[Wrf-users] Grell parameterization.

[jruiz at cima.fcen.uba.ar]

Hello
My name is Juan Ruiz and I ve been testing WRF over South America region.
We have performed a brief precipitation forecast verification using CMORPH
data. The verification period is 38 days between october and november
2005. One of the most interesting results is that using Grell cumulus
scheme the 48 hour BIAS is much smaller than the 24 hours BIAS over South
Eastern South America. For example at 24 hour forecast the BIAS for a 10
mm threshold is 1, for the same threshold but for 48 hour forecast the
BIAS is 0.7. Other thresholds shows similar results. Other
parameterization like for example K-F don't show such a strong depedence
of the BIAS on the forecast time. Changes in other physics (LSM, boundary
layer) don't change the dependance of the BIAS on the forecast time.
Has anybody found a similar behaviour? This could be a result of this
particular period (which is short).
Thancks a lot in advance.
Juan.