[RegCNET] Problem using NNPR2

[Prof.Mkankam Kamga François]

Hi all,
I have this particular problem using the NNPR2 reanalysis as BC in RegCM3.
The program becomes very slow as compared to when using ERA40 dataset. The
NNPR2 data I have is  an African window subset.
I notice that the program says that first and second derivatives of ps are
NaN.
Any idea what the problem may be? I lowered timestep as suggested by
Jeremy but  no improvement.
Best regards
Kamga


-- 
Prof. MKANKAM KAMGA François
Laboratory for Environmental Modelling and Atmospheric Physics
Department of Physics, Faculty of Sciences, University of  Yaounde I
P.O. Box 812 Yaoundé, Cameroon
Tel C. (237) 782 3868  H. (237) 231 5665

fmkankam at uycdc.uninet.cm , mkankam at cenadi.cm