<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
</head>
<body bgcolor="#FFFFFF" text="#000000">
Dear, <br>
<br>
My name is Angel Vázquez. I am working at Universidad de Cuenca,
Ecuador. I am a Computer Engineer, but I am currently working in an
environmental sciences research group. <br>
We are trying, among other things, to build the capacities to
perform downscaling of GCM´s results for Ecuador. Among those
capacities, we are exploring the use of WRF as a Regional Climate
Model.<br>
We are having some trouble with running WRF using ECHAM5 data.<br>
<meta http-equiv="CONTENT-TYPE" content="text/html; charset=UTF-8">
<br>
We have tried to run WRF by using ECHAM5 (considering data every 6
hours). I have downloaded all the variables from January 1, 2095,
and I have gathered them in a single file (by using the CDO library,
version 1.5.5). Then, we have generated a file for each time-step.
We have not modified any information within the downloaded GRIB
files so far. Once we have the files ready, we have created a Vtable
based on the Vtable from ECMWF (which I am attaching). When we run
geogrid.exe no problem is observed. Then, when running ungrib.exe
there is no running time message, but we can detect that information
from variable PSFC (surface pressure) has not been extracted. This
can be seen in the corresponding log file, which I am attaching.
While running metgrid.exe, we obtain the following warning and error
messages:<br>
<br>
WARNING: Fields PRES has missing values at level 200100 at (i,
j)=(1, 1)<br>
WARNING: Fields PSFC has missing values at level 200100 at (i,
j)=(1, 1)<br>
ERROR Missing values encountered in interpolated fields. Stopping<br>
application called MPI_Abort(MPI_COMM_WORLD, 52) – process 0<br>
<br>
Also, when editing the Vtable so that the PSFC won´t be extracted
from GRIB files, the execution of geogrid.exe, ungrib.exe, and
metgrid.exe does not stop. However, when running real.exe, it stops
without giving<br>
much information. The corresponding log file is also attached I have
also recompiled WRF and verified that with data download from GFS,
it works fine.<br>
<br>
Is it necessary to transform hybrid sigma-pressure coordinates
system into pressure levels?<br>
<br>
The WRF and WPS version we are using is 3.3.3. We use gcc and
gfortran compilers.<br>
<br>
<div class="moz-signature">Best regards,<br>
<br>
Angel Oswaldo Vázquez Patiño<br>
Researcher<br>
<br>
<a href="http://diuc.ucuenca.edu.ec/cta">Grupo de Ciencias de la
Tierra y del Ambiente</a><br>
Universidad de Cuenca<br>
Av. Víctor Manuel Albornoz s/n. Campus Quinta Balzaín<br>
<br>
Telf: (593) (7) 4051000 Ext. 4415</div>
</body>
</html>