------------------------------------------- Notes for Pre-processing data for Nudging: ------------------------------------------- ./Grid-Reference : Contains refernce files for topography, vertical model levels, and SEgrid pentagons that specfy the grid structure of the generated data. ./Gen_Data_SETNAME_ne30 : Directory to be used as a template for processing data from 'SETNAME'. Contains the ncl processing scripts and 2 example csh scripts used to process YOTC and ERA-I data. Generally, it is best to create a separate directoy for each dataset/resolution processed. Processing programs/routines: ------------------------------ ESMF_regridding.ncl makeIC_Create_field_Master_List.ncl makeIC_extract_analyses_info.ncl MAKEIC.f90 MAKEIC.stub makeIC_se_001.ncl makeIC_se_002.ncl makeIC_se_procedures.ncl Work directories for processing: --------------------------------- Config/ LOG/ TMP/ Example Processing scripts: ---------------------------- Gen_ERAI_ne30.csh Gen_YOTC_ne120.csh ----------------------------- Script CONFIGURATION SECTION: ----------------------------- RUNNUM : Index so that processing can run with concurrent scripts running REF_DATE : 'YYYYMMDD' Year month and day to begin processing NUM_DAYS : Number of days to process NAMELIST : './Config/Config_makeIC-'$RUNNUM'.nl' <-- Work directories MYLOGDIR : './LOG/LOG_002.'$RUNNUM'/' <-- Work directories MYTMPDIR : './TMP/TMP_002.'$RUNNUM'/' <-- Work directories MYOUTDIR : '/PATH-FOR-OUTPUT-DATA/' <-- SET THIS PATH INPUTDIR : '/PATH-TO-REANALYSES-INPUT/' <-- SET THIS PATH ESMF_interp : <-- NOTHING TO CHANGE HERE! ESMF_pole : <-- NOTHING TO CHANGE HERE! ESMF_clean : <-- NOTHING TO CHANGE HERE! TMP_clean : <-- NOTHING TO CHANGE HERE! CASE : 'ROOT-NAME-OF-OUTPUTDATA' <-- SET THE NAME OF YOUR DATA FILES DYCORE : 'se' <-- LEAVE ALONE PRECISION : 'float' <-- LEAVE ALONE VORT_DIV_TO_UV : 'True' <-- CALC U,V from VOR/DIV VALUES? SST_MASK : 'False' <-- YOUR CHOICE?? ICE_MASK : 'False' <-- YOUR CHOICE?? OUTPUT_PHIS : 'True' <-- YOUR CHOICE?? REGRID_ALL : 'False' <-- PROCESS OTHER CONTENTS IN DATASET? ADJUST_STATE_FROM_TOPO : 'True' <-- YOUR CHOICE?? MASS_FIX : 'True' <-- YOUR CHOICE?? <-- SELECT EITHER THE ne30 of ne120 files fname_grid_info : '../Grid-Reference/cami_0000-01-01_ne30np4_L30_c120315.nc' fname_phis_output : '../Grid-Reference/USGS-gtopo30_ne30np4_16xdel2-PFC-consistentSGH.nc' #fname_grid_info : '../Grid-Reference/cami-mam3_0000-01-ne120np4_L30_c110928.nc' #fname_phis_output : '../Grid-Reference/USGS-gtopo30_ne120np4_16xdel2-PFC-consistentSGH.nc' ftype_phis_output : 'SE_TOPOGRAPHY' <-- LEAVE ALONE fname : ( none none none none none none ) <-- NOTHING TO CHANGE HERE! ftype : ( none none none none none none ) <-- NOTHING TO CHANGE HERE! ftime : ( none none none none none none ) <-- NOTHING TO CHANGE HERE! hoursec : ( 00000 21600 43200 64800 ) <-- NOTHING TO CHANGE HERE! hourstr : ( 00 06 12 18 ) <-- NOTHING TO CHANGE HERE! fname[1] : "YYYYMM/ei.oper.an.ml.regn128sc.YYYYMMDDHH" <-- TEMPLATE FOR REANALYSES FILES ftype[1] : "Era_Interim_627.0_sc" <-- TYPE OF REANALYSES DATA ftime[1] : "1X" <-- # OF RECORDS IN FILE fields : ( U V T PS Q ) <-- FIELDS TO PROCESS field_findex : ( 1 1 1 1 1 ) <-- FILE CONTAINING THE FILED phis_findex : 1 <-- FILE CONTAINING 'PHIS' WHAT TO DO: ----------- (1) Set REF_DATE to the begining date of reanalyses data, and NUM_DAYS, the number of days of data to process. (2) Processing the data is SLOW, SLOW, SLOW, particularly at ne120. To get done in a timely mannor, multiple copies of the script need to be run concurrently. RUNNUM is an index so that concurrent runs don't interfere with eachother. Also, for example, the format for the YOTC data changes during the 2 year length, which requires separate processing scripts. (3) Set the path to the reanalyses source data and the path where processed output should be stored. (4) Set the CASE name, the root anme of your processed files. (5) Set { VORT_DIV_TO_UV, SST_MASK, ICE_MASK, OUTPUT_PHIS, } as you desire. { REGRID_ALL, ADJUST_STATE_FROM_TOPO, MASS_FIX } (6) Slelect the grid/phis values for the desired resolution (7) Determine the reanalyses files that contain the data you desire. Then set fname[N],ftype[N],ftime[N] values for each of the N files you need. (For ERA-I all of the values are contined in a single file, for YOTC the data is spread across 3 different files.) fname[n] - Is a template in which YYYY,MM,DD,HH will be replaced with the values at a given time. ftype[n] - Identifies the type of reanalyses file - see 'makeIC_extract_analyses_info.ncl' ftime[n] - "4X" - for daily files containing 4 time recoeds "1X" - for 1 record per file. (8) Add the data to be processed into fields. In 'field_findex' set the file index N that contains that data. (9) Set the index for the file which contains PHIS. (10) The PROCESSING Section may need some minor mods depending on the structure of the file names. Compare Gen_ERAI_ne30.csh to Gen_YOTC_ne120.csh to see the mod that was needed when the time of day is specified in Hours instead of Seconds. (11) **IMPORTANT** - Conservative mapping with ESMF is not possible using the descriptor files it generates for the SE grid. To make the 'conserve' option work, the ne30 and ne120 pentagons files were specified directly by hand. So the file 'makeIC_se_002.ncl' needs to be edited to use the correct files for the desired resolution. Search fo 'ne30' to get to the line that needs to be changed. Fixing this work around is on the TODO list... near the bottom. - Different versions of this ncl program can be created, one just needs to change the name of the NCL script onvoked by the Gen_XXX.csh script (near the bottom).