wiki:access/UtilsWG/UtilWgDocs/smocs_um_0.0.9

Version 2 (modified by Wenming Lu, 2 years ago) (diff)

--

USER Guide: SMOCS on NCI raijin

  1. Load the SMOCS
    module load smocs    #on raijin; current default ver=0.0.9
    module list          #to see what have been loaded
    2) python/2.7.6                 3) python/2.7.6-matplotlib      4) cdat-lite/6.0rc2-fcasttime   5) smocs/0.0.9
    
  1. Sample Data and Output
    Sample data:
    /short/dp9/aps_test_data/util_mosrs/smocs/fcst
    There are 42 UM pp files in this folder from APS2 ACCESS-C VT domain. For making the testing short, all pg files are not copied
    and only part of the forecast hours are copied.
    
    Standard Output:
    /short/dp9/aps_test_data/util_mosrs/smocs/output
    This folder contains all output json files for your reference. It also contains a smocs.ksh which will be explained in section3.
    
  1. Method 1

Before starting running smocs, we define few variables which are used in commands,

export MODELDIR=/short/dp9/aps_test_data/util_mosrs/smocs/fcst/
export PREFIX=xbnmya
export MODEL=aps2_vt
export YYYYMMDDZZ=2014050500

To get the smocs statistics,

UM_DecumDir.py ${MODEL} ${PREFIX} ${MODELDIR}   
  this step reads all pa files and produces hourly precipitation and evaporation from accumulated fields
  The output is aps2_vt_hprc_2014050500.json

UM_ScanDir.py ${MODELDIR} ${PREFIX} -xj ${MODEL}_hprc_${YYYYMMDDZZ}.json  -nobin -noperc
  this step read all files and produces statistics for all fields 
  if -xj ${MODEL}_hprc_${YYYYMMDDZZ}.json is specified, the step will combine the data from the previous step together
  The output is xbnmya_2014050500.json
  Note: This command uses huge computing resources. It will get killed before producing s result. You need to refer to 
  smocs.ksh to qsub this task to PBS and the output is smtstTS.e6654242.
  1. Method 2
      produce a json file for each file,
    FILES=`ls $MODELDIR`
    for F in $FILES
    do
      echo $F
      UM_ScanFile.py ${MODEL}  $F -y ${YYYYMMDDZZ} -nobin -noperc
    done
      the output aps2_vt_2014050500_xbnmya_pa002.json etc. are in standard output location for your reference.
    
    UM_DecumDir.py ${MODEL}  ${PREFIX} ${MODELDIR}
      this step is as same as the first step in section3.
    
    UM_ScanConcat.py ${MODEL} ${YYYYMMDDZZ} -x ${MODEL}_hprc_${YYYYMMDDZZ}.json
      this step combine all json files produced in step 1$2 and produces aps2_vt_2014050500.json.
      You may skip -x ${MODEL}_hprc_${YYYYMMDDZZ}.json not to include statistics of precipitation/evaporation fields.
    
  1. View json File
    bodoView.py aps2_vt_2014050500.json