curl -o Anaconda-2023.07-2-Linux-x86_64.sh https://repo.anaconda.com/archive/Anaconda3-2023.07-2-Linux-x86_64.sh
chmod u+x Anaconda-2023.07-2-Linux-x86_64.sh
sudo /usr/local/anaconda3/bin/conda update conda
sudo yum install mlocate-0.26-30.el9.x86_64
sudo dnf install gcc
sudo dnf install gcc-fortran
sudo dnf install gcc-toolset-12-gcc-gfortran
curl -o netcdf4.tar.gz https://downloads.unidata.ucar.edu/netcdf-c/4.9.2/netcdf-c-4.9.2.tar.gz
tar -xvf netcdf4.tar
sudo yum install libxml2-dev
sudo yum isntall m4
sudo yum makecache --refresh
sudo yum install wget
https://support.hdfgroup.org/ftp/HDF5/releases/hdf5-1.14/hdf5-1.14.2/src/hdf5-1.14.2.tar.gz
gunzip hdf5-1.14.2.tar.gz
tar -xvf hdf5-1.14.2.tar
sudo yum provides g++
In the hdf5 source directory: 
        ./configure --enable-parellel --enable-fortran -enable-cxx --prefix=/usr/local/hdf5 
	make
        make check
	make install
sudo yum install libcurl-devel
sudo yum install unzip
In the netcdf source directory:
        ./configure --enable-remote-fortran-bootstrap CPPFLAGS="-I/usr/local/hdf5/include" LDFLAGS=-L/usr/local/hdf5/lib
        make check
        make install
sudo yum install evince
sudo yum install openmpi


----PRAMS compile failed due to lack of proper netcdf library modules

----Try rebuild of HDF5 and netcdf
----And this time I have openmpi installed

FC=/usr/lib64/openmpi/bin/mpif90 CC=/usr/lib64/openmpi/bin/mpicc ./configure --enable-parellel --enable-fortran -enable-cxx --prefix=/usr/local/hdf5


make
make test
---the test hangs on t_pmulti_dset but I find nothing on the webs to fix this.  Cross fingers and move on
sudo make install


--re-installing netcdf
--in the netcdf directory
HDF5DIR=/usr/local/hdf5 CC=mpicc CPPFLAGS=-I${H5DIR}/include LDFLAGS=-L${H5DIR}/lib ./configure --enable-remote-fortran-bootstrap --enable-parallel 
make test
sudo make install

-----------still failing on PRAMS compile.  netcdf.mod not found.  Appears to be issue with netcdf fortran install

------ go to netcdf fortran source and install...maybe bootstrap method with netcdf-c failed.

./configure
make
make test
sudo make install

------ netcdf.mod now exists!!!
------- retry compile of PRAMS

------- got past the netcdf.mod issue, but appears to have trouble finding the mpi libraries.
------- try serial compile and see if we can at least get that to work.

------- got farther with serial build, but now have a compile error.  Must be something new to the current gcc:
/home/rafkin/msr/PRAMS_2.9/common/src/base_model/surface_file.f90:73:61:

   57 |          call surface_nest(beg_grid, end_grid, igd, nzg+nzs, npatch, 'FILEDATA')
      |                                                             2
......
   73 |       call surface_nest(beg_grid, end_grid, nzg+nzs, npatch, 'FILEDATA')
      |                                                             1
Error: Type mismatch between actual argument at (1) and actual argument at (2) (CHARACTER(8)/INTEGER(4)).

----looks like it was an actual coding bug.  The 'igd' should not be there.  Removed and it compiles in serial.

----------------still having some issues with parallel build.  Made some changes to include and library paths for universal_lib build.  It now finds useable MPI fortran and C libraries, but produces a compiler error.  Will have to investigate.

--------- errors are due to type mismatches deep in the MPI interface calls.  Not sure if this was intentional, but I can get past it by reverting to fortran 2003 standards with the -std=f2003 compile option.  Let's run with this for now and remember to come back and see if there's anaything to be done about the mismatch.