[Wrf-users] Time Series Start Date

[Capehart, William J]

We've been working with Timeseries for some time and have just noticed this "feature":

Is the time for the time column for the TS files series files leveraging the wrfinput file's "SIMULATION_START_DATE" which when running a model in "cycling" model with 3DVAR may not be the same as a given actual model start date ("START_DATE")?   We're seeing time series files with the first TS time step staring later and later at intervals of six hours (the same as our cycling period).

For some reason we found this problem during our migration to WRF 3.5 but never before.  Is this "new"?

(We have some obs nudging problems that have also suddenly emerged where the same data set that is read in a concurrent 3.4.1 run is happily ingested while the 3.5 runs halts in an error saying that it's read in the maximum number of observations — which is set to a value that is far far greater than the number of obs in the OBSDOMAIN file.   We will submit when we have more information.)

Thanks Much
================================================================
Bill Capehart <William.Capehart at sdsmt.edu>   Associate Professor
Department of Atmospheric Sciences           AES PhD Coordinator
213 Mineral Industries Building             Skype/EVO: wcapehart
South Dakota School of Mines and Technology  Ph: +1-605-394-1994
501 East Saint Joseph Street                Fax: +1-605-394-6061
Rapid City, SD 57701-3995                Mobile: +1-605-484-5692
=================== http://capehart.sdsmt.edu ==================

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20130512/096bba3f/attachment.html