Hydrologic Prediction

Assimilation of Satellite Snow Data

The problem of how to predict snow water equivalent (SWE) using numerical models still requires considerable attention. EOS satellite-based remotely sensed snow data provide more realistic information about the snow distribution and quantity than the model predictions. In studies involving the SAST model coupled with the NCAR GCM (CCM3), we replaced the modeled snow cover over the Rocky Mountains with satellite SMMR monthly snow coverage and SWE data. Significant consequences in terms of both the land surface and atmospheric processes were observed: (1) previously overestimated soil moisture became much closer to the long-term reanalysis values, (2) runoff in the Missouri River basin decreased, (3) regional surface temperature from March to September increased 1 ° -1.5 ° C and became closer to the observations, and (4) summer (JJA) precipitation in the SWUS and Great Plains improved from original substantial underestimation (Jin, 2002). We intend to follow up on these promising results by investigating the physical processes giving rise to the observed changes and implementing appropriate assimilation techniques using the EOS snow data.

When assimilated into CCM3, the community climate model, MODIS snow information resulted in significant changes of model's prediction of precipitation for long-term simulation.

MODIS snow products are available.