Simulation of Cold Processes in the CMIP6 Land-Historical Simulations

Model evaluation is a necessary component of climate research. While the snow and soil components of land-surface models have been previously evaluated using standalone simulations (Dirmeyer et al., 1999; Slater et al., 2001; Dirmeyer, 2011), the Land Surface, Snow and Soil moisture Model Intercomparison Project (LS3MIP; van den Hurk et al., 2016) is the first CMIP framework that coordinates both components and integrates them into the larger suite of CMIP experiments. Here we analyze output from the offline global land surface simulations, forced by historically observed meteorological conditions, and compare them to the fully coupled and AMIP historical CMIP6 simulations. In this context, we focus on high-latitude processes and variables, in particular snow and permafrost. We demonstrate that there are marked differences in the representation of snow between the coupled and offline simulations, in both their climatological extent and total mass, as well as their variability. For snow, differences in the balance of accumulation and ablation between the coupled and uncoupled simulations yield differences in climatological snow extent and snow mass. However, these differences do not mar the representation of historical snow variability. The strong influence of driving temperature and precipitation variability results in a clear improvement of historical snow variability in the offline simulations for both long term trends and interannual variability. For permafrost, errors in the functional relationships that control permafrost extent and soil temperature are large and much more important than differences due to improved representation of historical temperature, precipitation or other meteorological drivers. These differences suggest that the added value of standalone land model simulations in the suite of CMIP6 experiments will depend on the land-model variable of interest. References: Dirmeyer (2011), https://doi.org/10.1175/JHM-D-10-05010.1 Dirmeyer et al. (1999), https://doi.org/10.1175/1520-0477(1999)080<0851:TPPOTG>2.0.CO;2 Slater et al. (2001), https://doi.org/10.1175/1525-7541(2001)002<0007:TROSIL>2.0.CO;2 van den Hurk et al. (2016), https://doi.org/10.5194/gmd-9-2809-2016