Custom Query (210 matches)

Hi,

I would greatly appreciate your help with the following runtime error:

I have run WRF V3.3 and, in attempting to run FlexWRF using the WRF output, I get the following fatal error:

FLEXPART WARNING: NO WIND FIELDS ARE AVAILABLE.

I have checked my WRF output variable listing and compared it against the required variables for FlexWRF (as per the Brioude et al , 2013 paper in Geosci. Model. Dev.). I can also view the WRF output variables, so the data have been generated. I have even tried running with a larger nest, but get the same error.

Previously, I ran FlexWRF successfully run for the testcase (i.e. with Flexwrf.input.forward1 and the downloaded test WRF output files), so I am certain that it is not a compiler issue.

I have attached my input file, log file and output header file.

Thank you in anticipation for your kind assistance.

PeterD

Hi FPers,

I would like to thank all the develops of FLEXWRFv3.1 for your hard work. I run it perfectly with OUTGRID_COORD set to 0, which means the output will adopt the grids in wrf fields. However, when I try to set OUTGRID_COORD to 1 and expect to get a lat/lon grid output, I was confused about settings in the OUTGRID part and couldn't run the package. Actually, the descriptions of several parameters in this section are confusing to me. I hope someone can share with me a working input file with OUTGRID_COORD set to 1
, and explain a little bit more about how to set the following parameters right.

OUTLONLEFT OUTLATLOWER (these two were actually need to be set as meter coordinate when OUTGRID_COORD was set to 0 and wrf files in meter coordinate. The description says geographical lon/lat) NUMXGRD NUMYGRID (As I found out for my case, when OUTGRID_COORD was set to 0, these two were grids set within the wrf domain.) OUTGRIDDEF DXOUTLON DYOUTLON (I think this is a typo, should be DYOUTLAT.)

I tried my old settings in 8.22, OUTLONLEFT as -179, OUTLATLOWER as 0 , NUMXGRID as 360 and NUMYGRID as 90 for a full North Hemisphere run. The settings do not work here. The nc output feature is very nice, but the description for these settings may need to be specified more according to the griding method. I am now testing by guessing how these parameters are taken and digging into the fortran codes. Hope someone can help me out.

Many thanks, Bo Zhang Michigan Technological University

Hi!

I am trying to compile FLEXPART-WRF v3.1 on a linux x86_64 machine. I have successfully compiled it in serial mode, however I am having difficulty running the program once it is compiled with openMP. I compile as follows:

$ OMP_NUM_THREADS=2 make -f makefile.mom omp

and the compilation runs fine. However, when I run the program, it almost immediately throws a seg fault. I have a feeling this has to do with how openmp is setting the stack size on compilation. I tried adding the flag: "-fmax-stack-var-size=n" for various numbers of bit, n, and this compiles fine, (it does give plenty of warnings that this overrides -frecursive); however, then I run into issues with the variable maxomp in par_mod.f90 .

How can I get the compiler to set the appropriate stack size?