Project

General

Profile

Projection parameters on IMS Northern Hemisphere

Added by liren wei almost 2 years ago

Hello everyone,

I have to remap a IMS Northern Hemisphere Polar Stereographic 24 km field into latlon grid. The original grid is stored in ASCII txt file and it has been converted to netCDF. When I was trying to assign a projection coordinate system to netCDF file, I met a problem. Due to the early release of data, it is difficult to find the projection information on the IMS 24km dataset. I have to create a grid description file and attach it by "-setgrid". The grid description file (which named as "sgrid.txt"):

gridtype  = projection
gridsize  = 1048576
xsize     = 1024
ysize     = 1024
xunits    = "meters" 
yunits    = "meters" 
xfirst    = -12126597.0
xinc      = 23684.997
yfirst    = 12126840.0
yinc      = 23684.997
grid_mapping = crs
grid_mapping_name = polar_stereographic
earth_radius = 6371200.f
proj_params="+proj=stere +lat_0=90 +lat_ts=60 +lon_0=-80 +R=6371200. +units=m +no_defs" 

Using this grid description file, CDO can successfully remap my netcdf file . However there was an obvious error in output field (see the cdo_regrid.png). This means there must be some problem in the setting of projection parameters. Unfortunately, I can't find the errors.

Some information about this set of data is listed below:

1. There is a gpd file (grid description file) attached with the ASCII file:

; IMS Northern Hemisphere
; Polar Stereographic 24 km grid
; apparently 1 "bedient" along true latitude is 381 km.
; For a 1/16th mesh grid, then: 381/16 = 23.8125 km
; However, this scale does not match the values in the NOAA files
; ftp://sidads.colorado.edu/DATASETS/NOAA/G02156/metadata/imslat_24km.dat
; ftp://sidads.colorado.edu/DATASETS/NOAA/G02156/metadata/imslon_24km.dat
; so this scale is slightly smaller.
;
; The NOAA locations in these files at the lower left corner of each grid cell.
;
; The Earth radius is assumed to to be 6371.2, based on
; http://www.nco.ncep.noaa.gov/pmb/docs/libs/w3lib/w3fb05.html
;
Map Projection:                 Polar Stereographic
Map Reference Latitude:         90.0
Map Reference Longitude:        -80.0
Map Second Reference Latitude:  60.0
Map Equatorial Radius:          6371200 ; software w3fb05 documentation
Map Origin Latitude:             89.998344
Map Origin Longitude:            55.000000
Grid Map Origin Column:         511.5
Grid Map Origin Row:            511.5
Grid Map Units per Cell:        23684.997 ; slightly smaller than 1/16 bedient
Grid Width:                     1024
Grid Height:                    1024

2. There is a data description pdf file: [[https://nsidc.org/sites/nsidc.org/files/G02156-V001-UserGuide.pdf#%5B%7B%22num%22%3A31%2C%22gen%22%3A0%7D%2C%7B%22name%22%3A%22XYZ%22%7D%2C84%2C245%2C0%5D]].

The "proj4" projection parameters are indicated on page 24 (same as the "sgrid.txt"). Also, some parameters are listed in the table on Page 10. I set the file according to these parameters, but the results are wrong with these params.

Thanks in advance to the people that will answer me.

cdo_regrid.png (94.4 KB) cdo_regrid.png Spatial distribution of data

Replies (2)

RE: Projection parameters on IMS Northern Hemisphere - Added by Karin Meier-Fleischer almost 2 years ago

Hi Liren,

without your files it is not really possible to help. I can only say, when working with IMS data using Python that I had to use a different proj4 setting than described in the IMS User Guide.

Your setting:

proj_params="+proj=stere +lat_0=90 +lat_ts=60 +lon_0=-80 +R=6371200. +units=m +no_defs" 

Setting using Python:

proj_params="+proj=stere +lat_0=90.0 +lon_0=90.0 +x_0=0.0 +y_0=0.0 +no_defs" 

But as already said we need the IMS netCDF file you have generated from the original IMS ASCII data.

RE: Projection parameters on IMS Northern Hemisphere - Added by liren wei almost 2 years ago

Thanks for your reply! Fortunately I have solved this question. This error may raise from the incorrect "xfirst, yfirst, xinc, yinc" settings. I found a set of parameters that looked right by guess. I kept my "proj_params" and the result seems right. Anyway, thanks for your kindly help!

    (1-2/2)