdlVlng M- M- Soil moisture (run-off) Mo Radiative processec (through WGNE changed atmosphenc composition) Mo GEWEX M + Mo ECMWF ECMWF NCEP Mo M o GEWEX WGNE Mo CLIVAR ACC concentrates on the Arctic Ocean and sea ice. Its objectives are: 0 to provide the scientific basis for the representa- tion of the Arctic region in coupled global atmosphere-ocean models; 0 to develop an effective climate monitoring scheme in the Arctic; 0 to carry out scenario computations for specified large-scale atmospheric conditions, in order to evaluate possible impacts of climate change in the Arctic region. started in the early '90s. A good example of early achievements is provided in Fig. 3, which displays the first multi-year time series of sea ice draft i n Fram Strait measured with upward looking sonars mounted on top of ocean moorings. It underlines the importance of sea ice export through Fram Strait into the Atlantic. 2800 km3 are exported per year which is - if compared to river run-off - roughly 20 times the discharge of the Congo River. Astonishingly, the thickest ice is exported in summer and interannual changes are large, spanning the range from about 2000 to 4000 km3, ACSYS formally implements its goals since 1994 although some scientific activities had already thus contributing to large salinity anomalies in Greenland Sea waters. Grass1 1999: Polar Research 18(2), 119-125 123 Table 4. Existing international programmes/projects/activities. Status: I = implementation; 1P = implementation plan exists. Acronym Full title Hemisphere Status Focus Sponsoring institution ACSYS Arctic Climate System Study IABP International Arctic Buoy Programme IPAB International Programme of Antarctic Buoys ANSITP Antarctic Sea Ice Thickness Project iAnZone International Antarctic Zone Program CALM Circumpolar Active-Layer Monitoring Programme GCD Global Geocryological Data Base MAGICS Mass Balance of Arctic Glaciers and Ice Sheets i n Relation to the Climate and Sea Level Changes CLIVAR-D5 Southern Ocean Thennohaline ASPECt ISMASS WAIS* FRISP* EISMINT" CRYSYS* Circulation Focus in CLIVAR Climate Level Antarctic Sea-Ice Processes and Ice Sheet Mass Balance and Sea NH NH SH SH SH NH Both Both, focus on NH SH SH SH I I I I I Coupling of the circum-Antarctic Ocean and the atmosphere I Active layer thickness in continuous permafrost areas I Retrieval, documentation, archiving and distribution of permafrost data 1 Circumpolar mass balance studies Arctic Ocean circulation and sea Drifting buoys on sea ice Drifting buoys on sea ice Sea ice thickness from upward- ice looking sonars IP Southern Ocean variability and IP IP climate variability and environment change Sheet 200 years of past Antarctic climate Mass balance of the Antarctic Ice West Antarctic Ice Sheet SH I Understanding the dynamics of the Initiative WAIS Filchner-Riinne Ice Shelf Project SH I Understanding ice shelf-ocean European Ice Sheet Modelling SH I Understanding the dynamics of ice Initiative sheets, including ice shelves Canadian contribution to NASA's NH '? Use of the cryosphere to monitor EOS on the Cryospheric System interaction climate change WCRP Operational and WCRP WCRP SCOR (ICSU) research institutions IPA (rcsu) IPA (ICSU) IASC WCRP SCAR (ICSU) SCAR (ICSU) USA consortium European consortium Mainly European consortium Canadian initiative *National or multi-national activities with research topics falling within the CLIC domain (incomplete). Another achievement is the sea ice model intercomparison which has led to a recommenda- tion for climate model sea ice codes (Lemke et al. 1997): modelled sea ice dynamics using the viscous-plastic model are nearest to observations of sea ice extent and thickness and should thus be adopted for global coupled climate models. No overlaps, but still some gaps Soon after the start of ACSYS, discussions for a broader approach of research on climatically relevant cryospheric processes were initiated inside and outside WCRP. An expert meeting on Cryosphere and Climate organized by WCRP and hosted by the British Antarctic Survey in Cam- bridge, UK, from 3 to 5 February 1997, developed recommendations for WCRP concerning the broadening of climate-related cryospheric research (WCRP 1998). One of the main findings was that despite the large number of programmes/projects/activities, there were nearly no overlaps but still some gaps i n climate-related research of cryospheric processes. This is detailed i n Tables I to 3, which first show the processes relevant i n the different cryospheric elements and the contrasts between the hemi- spheres, and then point to data gaps and insuffi- cient modelling. The major recommendation to WCRP was that a global activity on cold regions and climate within WCRP should be developed whereby already successfully implemented pro- jects should not be disrupted. The recommendations of the meeting led to the establishment of a Task Group on Climate and 124 The cryosphere: an early indicator and global player Cryosphere (CLIC) by the Joint Scientific Corn- mittee (JSC) for WCRP. by seasonal climate anomaly predictions related to tropical sea surface temperature anomalies, which are available now from global modelling centres, shifts the main climate research topics to other processes and to longer time scales. Cryospheric processes thus come more into the focus of climate research. For example. improved seasonal predic- and Cryosphere: a WRCP initiative With the establishment of the JSC/ACSYS Task Group on Climate and Cryosphere (CLIC) at the nineteenth session of JSC in Cape Town. South Africa, in March 1998, a global Climate and Cryosphere project of WCRP became highly probable. That the CLIC Task Group is also charged with the formulation of a coordination plan can easily be understood if one looks at Table 4 which lists existing international programmes/ projects/activities engaged i n climate-related cryo- spheric research. Coordination will be facilitated by the fact that many of the ongoing activities are sponsored by committees or associations of the International Council for Science (ICSU) which is also a sponsor of WCRP, i n addition to WMO and UNESCO's IOC. The two major goals of CLIC - as far as I can see them now - will be: 0 understanding climatically relevant cryospheric processes, including their role as early indica- tors of climate variability and climate change; 0 understanding and predicting - to the extent possible - the influence of the cryosphere on global climate. A first outline of the science plan was drafted at the first CLIC Task Group meeting which took place i n Utrecht, The Netherlands, from 8 to 1 1 July 1998. The report of the meeting, including the outline, will be discussed by the ACSYS Scientific Steering Group i n Tokyo i n November 1998 before i t is submitted to the JSC for WCRP in March 1999 in Kiel, Germany. Outlook The improved understanding of seasonal to interannual climate variability, as it is expressed tions i n the Asian monsoon area may need better data on central Asian snow cover. The under- standing of the Arctic oscillation depends on improved data and modelling of sea ice - ocean - atmosphere interaction; and the understanding of the influence of the enhanced greenhouse effect on global thermohaline ocean circulation needs, in addition to oceanographic information, data on the long-term behaviour of snow extent, run-off or calving from ice sheets. It is time for CLIC, and since all institutions implementing or planning projects on climatically relevant cryospheric processes have indicated their preparedness to cooperate, I see no major obstacles for the proper coordination of the existing activities and the filling of the gaps identified i n a new global WCRP project Climate and Cryo- sphere. References Lemke, P.. Hihler, W. D., Flato. G., Harder. M. & Kreyrcher, M. 1997: On the improvement of sea-ice models for climate simulations: the Sea Ice Model Intercomparison Project. Ann. Clacid. 25, 183-1 87. Paul, F. 1997: Changes of glacier area i n the Austrian Alps hetween 1973 and 1992 derived from Landsat data. M a x Plunck Institute for Meteorology Report no. 242. Rahmstorf. S . 1996: On the freshwater forcing and transport of the Atlantic thermohaline circulation. Clirn. Dyn. 12. 799-8 I I. Rahmstorf. S. 1999: Decadal variability of the thennohaline circulation. I n A. N a v m a (ed.): Eryond El NiAo: clecudul and inrrrdrcadal c./imatr variubilin. Pp. 309-332. Berlin: Spring- er. WCRP (World Climate Research Programme) 1998: 0rgoni:u- t i o r i of internatiorml/y co-nrdinated resenrch in cr?.osphere and clitnutr: proceedings of a meeting oJ r.rperl.7 or1 crysphere arid climate. Cambridge. UK. 3-5 February 1997. R . G. Barry. ed. WMO/TD No. 867. Geneva: World Meteorological Organization. Grass1 1999: Polar Research 18(2). 119-125 I25