Why does gdalwarp from ortho to longlat truncate the poles? Lonlat should display -90 to +90 It is the same as eqc isn't it? They are synonyms for Plate-Carre or am I mistaken? Why is the gap showing up? It must be possible otherwise Antarctica would not be there on the plate-carre. Is there an interpolation argument I am missing?
Re-projecting an orthographic image to a equirectangular projection has some sort of problem. The poles are getting truncated. is there a parameter I am missing? Is there a difference between "eqc" "longlat" "plate-carree" ? Is this a fundamental defect in gdal tools?
gdal_translate
-of Gtiff
-a_ullr -1666666 1666666 1666666 -1666666
-a_srs
"+proj=ortho
+ellps=WGS84
+datum=WGS84
+lat_0=85
+lon_0=45"
kat.png
QQQQ-kat-ortho.tif
gdalwarp
-t_srs
"+proj=longlat
+datum=WGS84
+ellps=WGS84
+units=m"
QQQQ-kat-ortho.tif
QQQQ-kat-latlon.tif
gdal_merge.py
-o QQQQ-kat-eq-merged.tif
NE1_50M_SR_W_tenth.tif
QQQQ-kat-latlon.tif
convert QQQQ-kat-eq-merged.tif QQQQ-kat-eq-merged.png
convert QQQQ-kat-latlon.tif QQQQ-kat-latlon.png
Reprojecting an orthographic (ortho) or perspective (nsper) image to equirectangular (eqc) when the map includes polar latitudes,
REQUIRES the target extents argument: "-te -180 -90 +180 +90"
or some other set of -te corners that EXPLICITLY extends all the way up to +/-90 Latitudes, else the default xmax is set to Latitude 85. Perhaps gdal REALLY expects Mercator and snips off poles. ?? This does not appear in the documentation anywhere.
example:
gdalwarp -te -180 -90 +180 +90 -t_srs "+proj=eqc +datum=WGS84 +ellps=WGS84 +units=m" source_ORTHO.tif target_EQC.tif