[ncl-talk] smops interpolation

Dennis Shea shea at ucar.edu
Thu Sep 13 21:40:34 MDT 2018


SMOPS: soil Moisture Operational Products System
*https://www.ospo.noaa.gov/Products/land/smops/*
<https://www.ospo.noaa.gov/Products/land/smops/>
Values over land only

[1]
The* most important rule in data processing* is **Look at your data"*.
Sometimes you have to look *carefully!*
*This is a user responsibility.*

netcdf NPR_SMOPS_CMAP_D20170410 {
dimensions:
        Longitude = 1440 ;
        Latitude = 720 ;
variables:
        *short* *Blended_SM(Latitude, Longitude) ;*
                Blended_SM:long_name = "Blended Soil Moisture" ;
                Blended_SM:units = "m^3/m^3" ;
                *Blended_SM:FillValue = -999s ;    <=======*
                Blended_SM:valid_range = 0s, 10000s ;
                Blended_SM:scale_factor = 0.0001f ;
                Blended_SM:add_offset = 0.f ;
[SNIP]

The NPR_SMOPS_CMAP file *does not conform to any netCDF convention*. Likely
an oversight. It uses *FillValue* rather than _*FillValue.  *The *_ *is
important!!! NCL  recognizes only *_FillValue *(Climate and Forecast netCDF
Convention) *.* Hence, NCL's *short2flt *
<http://www.ncl.ucar.edu/Document/Functions/Contributed/short2flt.shtml>will
not properly unpack the  "*short* *Blended_SM"* variable. Likely, this is
the source of some of the interpolation issues.

How to fix this?

Replace:

           sm  = *short2flt( *
<http://www.ncl.ucar.edu/Document/Functions/Contributed/short2flt.shtml>SMOPS_fi->Blended_SM
)

with

           sm  = SMOPS_fi->Blended_SM    ; Read variable of type 'short'
           sm@*_FillValue *= sm@
*FillValue  ; add correct attribute to original variable*
           printVarSummary(sm)                    ; Note *_FillValue*
attribute
           sm  *:=* *short2flt*( sm )                    ; Use NCL's
overwrite syntax  *:=*

Add coordinates which are not on the file. Another, oversight by the file
creator!

          sm!0                          = "latitude"
          sm!1                          = "longitude"
          sm&latitude               = latitude
          sm&longitude            = longitude
          printVarSummary(sm)
          printMinMax(sm, 0)
          print("---")

============
[2] Using '*poisson_grid_fill*'

   [a] will 'fill-in' gaps in the satellite mosaic swaths     :-)
   [b] it will also put 'soil moisture' over the oceans and lakes.   :-(

You can use a*n* NCL function to mask out the oceans/lakes/ice/etc
   *http://www.ncl.ucar.edu/Document/Functions/Shea_util/landsea_mask.shtml*
<http://www.ncl.ucar.edu/Document/Functions/Shea_util/landsea_mask.shtml>

%> ncl 20_SMOPS.poisson_fill_mask.ncl

creates: SMOPS_FILL_MASK.png
============

[3] The following uses linint2_points_Wrap to perform the station
interpolations.

%> 20_SMOPS.poisson_fill_mask_linint2.ncl

On Wed, Sep 12, 2018 at 5:15 AM, Ehsan Taghizadeh <ehsantaghizadeh at yahoo.com
> wrote:

> Hi,
> Again Thank you so much for your helpful comments.
> Unfortunately after a few days I still have problem with interpolating
> data on station points. In spite of using "poisson_grid_fill" and "linmsg"
> to avoid getting Fill_Value on unstructured points, these don't work; all
> values for all dates and all points are Fill_value (-0.099900) (attached
> file: 20_SMOPS_Interp.txt). Could I have any suggestion to get interpolated
> values other than "-0.099900". I've used this script for some data set and
> it helps me a lot. But for some data set, like SMOPS data, it gives me just
> Fill_Value.
> I know using "knnsearch" in matlab gives values except "-0.099900",
> however I want to interpolate with ncl.
> I've attached the script (20_SMOPS_Interp.ncl) and also station points
> (2_stations.dat) and output file (20_SMOPS_Interp.txt). I've also put
> zipped input data (SMOPS_Data.tgz) on ftp.
>
> I'll be thankful for any help.
>
> Sincerely
> Ehsan
>
>
> On Friday, September 7, 2018, 8:25:16 PM GMT+4:30, Dennis Shea <
> shea at ucar.edu> wrote:
>
>
> [1] No
> [2] Yes
>
>       nlat = 720
>       latitude = *fspan*(-49.875, 49.875, nlat)
>       latitude at units = "*degrees_north*"
>       latitude at long_name = "latitude"
>       latitude!0        = "latitude"
>       latitude&latitude =  latitude
>
>       printVarSummary(latitude)
>       printMinMax(latitude,0)
>
>       mlon = 1440
>       longitude  = *lonGlobeFo* <https://www.ncl.ucar.edu/Document/Functions/Contributed/lonGlobeFo.shtml>(mlon, "longitude", "longitude", "degrees_east")  ; 0->360
>       longitude  = (/ longitude - 180. /)  ; subtract 180 from all values
>       longitude&longitude = longitude            ; update coordinates       printVarSummary(lon)
>
>       printVarSummary(longitude)
>       printMinMax(longitude,0)
>
>       f = *addfile*("NPR_SMOPS_CMAP_D20170 410.nc","r")
>
>       x = *short2flt* <https://www.ncl.ucar.edu/Document/Functions/Contributed/short2flt.shtml>( f->Blended_SM )
>
>       x&latitude = latitude
>       x&longitude = longitude
>
>       printVarSummary(x)
>
> [3]/[4]
>
>       xo = .... longitudes of land points...
>
>       yo = ... latitudes ....
>
>       xpts = *linint2_points*(x&longitude,x&latitude,x, False, xo,yo, 0)
>
>       printVarSummary(xpts)
>
>       printMinMax(xpts,0)
>
>
>
> On Tue, Sep 4, 2018 at 5:50 AM, Ehsan Taghizadeh <
> ehsantaghizadeh at yahoo.com> wrote:
>
> Thank you so much for your nice reply, as always.
> I did your comments, however I encountered some difficulties and I'll be
> thankful if I could have any help about them, too.
>
> 1. Should I change
>
> latitude = *fspan*(-49.875, 49.875, nlat)
>
> to
>
> latitude = *fspan*(-89.875, 89.875, nlat)
>
> or like longitude, use
>
> lat  = *latGlobeFo*(nlat, "lat", "latitude", "degrees_north")
>
>
> 2. Should I change
>
> latitude at units = "degrees_east"
>
> to
>
> latitude at units = "degrees_north"
>
>
> 3. How should I control FillValue (-999s), because after running ncl
> script, it writes -0.1 for Nan value of the data, both in grided and
> interpolated output files.
> 4. However the most difficulty is the interpolating part, because all
> interpolated values for all station points got -0.1, means Nan. There is no
> value except -0.1 for any station points. Could anybody help me about this?
> I've attached the output of interpolated file. I'll be thankful for any
> help.
>
> Sincerely
> Ehsan
>
> On Friday, August 31, 2018, 6:14:53 PM GMT+4:30, Dennis Shea <
> shea at ucar.edu> wrote:
>
>
> Sometimes, you have to read documentation associated with the file(s)
> *Soil Moisture* Operational Products System (*SMOPS*)
>
> The file has:
>
> netcdf NPR_SMOPS_CMAP_D20170410 {
> dimensions:
>
> *        Longitude = 1440 ;        Latitude = 720 ;*
> variables:
>         *short* Blended_SM(*Latitude, Longitude*) ;
>                 Blended_SM:long_name = "Blended Soil Moisture" ;
>                 Blended_SM:units = "m^3/m^3" ;
>                 Blended_SM:FillValue = -999s ;
>                 Blended_SM:valid_range = 0s, 10000s ;
>                 Blended_SM:*scale_factor *= 0.0001f ;
>                 Blended_SM:*add_offset* = 0.f ;
>
> [SNIP]
>                 :Product_Resolution = "0.25 degree" ;
>                 :Date_Start = "20170410" ;
>                 :Date_End = "20170410" ;
> [SNIP]
>  ====
> It does not have Latitude[*] and Longitude[*] coordinates. However,
> **you** can create them. Then, use ESMF or (say)* linint2_Wrap *
> <https://www.ncl.ucar.edu/Document/Functions/Contributed/linint2_Wrap.shtml>or
> *linint2_points_Wrap*
> <https://www.ncl.ucar.edu/Document/Functions/Contributed/linint2_points_Wrap.shtml>
> to perform the interpolation.
>
> ====
>
>       nlat = 720
>       latitude = *fspan*(-49.875, 49.875, nlat)
>       latitude at units = "degrees_east"
>       latitude at long_name = "latitude"
>       latitude!0        = "latitude"
>       latitude&latitude =  latitude
>
>       printVarSummary(latitude)
>       printMinMax(latitude,0)
>
>       mlon = 1440
>       longitude  = *lonGlobeFo* <https://www.ncl.ucar.edu/Document/Functions/Contributed/lonGlobeFo.shtml>(mlon, "longitude", "longitude", "degrees_east")  ; 0->360
>       longitude  = (/ longitude - 180. /)  ; subtract 180 from all values
>       longitude&longitude = longitude            ; update coordinates       printVarSummary(lon)
>
>       printVarSummary(longitude)
>       printMinMax(longitude,0)
>
>       f = *addfile*("NPR_SMOPS_CMAP_D20170 410.nc","r")
>
>       x = *short2flt* <https://www.ncl.ucar.edu/Document/Functions/Contributed/short2flt.shtml>( f->Blended_SM )
>
>       x&latitude = latitude
>       x&longitude = longitude
>
>       printVarSummary(x)
>
>
>
> On Fri, Aug 31, 2018 at 5:45 AM, Ehsan Taghizadeh <
> ehsantaghizadeh at yahoo.com> wrote:
>
> Hi,
> Until now I've asked some questions related to interpolating some dataset
> on some unstructured points (synoptic stations). Now I have a new data set
> which I'm not sure about appropriate method to interpolating. I put one of
> my data (NPR_SMOPS_CMAP_D20170410.nc) in ftp.cgd.ucar.edu. I'll be
> thankful if some one could refer me to correct script, like on (https://www.ncl.ucar.edu/
> Applications/ESMF.shtml <https://www.ncl.ucar.edu/Applications/ESMF.shtml>)
> or (https://www.ncl.ucar.edu/ Applications/regrid.shtml
> <https://www.ncl.ucar.edu/Applications/regrid.shtml>), which interpolates
> my data on some special unstructured points.
> However the printVarSummary(Blended_SM) result is as below:
>
> Variable: Blended_SM
> Type: float
> Total Size: 4147200 bytes
>             1036800 values
> Number of Dimensions: 2
> Dimensions and sizes:   [Latitude | 720] x [Longitude | 1440]
> Coordinates:
> Number Of Attributes: 5
>   _FillValue :  1e+20
>   long_name :   Blended Soil Moisture
>   units :       m^3/m^3
>   FillValue :   -999
>   valid_range : (  0,  1 )
>
>
> Sincerely
> Ehsan
>
> ______________________________ _________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/ mailman/listinfo/ncl-talk
> <http://mailman.ucar.edu/mailman/listinfo/ncl-talk>
>
>
>
> ______________________________ _________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/ mailman/listinfo/ncl-talk
> <http://mailman.ucar.edu/mailman/listinfo/ncl-talk>
>
>
>
> _______________________________________________
> 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/20180913/ad44b26a/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 20_SMOPS.poisson_fill_mask.ncl
Type: application/octet-stream
Size: 5239 bytes
Desc: not available
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20180913/ad44b26a/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SMOPS_FILL_MASK.png
Type: image/png
Size: 360047 bytes
Desc: not available
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20180913/ad44b26a/attachment-0001.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 20_SMOPS.poisson_fill_mask_linint2.ncl
Type: application/octet-stream
Size: 5549 bytes
Desc: not available
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20180913/ad44b26a/attachment-0003.obj>


More information about the ncl-talk mailing list