Improving the understanding and diagnosis of Earth system changes in cold regions

Abstract:

I review key hydrological state variables and fluxes relevant to cold regions, specifically snow, permafrost and seasonally frozen soils, lakes, and wetlands, and comment on the ability of current models to represent the associated processes, and the quality of the data sets upon which model development and diagnosis efforts rest. Although snow processes are relatively well represented in current generation land surface models, at least at large scales for deep mountain snowpacks, the representation of high latitude snow processes remain complicated by the role of snow redistribution, and of sublimation during the shoulder (especially spring) season. Most credible land surface models now include representations of permafrost, some of which perform well when forced with local climate data; however their performance over large areas is limited by spatial variability of key processes, including soil thermal characteristics. Likewise, many land surface models now represent the hydrology and energetics of lakes, which cover a substantial portion of the landscape in many high latitude environs. However, accurate representation of lakes requires knowledge of certain characteristics of their bathymetry and hydrological connectivity, information which is not always available. Likewise, the representation of wetlands in models, although improved in many cases, is limited by topography (and the role of microtopography, even at large scales). Nonetheless, increased attention to high latitude hydrological processes has demonstrably improved the fidelity of land surface representations over the last decade or so.