First commit

Developement.doc

@@ -0,0 +1,42 @@
+                Development notes for Wgrib2      1/2004
+
+
+
+Every option for wgrib is now a function call.  (Eventually some
+options will be macros; for example, -s = -var -d etc.)  I went
+to this approach for
+
+  (1) ease of development
+  (2) ease of extendability
+  (3) more versatile
+
+People kept making mods for wgrib, so I wanted a system where you
+just drop in the code and compile.  
+
+
+Steps to add a new option/function:
+
+  (1) decide on the type of function
+      currently inv (inventory) - writes to stdout, no access to decoded data
+                output - writes to stdout, dump-file, access to decoded data
+		setup - no access to decoded data
+
+  (2) write function
+	functions are called several times
+	a) first with mode == -1 (initialization stop)
+	b) for each data field (mode = 0 or mode = 1 verbose)
+	c) finishing with mode = -2 (finializing step)
+
+  (3) write header in the source code
+	format is
+		* HEADER:sort_order:name:type:number_args:description
+	ex      * HEADER:100:d:inv:0:date YYYYMMDDHH
+
+     For the name of the function and calling args, see
+	documentation in wgrib2
+
+To compile
+
+	./function.sh		updates the option table
+	./compile		compiles the code
+

INSTALLING

@@ -0,0 +1,172 @@
+1/2013                            General Installation 
+
+                       Quick Directions for the  Default Configuration
+                       "Any color that you like as long as it's black"
+
+Download source code from
+
+       ftp://ftp.cpc.ncep.noaa.gov/wd51we/wgrib2/wgrib2.tgz
+       http://www.ftp.cpc.ncep.noaa.gov/wd51we/wgrib2/wgrib2.tgz
+
+Untar the compressed tarball
+
+       gunzip wgrib2.tgz
+       tar -xvf wgrib2.tar
+
+       or if you have gnu tar:  tar -xzvf wgrib2.tgz
+
+Go to the main directory (grib2)
+
+       cd grib2
+
+Define the C compiler that you will be using.  Note icc and pgc have been
+found to be incompatible with the Jasper library and are not allowed.
+
+       sh/bash:    export CC=gcc
+       csh/tcsh:   setenv CC gcc
+
+Compile the wgrib2 using gnu make
+
+       linux:      make
+       MacOS:      gmake
+       Unix:       ???    ask system administrators
+
+Try executing wgrib2
+
+      wgrib2/wgrib2 -config
+
+Copy wgrib2 to SOMEPLACE
+
+      cygwin/linux/unis: cp wgrib2/wgrib2 SOMEPLACE
+
+
+                            Not So Quick Directions
+                            So you want some options
+
+
+Edit the file "makefile" and select the desired options:
+
+# netcdf3: write netcdf files with netcdf-3 library
+# netcdf4: write netcdf files with netcdf-4 library
+#  only one netcdf library can be installed. do not choose netcdf3 and ncdf4
+# regex: regular expression package used by (match,not), POSIX-2
+# tigge: enable -tigge option for tigge names
+# mysql: write to mysql files
+# ipolates: fortran interpolation library
+# udf: user defined functions
+# openmp: use OpemMP to thread the code
+# g2clib: install g2clib decoders (allow -g2clib 2)
+#
+# the flags are stored in wgrib2/config.h
+#
+
+Default options
+
+USE_NETCDF3=1
+USE_NETCDF4=0
+USE_REGEX=1
+USE_TIGGE=1
+USE_MYSQL=0
+USE_IPOLATES=0
+USE_UDF=0
+USE_OPENMP=0
+USE_G2CLIB=0
+
+In the default configuration, all the code is compiled
+using the C compiler.  
+
+
+On some machines, you will have to define the C compiler
+and C preprocessor.  On the NCEP's AIX, I have to define
+them by
+
+	export CC=/usr/vacpp/bin/xlc_r
+	export CPP=/usr/bin/cpp
+
+Then you have to run gnu make.  (gmake on some machines, make on others)
+
+	make		- on linux
+	gmake		- on AIX
+
+
+
+                          Comments on Netcdf
+
+The makefile is set up to use either netcdf v3 or netcdf v4.  Netcdf-4
+makes wgrib2 much bigger, so don't install it unless you really have to.
+The source code for netcdf-4 (and necessary hdf5) is also very big
+and will have to be downloaded separately.
+
+The netcdf-3 library is included with this distribution.  The library is
+compiled in enable-C-only mode which is not be that useful for
+fortran code.  Don't use try to link other codes to this library.
+
+The netcdf-4 library has to be downloaded with the HDF5 library and
+the shell variable FC has to be set to the default fortran compiler.
+
+
+                  IPOLATES, interpolation library
+
+IPOLATES is the interpolation library used by the -new_grid options.
+In order to use this library, you must have a f90/f95/f2003 fortran compiler.
+In order to enable the -new_grid options, the makefile has to have USE_IPOLATES 
+enabled and the shell variable, FC, must have the name of the fortran compiler. 
+You will have to change the source code if you are not using g95, gfortran, 
+openf95 or xlf.
+
+                              MYSQL
+
+Install MySQL requires the appropriate headers and libraries to be installed.
+The makefile may have to be modified to reflect the locations of the headers
+and libraries.
+
+
+
+                     INSTALLING on windows XP
+
+The code including the IPOLATES option compiles in the cygwin development 
+system.  You'll have to install extra packages such as the C, gfortran 
+compilers and various scripting utilities. 
+
+    Before you compile, you set the default compilers.
+
+    export CC=gcc
+    export FC=gfortran
+    make
+
+                        INSTALLING on FreeBSD
+
+Set appropriate CC and FC environment variables and try make. wgrib2 v1.9.8
+was compiled with clang and gfortran.
+
+
+                        INSTALLING on Apple OS-X
+
+Has been done. No details.
+
+
+                        INSTALLING on NCEP's AIX
+
+Define default C, C preprocessor and fortran
+
+   export CC=/usr/vacpp/bin/xlc_r
+   export CPP=/usr/bin/cpp
+   export FC=xlf_r
+   gmake
+
+
+			Static vs Dynamic Libraries
+
+The various libraries can be created in both dynamic and static versions.  
+I've tried to make the add-on libraries static.  
+
+                           Testing Package
+
+Have a testing script that tests an known-good version with a test
+version. For new machines, I use the gcc compiled version as the 
+"known-good" version.  Not ideal.
+
+                             Problems
+
+Some optional packages require changes to makefile
+

README

@@ -0,0 +1,60 @@
+                                 wgrib2
+
+wgrib2 is a program to read/write grib2 files.
+
+
+For documentation, see
+
+    http://www.cpc.ncep.noaa.gov/products/wesley/wgrib2/index.html
+
+
+For documentation of commonly used program options, type
+
+    wgrib2
+
+
+For documentation of all program options, type
+
+   wgrib2 -help all
+
+For documenation of any program option that is poorly described on-line
+or by the "wgrib2 -help all", check the source code by searching
+for "f_{option}" where the option does not include the minus sign.
+
+
+Default makefile options
+
+      USE_NETCDF4=0
+      USE_NETCDF3=1
+      USE_REGEX=1
+      USE_TIGGE=1
+      USE_MYSQL=0
+      USE_IPOLATES=0
+      USE_UDF=1
+
+To compile without netcdf, use USE_NETCDF3=0 and USE_NETCDF4=0
+To compile without netcdf v3, set USE_NETCDF3=1 and USE_NETCDF4=0
+To compile without netcdf v4, set USE_NETCDF3=0 and USE_NETCDF4=1
+ For netcdf4, the netcdf4 and hdf5 libraries are usually not included 
+  in the wgrib2.tgz file to save space.
+
+If you have an old C compiler system that doesn't include <regex.h>, then
+you have to set USE_REGEX=0.  This turns off all the regular expressions
+which would be a pain.
+
+No harm to keep TIGGE support on.  USE_TIGGE=1
+
+Support for Mysql is an option.  You have to modify the makefile to indicate 
+the locations of the mysql includes and libraries and set USE_MYSQL=1
+
+The option -new_grid uses the ipolates library to do the interpolation.
+The required libraries are written in fortran and their is no support
+for compiling these libraries.  The source code and makefile will
+have to be modified to use the ipolates option.  Consult a local
+expert if you want to install this optional package.  No help is available 
+from NCEP for installing the package.
+
+User Defined Functions (UDF) allow you to run shell commands from
+within wgrib2. UDF are not available on windows machines unless
+<unistd.h> is implimented by the compiler (cygwin, MinGW).  For
+the security paranoid, turn off UDF by USE_UDF=0.  

README.EUMETSAT

@@ -0,0 +1,87 @@
+			Using EUMETSAT grib2 fields
+
+
+Some of the EUMETSAT data is on a space view perspective grid.
+
+    -sh-3.00$ wgrib2 LMPEF_CLM_110328_0945.grib2 -grid
+    1:0:grid_template=90:winds(N/S):
+          Space view perspective or orographic grid (3712 x 3712) units 1e-06 
+	            input EW:SN output WE:SN res 0
+          sub-sat point: lat 0.000000 lon 0.000000 ix=1856.000000 iy=1856.000000
+          diameter of earth dx=3622 dy=3610 grid cells ori_angle 0.000000
+          sat. altitude=6.610700 (equatorial radii) grid_origin Xo=0 Yo=0
+
+wgrib2 v1.9.3 adds support for space view perspective grids with the restriction
+that orientation angle (ori_angle) is 0 and sub-sat point has a latitude of 0.
+
+The rest of the text assumes that you have a space view perspective file.
+
+
+			Displaying with GrADS
+
+My version of GrADS is not using the latest g2clib can couldn't handle product
+definition template 4.30.  So here is what I did.
+
+    * make a copy of the grib file product with def template 4.0
+    wgrib2 LMPEF_CLM_110328_0945.grib2 -set_pdt 0 -set_var CLOUDM -grib new.grb
+    1:0:d=2011032809:CLOUDM:no_level:-1 missing fcst:
+
+    * make a control file
+    g2ctl -raw new.grb >new.ctl
+
+    * make a index file
+    gribmap -i new.ctl
+
+Note that my grib file only had one field.  Some EUMETSAT grib files have
+mutiple fields on different grids.  g2ctl will get confused files that have
+more than one grid.  Your grib file must only have one grid.
+
+When you run GrADS on your file, you will see the raw satellite data.
+
+
+			Getting grid point information
+
+
+To get the value of the grid point nearest to a specified lat-lon.
+
+	-sh-3.00$ wgrib2 new.grb -lon 10 30
+	experimental space_view2ll scan=12
+	experimental space_view2ij
+	1:0:lon=10.009994,lat=30.005756,val=1
+
+To get the value of a grid point
+
+	-sh-3.00$ wgrib2 new.grb -ijlat 1 1
+	experimental space_view2ll scan=12
+	1:0:(1,1),lon=999.000000,lat=999.000000,val=3
+	-sh-3.00$ wgrib2 new.grb -ijlat 1000 1000
+	experimental space_view2ll scan=12
+	1:0:(1000,1000),lon=-27.386283,lat=-24.929402,val=0
+
+If the grid point is off the earth, a lat-lon of 999 is used.
+The convention of (i,j) is different that used by EUMETSAT.  (1,1) is 
+the bottom left corner which is consistent with wgrib2 convention.
+
+
+			Regridding the data
+
+
+To space view perspective is not as convenient as a latitude-longitude grid
+for many plotting programs.  You can use the -lola option to regrid the
+data to a lola grid.
+
+   * converting to lola grid by nearest neighbor
+    -sh-3.00$ wgrib2 new.grb -lola -60:1201:.1 -60:1201:.1 lola.grb grib
+    experimental space_view2ll scan=12
+    experimental space_view2ij
+    1:0:d=2011032809:CLOUDM:no_level:-1 missing fcst:
+
+    * the grid of lola.grb
+    -sh-3.00$ wgrib2 -V lola.grb
+    1:0:vt=2011032809:no_level:-1 missing fcst:CLOUDM Cloud Mask [-]:
+        ndata=1442401:undef=0:mean=1.53046:min=0:max=3
+        grid_template=0:winds(N/S):
+            lat-lon grid:(1201 x 1201) units 1e-06 input WE:SN output WE:SN res 48
+            lat -60.000000 to 60.000000 by 0.100000
+            lon 300.000000 to 60.000000 by 0.100000 #points=1442401
+

README.Mac

@@ -0,0 +1,25 @@
+A user had trouble making a 64 bit version of wgrib2 with the Leopard OS.
+Some of the libraries where 32 and other 64 bits.  To get everything
+to work, he had to get the "-m64" flag to all the compiles.  He
+did this by setting the following environment variables before
+doing the "gmake -f makefile"
+
+export CC=/usr/bin/gcc
+export CXX=/usr/bin/c++
+export F77=ifort
+export CFLAGS="-O2 -m64"
+export CXXFLAGS="-O2 -m64"
+export FFLAGS="-O2 -m64"
+
+Another user had to the above and change makefile
+
+      wCPPFLAGS=-O2
+
+    to
+
+      wCPPFLAGS=-O2 -m64 
+
+-------------------------------------------------
+
+Another user Mac user had 32-bit MySQL libraries and had to
+remove the -m64  (2/2010)

README.clang

@@ -0,0 +1,8 @@
+Compiling with clang (C language frontend for LLVM)
+
+tested with a Ubuntu 12.04 system
+
+export CC=clang
+export FC=gfortran -fplugin=dragonegg
+
+note: at time of testing, clang did not have OpenMP support