Re: ESMF Interpolation with Missing Values

From: Mary Haley <haley_at_nyahnyahspammersnyahnyah>
Date: Mon Feb 11 2013 - 09:16:21 MST

Did you verify that you have the correct missing value set for the input data?
That is, is _FillValue set to -9999?

Can you provide me with the script and all the necessary files so I can take a look?

You can use our ftp site, if necessary:

http://www.ncl.ucar.edu/report_bug.shtml#HowToFTP

Thanks,

--Mary

On Feb 8, 2013, at 4:00 PM, Barton, Neil P. wrote:

> Hi Mary,
>
> Unfortunately, I am receiving the same answers if I include or not-include
> the below line. Any other ideas?
>
> Thank you very much,
>
> Neil
>
> On 2/8/13 1:51 PM, "Mary Haley" <haley@ucar.edu> wrote:
>
>> Hi Neil,
>>
>> Try:
>>
>> opt@SrcMask2D = where(.not.ismissing(datS),1,0)
>>
>> This provides a mask array where you don't want the regridding to occur.
>>
>> --Mary
>>
>> On Feb 8, 2013, at 1:28 PM, Barton, Neil P. wrote:
>>
>>> Hello,
>>>
>>> I'm having problems interpolating data using ESMF_regrid when my source
>>> data have missing values. I am assigning the source data a _FillValue
>>> before the ESMF_regrid call. This _FillValue assignment seems to work
>>> when reading the source data, but not during the interpolation. For
>>> example when I interpolate Temperature data, I receive the below output.
>>>
>>> (0) ESMF_regrid_with_weights: dstData
>>> (0) Dimensions: 405 120 120
>>> (0) minSrcData: 216.432
>>> (0) maxSrcData: 295.866
>>> (0) minDstData: -15.5592
>>> (0) maxDstData: 283.523
>>>
>>> As shown, the minDstData is much less than the minSrcData (and
>>> unrealistic), and I am assuming this is because the _FillValue is set to
>>> 9999.0. Any help is greatly appreciated. I attached my ncl script to
>>> this email.
>>>
>>> Thank you very much,
>>>
>>> Neil
>>> <MERRA_interp.ncl>_______________________________________________
>>> ncl-talk mailing list
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Feb 11 09:16:37 2013

This archive was generated by hypermail 2.1.8 : Wed Feb 13 2013 - 09:25:58 MST