Snow Data Assimilation in a Catchment-Based Land Surface Model
Chaojiao Sun, Jeffrey P. Walker and Paul R. Houser
Snow plays an important role in both the global energy and water budgets, as a result of its high albedo, thermal properties and being a medium-term water store. However, it is difficult to forecast snow accurately in atmospheric and hydrologic models, since the scale of its variability is usually smaller than that resolved by most models, and there are errors in model forcing data. To offset such errors, observed snow water equivalent may be assimilated into the snow model by the method of direct insertion, but this may induce artificial moisture transfer. Hence, a data assimilation scheme that optimally merges snow observations with model forecasts is needed.
We present a data assimilation study for North America with the catchment-based land surface model (CLSM) used by NASA Seasonal-to-Interannual Prediction Project (NSIPP), which includes a three-layer snow model. We use a one-dimensional Kalman filter for each catchment of the CLSM to sequentially update model states and model error covariances. This scheme treats artificial systematic snow melt as a result of bias in the land surface and/or air temperature.
Results from identical-twin experiments are presented, in which the observations and true states are generated by the same model. A retrospective forcing data for the past 20 years for North America is used. Special attention is paid to the updating of land surface temperature through snow data, so that erroneous snowmelt may be prevented. In a follow-up study, we will assimilate SMMR and SSM/I satellite-measured snow data into the model.