RyanChangeLog.txt

	NEW ROMS RYAN's CHANGES LOG FILE

This file log's changes that I make to the most recent ROMS
version. 

To keep updated on the latest ROMS version, in the base
directory use the command: 
	  svn update
To revert to the unchanged version of a file, use the command:
   	  svn revert -filename-
The original command to checkout the version was:
 svn checkout --username Ryan_Holmes https://www.myroms.org/svn/src/trunk ROMS/
 
File began: 23/4/14 by Ryan Holmes

----------------------------------------------------------------------

23/4/14 - Setup to compile on cluster;
	1 makefile_original saved (revision 734)
	2 makefile altered for cees-cluster run on PEQ_BMIX, saved as
	  makefile_cees-cluster
	3 Linux-gfortran - the lines containing: 
	
		ifeq "${FC_TEST}" ""
		$(SCRATCH_DIR)/ran_state.o: FFLAGS += -fno-strict-overflow
		endif
	  had been commented out in the OLD_ROMS version, may need to
		do it again? -> Yes I did!!
	4 Analytical files should be in the INC directory of the
		simulation run.
        5 Changed mod_scalars.F at line 1335 to use the higher K0
		parameters in KPP. This is surround by a cpp option to
		ONLY ACTIVATE IF PEQMIX IS ACTIVATED!

16/5/16 - This working copy seems to be too old to work with
          svn. Therefore I have downloaded the most recent version of
          ROMS retaining the make files and the Linux-gfortran file
          from above. I have not changed the mod_scalars.F
          file. Only needed to change nc-config to nf-config in
          Linux-gfortran.mk (nc-config masked by anaconda version on
          Path).
        - Upwelling test case successfully run with mpi, results in
          Testing/

17/8/16 - Redownloaded ROMS trunk as a git repository using the
          command
          git svn clone --prefix=origin/ --username Ryan_Holmes https://www.myroms.org/svn/src/trunk ./
        - Now should be able to make changes and track them using git
          (no longer an svn repository). Should be able to include
          remote updates using git svn rebase, git svn fetch
          etc. However, these are currently seg faulting leaving an
          error signal 11.
        - Changes from above included here (Linux-gfortran.mk and
          makefile), plus OLD_ROMS (contains NUMDIF) and
          OLD_ROMS_14_8_13 saved in separate folders.

September-October 2016:
        - Developed and implemented an atmospheric boundary layer
          model (cheapAML from MITgcm) inside ROMS. Contained in git
          repositories. The log for implementing this is in
          ABLM_log_Ryan.txt
        - Developed and implemented simple statistical coupled model
          (STATS_ENSO cpp option).
        
11-11-16:
        Wrote in option (activated via SET_WINDSPEED cpp option) to
        explicit set the wind magnitude Wmag in bulk_flux.F90. This
        option takes a wMag field from the forcing files and uses this
        for the wMag in the latent, sensible, and wind stress
        formulations instead of the wind magnitude calculated from the
        Uwind and Vwind inputs. This is done in order to maintain the
        correct fluxes when wind variability is removed.

17-02-17:
        When you update make sure you change the KPP K0 parameter
        again!!!!!!!!!!!!!!!!
        git svn fetch and git svn rebase do work, even if they seg
        fault and don't finish cleanly.