CDAAC atmPhys files now have different MetOp leo_ids

[Ian Culverwell]

They're now called something like

atmPhs_MTPB.2013.151.15.59.G32_2010.2640_nc

with the global attribute

fileStamp = "MTPB.2013.151.15.59.G32"

which defines

leo_id = "MTPB"

This fails the ropp_pp GRAS data identification tests, which looks for

leo_id = "ME??".

So ropp_pp won't carry out GRAS-specific processing on these data.

Solution: look for "MT??" as well.

[Ian Culverwell]

These changes have been incorporated in (the f90 bits of) r4210.

Note that although ropp_pp_occ_tool now recognises (eg) atmPhs_MTPB.2013.151.15.59.G32_2010.2640_nc as GRAS data, it still fails to process it, because there aren't enough data points:

 imin, imax, w_smooth, ocd =            1        3079          59           1
INFO (from ropp_pp_preprocess):  GRAS data preprocessing
 imin, imax, w_smooth, ocd =           25          25          57           1
INFO (from ropp_pp_cutoff_amplitude):  Cut-off (amplitude/LCF criterion). Keep data       25 to       25
... (from ropp_pp_preprocess):  Merged RS+CL data size:      1
 
WARNING (from ropp_pp_preprocess):  Error: Too few data

It's like this in all the CDAAC GRAS datasets I've looked at.

Open #371 ('whenever') to look into this. Meanwhile, close this ticket for ROPP8.0.