[gradsusr] .he5 format descriptor file

[James E. Johnson]

On Monday, October 03, 2011 05:20:30 pm SMITH, JONATHAN WYNN. (GSFC-6140) 
wrote:
> Hello:
>  
> I was attempting to set up a descriptor file for following he5 file:
>  
> OMI-Aura_L2-OMDOMINO_2006m0522t0926-o09846_v003-2010m1007t075325.he5
> ___________________________________________________________________________
> ___ Here is the descriptor file I’ve started:
>  
> dset
> /Users/jonathanwsmith/DATA/OMI_DOMINO_NO2/FIRST_LEG/OMI-Aura_L2-OMDOMINO_%
> y4m%m2%d2t%h2%n2-o10074_v003-%y4m%m2%d2t%h2%n256.he5
> 
> chsub
> 
> options template
> 
> title OMIDOMINO
> 
> undef -9999
> 
> dtype hdf5_grid
> 
> xdef 360 linear -180 1
> 
> ydef 180 linear -90 1
> 
> zdef 1 linear 1 1
> 
> tdef 3 LINEAR 07JUN2006 1dy
> 
> vars 2
> Tropospheric~Vertical~Column=>tvono2 0 y,x Tropospheric_Vertical_Column
> 
> Averaging~Kernal=>avgk 0 y,x Averaging_Kernal
> 
> endvars
>  
> From this I am wondering what dset template would represent the 10074. 
> There is also no template for seconds ( the 2 numbers before the .he5) 
> Do I need the chsub?
>  
> Will grads work with .he5 files?
>  
> Thanks
> Jonathan Wynn Smith
> Student Trainee
> jonathan.smith at nasa.gov
> NASA-Goddard Space Flight Center
> Code 613.3-Atmospheric Chemistry and Dynamics Branch
> Greenbelt, MD 20771
> 301-614-5956
>  
> Doctoral Candidate
> jonathan.smith at bison.howard.edu
> Howard University Program in Atmospheric Sciences
> 2355 6th St NW/408 Thirkield Bldg.
> Washington, DC 20059

I don't think it will be easy or make much sense for you to try and read an 
OMI level-2 orbital/swath data file with GrADS. GrADS is designed for Level-3 
gridded (hence the Gr in GrADS) data files. You can try, but the file doesn't 
have regular Latitude and Longitude dimensions that GrADS expects. Instead it 
has along-track (nTimes) and across-track (nXTrack) dimensions. You would have 
to 'fake' GrADS into thinking they were Latitude and Longitude dimensions, and 
then be careful what you do since they are not Latitude and Longitude. You 
might be able to treat each pixel point as a "station" and create a station 
descriptor to grid the data, but even this is complicated. Otherwise you would 
have to grid the data with some other method to do anything useful in GrADS.

For your info: the first date string in the file name is the begin date/time 
for the file (orbit). The oXXXXX before the v003 is the orbit number for this 
file. The second date string after the v003 is the production date time (when 
the file was created). The next file will be one orbit later (approx. 98 
minutes after this one).

The simplest descriptor file to read your file is:

DSET OMI-Aura_L2-OMDOMINO_2006m0522t0926-o09846_v003-2010m1007t075325.he5
DTYPE hdf5_grid
TITLE OMI DOMINO
UNDEF -1.26765060e+30 _FillValue
XDEF    60 linear 1 1
YDEF  1644 linear 1 1
ZDEF    34 linear 1 1
TDEF     1 linear 09:26z22may2006 1hr
VARS 2
/HDFEOS/SWATHS/DominoNO2/Data~Fields/TroposphericVerticalColumn=>no2trop    0     
y,x    NO2 troposheric vertical column density
/HDFEOS/SWATHS/DominoNO2/Data~Fields/AveragingKernel=>avgkern               0   
z,y,x    Height dependent averaging kernel
ENDVARS

Then you could examine the 300th row of data using:

open my.ctl
set x 0 61
set y 300
d no2trop
(should make a XY plot of values, and ignore the fact that the X-axis isn't 
really longitude, but the index values)


If you want to set GrADS to use a template, I think it would be easiest to do 
this using the orbit number it increments by 1, the begin dates are only 
approximate at 98 minutes, and the product date time can't be used since they 
can be anything. I would first create a symbolic link to file using a simple 
shell script to do this:

for file in *.he5
do
  ln -s $file `echo $file | sed 's/v003-.*.he5/v003.he5/' | sed `
done

Your should have your original file and a sym link to it named

OMI-Aura_L2-OMDOMINO_o09846.he5

Then in the descriptor replace the

DSET OMI-Aura_L2-OMDOMINO_2006m0522t0926-o09846_v003-2010m1007t075325.he5

with

DSET ^OMI-Aura_L2-OMDOMINO_o%tm5.he5
OPTIONS template

and the TDEF to

TDEF 99999 linear 10:02z15jul2004 1hr

and then use set t 9846 to get orbit file 9846.
------------------------------------------------------------------------------
| James E. Johnson                      | address:                           |
| Wyle Information Systems, Inc.        |  NASA Goddard Space Flight Center  |
| e-mail: james.johnson at nasa.gov        |  Distributed Active Archive Center |
| phone:  301-614-5121                  |  Code 610.2, Bldg 32, Room S130G   |
| fax:    301-614-5268                  |  Greenbelt, MD  20771              |
------------------------------------------------------------------------------