[ncl-talk] Difference plot contour map emply

[Mary Haley]

If you do a difference plot between two variables that have differently
ordered latitudes, this will definitely be an issue unless you reverse the
latitudes of one of the dimensions. NCL doesn't look at coordinate arrays
when you simply subtract two arrays. Also, when you think about what this
means for the coordinates of the differenced array, then you really have
problems.

See the attached script which illustrates this.

You need to make sure you reverse the latitude dimension (assuming xarray1
is the one with the reversed latitudes):

   xdiff = xarray1(::-1,:) - xarray2
   copy_VarMeta(xarray2,xdiff)        ;

--Mary

On Wed, Apr 13, 2016 at 2:53 PM, Debasish Hazra <debasish.hazra5 at gmail.com>
wrote:

> Hi,
>
> I have attached the code I used to calculate difference between two models
> at a single time step. Model 1 (grib2 file) dimension is [lat | 181] x [lon
> | 360]  and coordinates are  lat_0: [90..-90], lon_0: [
> 0..359].  _FillValue :   1e+20
>
> second one (netcdf) dimension [lat | 180] x [lon | 360] and coordinates
> are lat: [-89.5..89.5] , lon: [-179.5..179.5].  _FillValue :   -9999
>
>   I regrid the 2nd to the first one co-ordinate and make a difference plot
> over different regions.
>
> Q1. If the 1st model latitude is [90..-90] and regrided 2nd model latitude
> coordinates are latitude is [-90..90], does it have any impact on the
> difference plot ?
>
> I have attached the code, model 1 output. I ftp the second model output
> (icap_2015091500_MME_modeaod550.nc) as that is of larger size.
>
> Thanks.
> Debasish
>
>
> On Tue, Apr 12, 2016 at 1:57 PM, Mary Haley <haley at ucar.edu> wrote:
>
>> Debasish,
>>
>> Without seeing your code, we have no way of helping you debug it.
>>
>> I recommend using print statements to verify that your data after
>> interpolation is correct. If your interpolated data array is called
>> (data_interp):
>>
>> printVarSummary(data_interp)
>> printMinMax(data_interp,0)
>> print(num(ismissing(data_interp)))    ; count the number of missing values
>>
>> If you continue to have problems, it would help if we could see a plot or
>> some code.
>>
>> --Mary
>>
>>
>> On Mon, Apr 11, 2016 at 6:57 AM, Debasish Hazra <
>> debasish.hazra5 at gmail.com> wrote:
>>
>>> Hi,
>>>
>>>  I am using NCL (v6.3) to make a difference plot between two model
>>> outputs. First one (grib2 file) dimension is [lat | 181] x [lon | 360]  and
>>> coordinates are  lat_0: [90..-90], lon_0: [ 0..359].  _FillValue :   1e+20
>>>
>>> second one (netcdf) dimension [lat | 180] x [lon | 360] and coordinates
>>> are lat: [-89.5..89.5] , lon: [-179.5..179.5].  _FillValue :   -9999
>>>
>>> To make both of them equal dimension before doing differences, I used "linint2"
>>> function to interpolate 2nd one to the dimension of first. However, after
>>> doing simple difference plot between the interpolated one and the first
>>> output, I am getting all constant value and map plot remains empty.
>>> Individual plot of these 2 dataset show lot of differences. Is it because I
>>> did not assign fillvalue after interpolation , or something else ? Any
>>> suggestions.
>>>
>>> Thanks
>>> Debasish
>>>
>>> _______________________________________________
>>> ncl-talk mailing list
>>> ncl-talk at ucar.edu
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20160414/679f65c5/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: diff.ncl
Type: application/octet-stream
Size: 1851 bytes
Desc: not available
Url : http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20160414/679f65c5/attachment.obj