Dear CIS users,
I am stuck with my MODIS Aerosol Optical Depth dataset in HDF format.
I am trying to do the following tasks (with no sucess):
1) First, I need to convert my hourly dataset into a day mean dataset;
2) Secondly, I must convert the time format from Julian Dates to Gregorian calendar ;
3) And finally, I would like to convert the HDF dataset into a Netcdf format, to be read in other softwares (e.g., CDO, Grads), in which I need to cross this information with precipitation in regular grids.
Someone could help me please?
CIS software is able to do all this tasks?
Thanks in advance.
Prof. Roger Torres
I'm stuck with my large Modis AOD dataset
Thu, 03/23/2017 - 23:15
#1
I'm stuck with my large Modis AOD dataset
Dear Prof. Torres
I'm sorry to hear that your stuck, CIS should be able to do this task quite easily - what have you tried so far?
I presume you are using Level 2 MODIS data, in which case to aggregate to a daily mean you would need to specify a horizontal grid as well. Assuming a typical layout of MODIS files, you can do this using the aggregate command like so:
cis aggregate :/MYD04_L2/2008/203/*.hdf x=[-180,180,1],y=[-90,90,1] -o 20080721_modis_aod.nc
Here I've taken all the files from the 2008/203 folder (which is the 203rd day of the year - assuming you have mirrored the NASA repository), and aggregated them onto a one degree by one degree grid, so you now have a one day average. You could then simply loop over each day using Python/Bash whichever scripting language you prefer. You would just need to check which variables you want from each file, and specify the right horizontal grid. This will be output to a NetCDF file which CDO can work with automatically.
Alternatively, you can go to your precipitation grid directly using one collocation command:
cis col :/MYD04_L2/2008/203/*.hdf 20080721_precip_file.nc -o 20080721_modis_aod_on_precip.nc
This is assuming the precip files are cf-conformant so that CIS can read them.
Hopefully that helps, if you have any other questions please just ask.
Duncan