[RegCNET] New NCEP files

[hanzenyu]

Hello all,

I suggest add attribution "save" to ivar in both ICBC.f and SST_1DEG.f, in order to to store "state" across invocations of the subroutines. But there may be no need to add this modification under some compiler.

      integer,save :: ivar 


Best Ragards,
HAN Zhenyu

2011-03-08 



韩振宇
HAN Zhenyu

Graduate University of the Chinese Academy of Sciences

LASG,Institute of Atmospheric Physics,
Chinese Academy of Sciences

Email: zhenyuhan at mail.iap.ac.cn
 



发件人： bixq 
发送时间： 2011-01-25  00:18:18 
收件人： regcnet 
抄送： GIORGI FILIPPO 
主题： Re: [RegCNET] New NCEP files 
 
Dear all:
Thanks to Steve Hostetler of OSU for him find out the problem of RegCM
related to NNRP1 data of year 2010 and his suggestion of solution to
the problem.
The following two versions of RegCM3 has been updated to fix the problem
related to 2010 NNRP1 data and/or monthly OISST probelms.
http://www.ictp.it/~pubregcm/RegCM3/regcm.tar.gz
http://www.ictp.it/~pubregcm/RegCM3/regcm_beta_20090916.tar.gz
Dr. Ha of Vietnam also mentioned the monthly OISST problem to me before,
the reason is that CDC/NOAA occasionally change the order of fields in
their netCDF monthly OISST files. This problem has been solved in the
above two version.
In RegCM4.0, there is also problem related to 2010 NNRP1 data, so version
4.0 should also be updated soon.
Best regards !
Xunqiang Bi
On Thu, 20 Jan 2011, Steve Hostetler wrote:
> In an attempt to save you from dealing with user panic/work:
>
> Beginning with 2010, NCEP (RI for now, RII is not updated yet) has started 
> writing some of their files (eg air.2010.nc  and the weekly Oi_sst) in a 
> different order than previous files (eg air.2009.nc).  This causes errors in 
> reading the offset, scale and values in the V3 SST_1DEG.f and V3/V4 ICBC.f 
> codes because they are hardwired to read a certain varid (5 in ICBC) 5. ( V4 
> mod_sst_1deg.f90 has been fixed for this issue.)  In both codes, calls to 
> nf_inq_varid() need to be made to get the correct varid in order to free them 
> from being hardwired.  The ICBC code needs nf_inq_varid() in several places. 
> In addition, V3/V4 SST_1DEG.f/mod_1deg.f90 need to have a trap to stop them 
> in case they attempt to read past the number of records in the OI_SST file. 
> I just use:
>
>       status=nf_inq_unlimdim(inet,recid)
>       status=nf_inq_dim(inet,recid,recname,nrecs)
>       if (it .gt. nrecs) stop '>>>> NUMBER OF SST TIME SLICES EXCEEDED'
>
> The netCDF error trap would work just as well.  As it is, if it does attempt 
> to make SSTs for dates that exceed the SST data, all temperatures are set to 
> the last good values.
>
> Steve
>
> __________________________
> Steve Hostetler
> US Geological Survey
> Department of Geosciences
> Wilkinson Hall Room 104
> Oregon State University
> Corvallis, OR 97331
> tel +1 541 737-8928
> fac +1 541 737-1200
>
>
_______________________________________________
RegCNET mailing list
RegCNET at lists.ictp.it
https://lists.ictp.it/mailman/listinfo/regcnet
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ictp.it/pipermail/regcnet/attachments/20110308/324d5458/attachment.html>