How do I find and use the OMI NO2 data products?
The OMI NO2 products OMNO2 (level-2 data, one orbit per file) and OMNO2G (global daily data from 14 or 15 orbits binned into 0.25° x 0.25° grids), are available from our data access site http://mirador.gsfc.nasa.gov/. These products contain NO2 data: total and tropospheric observed vertical column density, as well as estimated values below cloud.
Since in the OMNO2G product each lat/lon grid the data is stored or binned (it is not averaged), overlapping of swaths at edges, in particular at mid- and high latitudes, may result in more than one pixel in the bins. The third dimension (equal to 15) in L2G parameters identifies the pixels coming from different orbits.
Please make sure to use Vcdquality flag to screen the data. The Vcdquality flag zero identifies good retrievals. Also, make sure to read the README documents provided by the OMI science team. Links to these documents are available from the NO2 product sites:
In the Mirador, the subset options are available only when you put the data in shopping cart.
For the OMNO2G data product, there is a download option called "OPenDAP". If you select that option, you will notice that it gives the full structure of the data file. You can subset data and download in ASCII, by selecting the proper array indices.
The data in the OMI file is about 22 minutes later than the filename indicates. Does this indicate a problem?
No, There is nothing wrong in the time conversion. The time difference of about twenty minutes between time reported in the file name and the start time of the data reported is normal. OMI data is retrieved and reported only for the daytime orbit. The OMI team always uses the descending node equator crossing time as the 'Start' for the data file name. Since it is still dark when OMI crosses the equator on the descending track, the 'actual' data values don't commence until 20 - 30 minutes later, when the instrument crosses over Antarctica.
What is the simplest way to read ACOS and other HDF5 data in IDL?
Different data providers have different views on how to construct their HDF5 files. However, the basic way to open an HDF5 file and read data from there is very simple and should hold for any HDF5 file, with only difference on the file and dataset names. The following is an IDL example for ACOS xCO2 data. The use of 'H5D_GET_STORAGE_SIZE' function is rarely needed, but ACOS providers sometimes allow for datasets of Null-length in their v.2.8, which may cause problems depending on user's expectations:
sz=H5D_GET_STORAGE_SIZE(d_id) ;check the size of “xco2“ data field
IF sz EQ 0 THEN BEGIN
xCO2=H5D_READ(d_id) ;data are now in variable xCO2
;Close file access:
The most robust way to check the names of datasets, which in HDF5 look like full paths, is to use the command line utility h5dump with option "-n"
h5dump -n <file name>