debiasing model data

Could anyone suggest an approach to debias monthly model precipitation
data against observations, both in terms of correcting the mean and the
variance? I am using observational data, a model 20th century
simulation,
and a transient model simulation of the future.

When correcting the mean, a percentage correction usually works, but
over arid regions, where unreasonable values can result. I am not sure
how to correct the variance.

Sorry if this is not directly an NCL question.

Thanks,
Michael
_______________________________________________
ncl-talk mailing list
ncl-talk_at_ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Fri Feb 23 2007 - 11:43:17 MST