Synthetic study has
been undertaken using the bias-corrected ECMWF
forcing for 1987. To mimic the space-borne
observations inferred from the passive microwave brightness
temperature measurements, we have used model
output of the total snow water equivalent of the
snowpack at each catchment.
The
snow observation and truth data sets were generated by the
CLSM after it had been spun-up to equilibrium. A
degraded simulation
was then made using a poor initial condition for
the snow states. Specifically, it is
prescribed that there is no snow present.
The
assimilation run starts from the poor initial state, and
uses the observations generated from the truth run. The assimilated
results are then compared with the truth and
degraded model simulations. Observations are
assimilated once every three days.