[ncl-talk] Station obs overlay on map

[Dave Allured - NOAA Affiliate]

Debasish, you have this at the start of your program:

     obs_lat = z->hdr_lat           ; indexed by nhdr
     obs_lon = z->hdr_lon         ; indexed by nhdr
     obs_hid = z->obs_hid        ; indexed by nobs

However, the first two lines are not right.  You need obs coordinates
indexed by the obs index, not the header index.  As you already said,
"obs_hid" cross-references the obs index to the header index.  NCL  will
conveniently cross-reference the coordinates by using vector subscripting.

     hdr_lat = z->hdr_lat
     hdr_lon = z->hdr_lon
     obs_hid = z->obs_hid

     obs_lat = (/ hdr_lat(obs_hid(:)) /)        ; now indexed by nobs
     obs_lon = (/ hdr_lon(obs_hid(:)) /)       ; now indexed by nobs

After this, use only the new obs_lat and obs_lon for station coordinates in
the rest of the program.  The only other obs variables that you should need
for plotting are obs_vid to select only the PM2.5 values, and obs_val for
the actual data values.


On Fri, Jul 24, 2026 at 10:29 PM Debasish Hazra <debasish.hazra5 at gmail.com>
wrote:

> Thanks Dave. Made some inroads related to the input file. As there can be
> multiple variables stored for each station, that is why the number of
> observations is larger. obs_var is the list of possible variables and
> obs_vid is for the index of the obs_var for each observation value. There
> is also the obs_hid, which lists the index of the header variables,
> including lat/lon/etc for each observation. Attached is an input file, with
> ncl program with more print statements to show obs_var, obs_vid and an
> output txt file that shows print statements output.
> So in this case, based on obs_vars, trying to read pm25 (2nd variable in
> obs_vars), with corresponding obs_vids and obs_val and then group them in
> selected bins to overlay on map. Any help on this is appreciated
> Thanks,
> Debasish
>
> On Mon, Jul 20, 2026 at 2:43 PM Dave Allured - NOAA Affiliate <
> dave.allured at noaa.gov> wrote:
>
>> Debasish, the problem starts with the Airnow input file.  It has obs_lat
>> and obs_lon on dimension nhdr, but pm25_data is on dimension nobs.  These
>> are two different sizes.  From the Airnow file, can you determine how to
>> line up the coordinates for each of the 3982 obs values?  My guess is that
>> there is some kind of cross referencing in the Airnow file, such as site ID
>> arrays.
>>
>>
>> On Mon, Jul 20, 2026 at 11:20 AM Debasish Hazra <
>> debasish.hazra5 at gmail.com> wrote:
>>
>>> Thanks Dave, attached the revised ncl code with print statements you
>>> have suggested along with print output and error in text message.
>>>
>>> On Mon, Jul 20, 2026 at 12:51 PM Dave Allured - NOAA Affiliate <
>>> dave.allured at noaa.gov> wrote:
>>>
>>>> Debasish, please check your subscripting.  Immediately before the line
>>>> that gets the error, add printVarSummary for lat_new, obs_lat, and
>>>> "indexes".  Also include:
>>>>
>>>>     printMinMax (indexes, 0)
>>>>     print ("num missing = " + num (ismissing (indexes)))
>>>>     print (i)
>>>>     print (npts_range)
>>>>
>>>> Show us all of this output immediately preceding the error, including
>>>> the error message itself.  Do not show repeated outputs, only the last set
>>>> of debug print messages before the error occurs.
>>>>
>>>>
>>>> On Mon, Jul 20, 2026 at 8:05 AM Debasish Hazra via ncl-talk <
>>>> ncl-talk at mailman.ucar.edu> wrote:
>>>>
>>>>> Thanks Karin for your reply. The file is big to attach, here is the
>>>>> link of actual file :
>>>>> https://nomads.ncep.noaa.gov/pub/data/nccf/com/gens/prod/gefs.20260718/00/chem/pgrb2ap25/
>>>>> I am getting error in this line : lat_new(i,0:npts_range-1) =
>>>>> obs_lat(indexes)
>>>>> error says : fatal:Subscript out of range, error in subscript #0
>>>>> fatal:["Execute.c":8637]:Execute: Error occurred at or near line 68 in
>>>>> file Test_read.ncl
>>>>>
>>>>> On Mon, Jul 20, 2026 at 3:17 AM Karin Meier-Fleischer <
>>>>> meier-fleischer at dkrz.de> wrote:
>>>>>
>>>>>> Hi Debasish,
>>>>>>
>>>>>> the file gefs.chem.t00z.a2d_0p25.f018.grib2 is missing.
>>>>>>
>>>>>> Can you provide more Information about the error.
>>>>>>
>>>>>> Regards, Karin
>>>>>>
>>>>>> Am 18.07.26 um 04:26 schrieb Debasish Hazra via ncl-talk:
>>>>>> > Hi,
>>>>>> > I am trying to do a plot with station data in netcdf overlay on
>>>>>> model
>>>>>> > output following this example here
>>>>>> > :https://www.ncl.ucar.edu/Applications/Scripts/polyg_8.ncl
>>>>>> > But getting subscript range error. Attached is the ncl code and
>>>>>> netcdf
>>>>>> > station data used for this. Any help on this is greatly appreciated.
>>>>>> >
>>>>>> > Thanks
>>>>>> > Debasish
>>>>>
>>>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.ucar.edu/pipermail/ncl-talk/attachments/20260729/ead84a1c/attachment.htm>