hsaf_grib reader projection offset

35 views
Skip to first unread message

José Daniel Gómez de Segura

unread,
Mar 3, 2022, 7:24:00 AM3/3/22
to pytroll
Hi all,

I'm trying to read HSAF grib products from their FTP site into a scene.

Apparently, the data is placed a few km to the North in the Northern hemisphere as compared to what we could agree as 'ground truth', i.e. the same data presented in EumetView.

Being playing around with QGIS, grib to netcdf translators (ecCodes and wgrib2) and the result is always the same. These are some images related to the issue:

From Eumetsat (Andorra, instantaneous_rainrate_03b, 20220225T07:30 UTC)
eumetsat1.png
Reading the grib into QGIS
grib_in_qgis2.png
From satpy into QGIS
satypy_QGIS2.png

Same sequence over Portugal
eumetsat2.png
grib_in_qgis.png
satypy_QGIS.png

Steps to reproduce:
1. Download grib from HSAF (ftphsaf.meteoam.it) In the example, /h03B/h03B_cur_mon_data/h03B_20220225_0730_fdk.grb.gz
2. Decompress it
3. Read with satpy and save a geotiff
from satpy.scene import Scene
filename=''H03b/h03B_20220225_0730_fdk.grb'
scn=Scene(reader='hsaf_grib', filenames=[filename])
scn.load(['instantaneous_rainrate_3b'])
scn.save_datasets(filename='/tmp/out.tiff')
4. Open both the grb and tiff files with QGIS

I don't discard my own errors, but all my trials with different approaches in Python lead to the same problem.

Do you think this could be a bug? I wouldn't want to have to translate the files into netcdf...

Thanks a lot,

José
Tecnalia


Reply all
Reply to author
Forward
0 new messages