I tried to use m3tshift for the COMBINE file. And it seems the deviation from the date is not happening for the COMBINE file diff_combine_o3.txt (405.7 KB)
The AMET run is successful. But the script didn’t extract data for the specific clim_region “South”. I plotted a spatial plot to verify that using the above setup.
Glad to know that this temporary workaround of fixing the non-compliant time stamps in the COMBINE file (inherited from the non-compliant daily files) with m3shift worked as expected. And yes, comparing the metadata between the initial file created by COMBINE and the one fixed with m3shift should show differences for only a handful of attributes directly related to the creation and writing of the file as well as the program creating the file (aside from the time stamps fixed by m3tshift, of course). That’s because both programs are fortran tools that followed the proper approach of linking in the I/O API library and using its functions to read/create/write files, while whatever tool created the daily files didn’t follow that approach.
Your remaining question is one that’ll have to be answered by @wyat.appel since the AMET analysis scripts are outside my area of expertise.
Also note that you can only specify a single AMET_ADD_QUERY environment variable. It looks like you have possibly tried to set the AMET_ADD_QUERY environment variable twice. If you want to apply both of the those queries, you’ll need to do it in a single query string. For example:
setenv AMET_ADD_QUERY “and (d.month=05 or d.month=06 or d.month=07 or d.month=08 or d.month=09) and d.O3_8hrmax_ob >= 60”
The string above will limit the analysis by May-Sept and limit the MDA8 O3 obs to just those values above 60. The way you had it written, the second setenv would simply overwrite the first setenv, so you’d only be limiting the obs, not applying the May-Sept criteria.
