CRYOMET - Bridging models for the terrestrial cryosphere and the atmosphere

In early summer, when people in the Norwegian mountains exchange their skis to hiking boots, obstacles are encountered along the emerging trails: snowfields from the past winter. The pattern of snowfields, sprinkled across snow-free areas is a common sight in most mountain areas well into summer. The main reasons for this pattern are strong winds and frequent snow storms which move around freshly fallen snow. Such snowdrift leads to snow depths of several meters in some locations, while other areas remain snow-free all winter. The resulting snow distribution, along with the surface air temperature, has important consequences for many phenomena, such as the spatial distribution of permafrost or the location of glaciers. Furthermore, the snow distribution influences the energy exchange between the land surface and the atmosphere. CryoMet aims for an improved representation of spatially variable snow depths in atmospheric modelling schemes. One of the major problems is that the models simulating climate variables are operating on coarse spatial scales so that the spatial variability of the snow cover and other climate variables is not captured. Therefore, coupling models between the atmosphere and the land surface is a challenge, and demands up-scaling and down-scaling of earth system models. During the CryoMet project some major results were achieved: 1. We used dynamical downscaling to produce maps of energy balance components to drive permafrost and glacier mass balance models. We produced such maps for Svalbard in a ground resolution of 3 km, for local areas in a ground resolution of 1 km. Such maps are of interest also for other disciplines, such as biology, 2. We combined the results obtained by downscaling for Svalbard with satellite-based surface temperature for the whole north-Atlantic region, producing a 10-years time series of the energy balance components. We used this information to force simple permafrost models, which in turn formed a basis for funding from the European Space Agency to apply this concept globally. 3. We measured snow variability and ground temperatures at several field sites in Norway and Svalbard and successfully reproduced ground surface temperatures as probability density functions (PDFs) in models. 4. This was a major step towards the implementation of sub-grid variability in different models. We applied these PDFs for a new permafrost model approach over the Nordic area. Furthermore, we used these results to parameterize the snow distribution in a weather model, and it is obvious that these changes have large influence on the results. 5. The new land-surface scheme CryoGRID 3 dedicated to permafrost processes was assembled during this project, in close collaboration with international partners from Alfred-Wegener-Institute Potsdam, Germany, and LGGE Grenoble, France. In addition to UiO, CryoGrid 3 is today already used by 5 international partners for a variety of topics. The achievements from this project have triggered new projects, funded by the Research Council of Norway (RCN) and the European Space Agency (ESA), The project inspired also the establishment of a cross-disciplinary research group at the University of Oslo, which major aim is to evaluate land-atmosphere interaction in cold climates (LATICE). This initiative is awarded as one of UiO?s Faculty of Mathematics and Natural Sciences focus areas.

Glaciers and permafrost are the two main components of high-latitude mountain and Arctic environments, and both will be affected by the changing climate system. Earth System Models (ESMs) provide projections of the climate evolution for the coming decade s. However, they operate on coarse spatial grids with grid spacing of 50 to 300 km, which is too coarse to resolve the variations in topography and land surface properties. Therefore, downscaling is a prerequisite for a detailed impact assessment, especia lly in mountain areas where topography and surface properties can vary on the scale of meters. Here, snow is a crucial moderating factor. CRYOMET will test and develop seamless downscaling procedures, that are designed to bridge the gap from ESM models t o the process scale. This will be achieved by combining deterministic downscaling using the state-of-the-art regional climate model PolarWRF with probabilistic downscaling of snow depth using snow redistribution models. This scaling concept is aimed to be capable of bridging about five orders of magnitude in space without inflicting a scaling gap. In a second step, CRYOMET will upscale the surface energy balance and thus the feedback of the spatially variable snow depth. CRYOMET will explore the potent ial of this scaling concept, which could be transferred to other research areas to significantly advance modelling capabilities for arctic ecosystems. Extensive field data sets for the ground thermal regime and the surface energy balance are available to this project. Those will facilitate validation and improvement of the formulations of the surface energy balance in the employed atmospheric models. The validation sites follow environmental gradients, with the main focus on Svalbard and mountain sites i n both northern and southern Norway. As a spin-off, CRYOMET aims for spatially refined model representations of both the thermal regime of permafrost and glacier mass balance for several areas on Svalbard.