Problem with geogrid

All other topics about postprocessing model data (GrADS and other software), about other numerical weather prediction software (including WRF-NMM and WRF-ARW discussion unrelated to UEMS/WRF EMS), and general meteorology talk go in this forum.
adem
Posts: 8
Joined: Wed Jun 06, 2012 7:55 am

Re: Problem with geogrid

Post by adem » Thu Jun 07, 2012 9:01 am

I can give the file only via mail if you send me your adress email by private message...
I understand if you don't want...

meteoadriatic
Posts: 1566
Joined: Wed Aug 19, 2009 10:05 am

Re: Problem with geogrid

Post by meteoadriatic » Thu Jun 07, 2012 9:24 am

Got it. Yes it is almost empty. File size is also much smaller than it should be.

It smells like compiler issue to me. Which compiler did you used?

adem
Posts: 8
Joined: Wed Jun 06, 2012 7:55 am

Re: Problem with geogrid

Post by adem » Thu Jun 07, 2012 9:27 am

I use the version 4.6 of gcc and gfortran.
this is configure.wps :

Code: Select all

# configure.wps
#
# This file was automatically generated by the configure script in the
# top level directory. You may make changes to the settings in this
# file but be aware they will be overwritten each time you run configure.
# Ordinarily, it is necessary to run configure once, when the code is
# first installed.
#
# To permanently change options, change the settings for your platform
# in the file arch/configure.defaults, the preamble, and the postamble -
# then rerun configure.
#

.SUFFIXES: .F .f .c .o

SHELL           	=       /bin/sh

NCARG_LIBS		=	-L/home/ambre/NCL/ncl_ncarg-6.0.0/lib -lncarg -lncarg_gks -lncarg_c \
				-L/usr/lib/x86_64-linux-gnu/ -lX11 -lf2c

NCARG_LIBS2		=	# May be overridden by architecture specific value below

FDEFS			=	-DUSE_JPEG2000 -DUSE_PNG

# Listing of options that are usually independent of machine type.
# When necessary, these are over-ridden by each architecture.

ARFLAGS			=	

PERL			=	perl

RANLIB			=	echo

WRF_DIR			=	../WRFV3

WRF_INCLUDE     =       -I$(WRF_DIR)/external/io_netcdf \
                        -I$(WRF_DIR)/external/io_grib_share \
                        -I$(WRF_DIR)/external/io_grib1 \
                        -I$(WRF_DIR)/external/io_int \
                        -I$(WRF_DIR)/inc

WRF_LIB         =       -L$(WRF_DIR)/external/io_grib1 -lio_grib1 \
                        -L$(WRF_DIR)/external/io_grib_share -lio_grib_share \
                        -L$(WRF_DIR)/external/io_int -lwrfio_int \
                        -L$(WRF_DIR)/external/io_netcdf -lwrfio_nf \
                        -L$(NETCDF)/lib  -lnetcdf \
			-L/usr/local/hdf5/lib -lhdf5

#### Architecture specific settings ####

COMPRESSION_LIBS	=  # intentionally left blank, fill in COMPRESSION_LIBS below

COMPRESSION_INC		=  # intentionally left blank, fill in COMPRESSION_INC below

#
#   Settings for Linux x86_64 g95 compiler     (serial) 
#
#
COMPRESSION_LIBS    = -L/usr/local/lib -ljasper -lpng -lz
COMPRESSION_INC     = -I/usr/local/include/libpng12/include -I/usr/local/include/jasper \
			-I/usr/local/include
FDEFS               = -DUSE_JPEG2000 -DUSE_PNG
SFC                 = gfortran
SCC                 = gcc
DM_FC               = 
DM_CC               = 
FC                  = $(SFC)
CC                  = $(SCC)
LD                  = $(FC)
FFLAGS              = -ffree-form -O -fconvert=big-endian -frecord-marker=4
F77FLAGS            = -ffixed-form -O -fconvert=big-endian -frecord-marker=4
FCSUFFIX            = 
FNGFLAGS            = $(FFLAGS)
LDFLAGS             = 
CFLAGS              = -w
CPP                 = /usr/bin/cpp -C -P -traditional
CPPFLAGS            = -D_UNDERSCORE -DBYTESWAP -DLINUX -DIO_NETCDF -DBIT32 -DNO_SIGNAL
ARFLAGS             =
RANLIB		    = ranlib

########################################################################################################################
#
#	Macros, these should be generic for all machines

LN		=	ln -sf
MAKE		=	make -i -r
RM		= 	/bin/rm -f
CP		= 	/bin/cp
AR		=	ar ru

.IGNORE:
.SUFFIXES: .c .f .F .o

#	There is probably no reason to modify these rules

.c.o:
	$(RM) $@
	$(CC) $(CPPFLAGS) $(CFLAGS) -c $<	

.f.o:
	$(RM) $@ $*.mod
	$(FC) $(F77FLAGS) -c $< $(WRF_INCLUDE)

.F.o:
	$(RM) $@ $*.mod
	$(CPP) $(CPPFLAGS) $(FDEFS) $(WRF_INCLUDE) $< > $*.f90
	$(FC) $(FFLAGS) -c $*.f90 $(WRF_INCLUDE)
#	$(RM) $*.f90

meteoadriatic
Posts: 1566
Joined: Wed Aug 19, 2009 10:05 am

Re: Problem with geogrid

Post by meteoadriatic » Thu Jun 07, 2012 9:40 am

And you compiled WRF before, with same compiler? And netcdf before WRF, again with same compiler?

I know, this probably won't be advice that you want to hear, but if I'm you, I would try with Intel. I have no positive experience with gfortran and whenever I tried to use it I had problems. Many of people said it is straightforward to use gfortran for WRF, but I have different experience. You will also get much better performance, especially if you have Intel CPU. Intel is free for non-commercial use.

If you want to try this approach, look here:

http://software.intel.com/en-us/article ... -download/
(click on Intel® Parallel Studio XE 2011 for Linux*)

adem
Posts: 8
Joined: Wed Jun 06, 2012 7:55 am

Re: Problem with geogrid

Post by adem » Thu Jun 07, 2012 10:16 am

And you compiled WRF before, with same compiler? And netcdf before WRF, again with same compiler?
I compiled netcdf, then ncl, wrf and wps...
May be you are rigth it's the compiler... But Some people use gfortran and it compiles good...
I try ifort.
Thanks

meteoadriatic
Posts: 1566
Joined: Wed Aug 19, 2009 10:05 am

Re: Problem with geogrid

Post by meteoadriatic » Thu Jun 07, 2012 10:18 am

I know, but with Intel I never had issues. And I can help in every step; with gfortran I'm bit out of knowledge.

Post Reply