Re: remap MODIS data to WRF domain by ESMF

Peng,

The ESMF regridding code doesn't take into account the map projection of your data.

By "curvilinear" we simply mean that your lat/lon coordinates are 2-dimensional, which is the case here for both your source and desination grids.

If you are using ESMF_regrid, you don't really need to worry about whether your coordinates are 2D or not, but you do need to provide the coordinates via the SrcGridLat/SrcGridLon/DstGridLat/DstGridLon options:

Assuming the lat/lon arrays have been read off both files:

    Opt@SrcGridLat = modis_lat2d ; source grid
    Opt@SrcGridLon = modis_lon2d
    Opt@DstGridLat = wrf_lat2d ; destination grid
    Opt@DstGridLon = wrf_lon2d

--Mary

On Oct 26, 2012, at 3:56 PM, Liu, Peng wrote:

> Dear Mary,
> Thank you for the help. I am trying to use MODIS L3 data which is in sinusoidal projection. I just want to make sure
> that sinusoidal projection is curvilinear so that ESMF can deal with it, doesn't it?
>
> Thanks again
>
> Peng
> ----- Original Message -----
> From: Mary Haley <haley@ucar.edu>
> To: Peng Liu <pliu34@gatech.edu>
> Cc: ncl-talk@ucar.edu
> Sent: Fri, 26 Oct 2012 11:06:00 -0400 (EDT)
> Subject: Re: remap MODIS data to WRF domain by ESMF
>
> Peng,
>
> You can set the special SrcMask2D resource to indicate where your missing values are. You have to give it an integer array that is the same size as your data, where "1" indicates valid values, and 0 indicates missing values.
>
> Something like:
>
> Opt@SrcMask2D = where(.not.ismissing(data),1,0)
>
> --Mary
>
> On Oct 26, 2012, at 8:21 AM, Liu, Peng wrote:
>
>> Dear NCL usres,
>> I am trying to remap the 1km MODIS data to a 4km resolution WRF domain and have a question about dealing with satellite data with ESMF, that is how the regrid method deal with the missing value in satellite data?(e,g. I need surface temperature from MODIS,hence no data exist when cloud exists)
>>
>> Thanks a lot
>>
>> Peng
>> ----- Original Message -----
>> From: Dennis Shea <shea@ucar.edu>
>> To: Peng Liu <pliu34@gatech.edu>
>> Cc: ncl-talk@ucar.edu
>> Sent: Thu, 25 Oct 2012 16:26:01 -0400 (EDT)
>> Subject: Re: remap MODIS data to WRF domain
>>
>> ESMF: 6.1.0 (released very soon)
>>
>> There are numerous ESMF examples:
>>
>> http://www.ncl.ucar.edu/Applications/ESMF.shtml
>>
>> If the 1km MODIS is a 'swath', then there are 2 examples that
>> start with swath data (source file): Examples 8 and 15
>>
>> The WRF grid is a curvilinear grid. There are numerous
>> examples where the destination grid is curvilinear.
>> One has WRF as a destination grid: Example 20.
>>
>> You can extract and modify as necessary.
>>
>> On 10/25/12 1:24 PM, Liu, Peng wrote:
>>> Dear NCL users,
>>> I am trying to remap the 1km MODIS data to a 4km resolution WRF domain, and I wonder is there any example that I can refer to and what functions in NCL
>>> are able to achieve this?
>>>
>>> thanks in advance
>>>
>>>
>>> Peng
>>> _______________________________________________
>>> 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

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Oct 29 09:34:06 2012