[ncl-talk] interpolation to finer resolution

[Debasish Hazra]

Thanks Mary, that resolved the issue.

Debasish

On Tue, Sep 26, 2017 at 11:03 AM, Mary Haley <haley at ucar.edu> wrote:

> Hi Debasish,
>
> The linint2 routine requires that your lat/lon arrays be monotonically
> increasing. In the icecvr7 variable, the latitudes are decreasing:
>
> Variable: icecvr7
> Type: float
> Total Size: 260640 bytes
>             65160 values
> Number of Dimensions: 2
> Dimensions and sizes: [lat_0 | 181] x [lon_0 | 360]
> Coordinates:
>             lat_0: [-90..90]     <---------------------------------
>             lon_0: [ 0..359]
>
> I see that you went through some trouble to reverse the latitude array by
> creating a new one, but the problem is that only the latitude array was
> reversed, and not the lat dimension of the icecvr7 variable, which is also
> required.
>
> To reverse a dimension, you can simply use the ::-1 syntax:
>
>      icecvr7 = e1->ICEC_P0_L1_GLL0(::-1,:) ; reverse the latitudes
>
>
> This will have the effect of reversing both the latitude coordinate array,
> *and* the leftmost dimension of icecvr7 at the same time.
>
> When using linint2, you also need to create the output latitudes as
> increasing as well:
>
>      newlat = fspan(-89.75,89.75,360)
>
> Your linint2 section will now look like this:
>
>      icecvr7 = e1->ICEC_P0_L1_GLL0(::-1,:) ; reverse the latitudes
>
>
>      newlat = fspan(-89.75,89.75,360)
>      newlon = fspan(0.25, 359.75, 720)
>      newlat at units = "degrees_north"
>      newlon at units = "degrees_east"
>
>      new_icecvr7 = linint2(icecvr7&lon_0,icecvr7&
> lat_0,icecvr7,True,newlon,newlat,0)
>
> If you need for new_icecv7 to be north-to-south in the latitude dimension,
> then use ::-1 again. For plotting, however, it doesn't matter which
> direction the latitudes are going:
> new_icecvr7 = new_icecvr7(::-1,:)
>
> The rest of the code should be fine and you should get the attached plot.
>
> -Mary
>
>
> On Tue, Sep 26, 2017 at 8:40 AM, Debasish Hazra <debasish.hazra5 at gmail.com
> > wrote:
>
>> Hi,
>>
>> I am trying to interpolate ice cover from 1 degree to 0.5 degree using "linint2",
>> but the attached plot shows near empty field after interpolation. Do I need
>> to use different interpolations for the data set. Attached is ncl plot and
>> resulting figure. I have uploaded data file "ocnf2015040900.grb2" in ftp
>> server. Any help.
>>
>> 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/20170926/acb8f495/attachment.html>