The WCS driver stores data in folder $HOME/.gdal/wcs_cache. The folder
contains a text file db which contains keys and URLs and responses from
the server
// key.xml = service file
// key.xml.aux.xml = metadata file
// key.xml = Capabilities response
// key.aux.xml = Global metadata
// key.DC.xml = DescribeCoverage response
The issue could perhaps be debugged with those files.
I tried
gdalinfo -if WCS -oo CLEAR_CACHE
"WCS:https://elevation.nationalmap.gov/arcgis/services/3DEPElevation/ImageServer/WCSServer?version=2.0.1&coverage=DEP3Elevation";
(CLEAR_CACHE clears the whole cache so use with caution) but it returns
504 and a HTML file describing an error.
Best,
Ari
Carl Godkin kirjoitti 2.11.2020 klo 16.38:
Hi Jukka,
That's very interesting. Thanks a lot. I wonder if I can figure out
how the X dimension is being changed to (or interpreted as) such a
large range.
Certainly something that I can dig into step by step.
I can try sending your findings along to the USGS and hope that
someone there takes an interest. This is a potentially very useful
facility and I would be surprised if no one else had tried using GDAL
tools with it!
Thanks again,
carl
On Sun, Nov 1, 2020 at 11:37 PM jratike80
<jukka.rahko...@maanmittauslaitos.fi
<mailto:jukka.rahko...@maanmittauslaitos.fi>> wrote:
Hi,
By adding "-- debug on" into your command it is possible to get
details
about what GDAL is doing. Here are the main parts from the log
gdal_translate
"WCS:https://elevation.nationalmap.gov:443/arcgis/services/3DEPElevation/ImageServer/WCSServer?version=2.0.1&coverage=DEP3Elevation";
test.tif -projwin -107.03 37.28 -107 37.25 -projwin_srs EPSG:4326
-tr 100
100 --debug on
GDAL:
GDALOpen(WCS:https://elevation.nationalmap.gov:443/arcgis/services/3DEPElevation/ImageServer/WCSServer?version=2.0.1&coverage=DEP3Elevation,
this=000002944175BD50) succeeds as WCS.
GDAL: Using GTiff driver
Input file size is 40075015, 20498394
GDAL: Using GTiff driver
0GDAL: GDAL_CACHEMAX = 1632 MB
GDAL: GDALDatasetCopyWholeRaster(): 33*42 swaths, bInterleave=0
WCS: DirectRasterIO(8122982,14330794,3340,4196) -> (33,42) (1 bands)
WCS: Requesting
https://elevation.nationalmap.gov:443/arcgis/services/3DEPElevation/ImageServer/WCSServer?SERVICE=WCS&REQUEST=GetCoverage&VERSION=2.0.1&COVERAGEID=DEP3Elevation&SUBSET=x%28-20037507.0672000013,-11911185.14836644%29&SUBSET=y%284474012.1150791775,4478208.115037268%29&SCALESIZE=x%2833%29,y%2842%29&Format=image/tiff
HTTP:
Fetch(https://elevation.nationalmap.gov:443/arcgis/services/3DEPElevation/ImageServer/WCSServer?SERVICE=WCS&REQUEST=GetCoverage&VERSION=2.0.1&COVERAGEID=DEP3Elevation&SUBSET=x%28-20037507.0672000013,-11911185.14836644%29&SUBSET=y%284474012.1150791775,4478208.115037268%29&SCALESIZE=x%2833%29,y%2842%29&Format=image/tiff)
HTTP: libcurl/7.70.0-DEV OpenSSL/1.1.1g zlib/1.2.3
WCS: GDALOpenResult() on content-type: multipart/related;
boundary="wcs";
start="GML-Part"; type="text/xml"
GDAL: GDALOpen(/vsimem/wcs/000002944175BD50/wcsresult.dat,
this=0000029441C0D9A0) succeeds as GTiff.
GDAL: GDALClose(/vsimem/wcs/000002944175BD50/wcsresult.dat,
this=0000029441C0D9A0)
...10...20...30...40...50...60...70...80...90...100 - done.
GDAL: Flushing dirty blocks:
0...10...20...30...40...50...60...70...80...90...100 - done.
Image test.tif looks somehow corrupted from the right side.
The only WCS request that GDAL sends can be found from the log. I
removed
url-encoding for clarity
https://elevation.nationalmap.gov:443/arcgis/services/3DEPElevation/ImageServer/WCSServer?SERVICE=WCS&REQUEST=GetCoverage&VERSION=2.0.1&COVERAGEID=DEP3Elevation&SUBSET=x(-20037507.0672000013,-11911185.14836644)&SUBSET=y(4474012.1150791775,4478208.115037268)&SCALESIZE=x(33),y(42)&Format=image/tiff
Unfortunately the server returns the GeoTIFF within a multipart
content with
some GML. I do not know how to extract the GeoTIFF from the multipart
package so I could compare the results obtained by using WCS service
directly with curl or a browser. But because GDAL does not do much
for the
result, just extract the tiff part and write it into separate
GeoTIFF, I
suppose that it is the WCS server that generates the corrupted output.
However, the request that GDAL sends feels odd to me. Compare the
projwin in
gdal_translate command with subsets in the GetCoverage
-projwin -107.03 37.28 -107 37.25
SUBSET=x(-20037507.0672000013,-11911185.14836644)
SUBSET=y(4474012.1150791775,4478208.115037268)
Subract the numbers and notice that subset x is 8126321.919 meters
while
subset y is 4196 meters.
I can see number 4196 also in this line of the log
WCS: DirectRasterIO(8122982,14330794,3340,4196) -> (33,42) (1 bands)
Perhaps the meaning is to create SUBSET=x() so that it is 3340
meters wide,
not 8126322? And maybe the ESRI WCS server creates a corrupted
tiff when it
receives a somewhat lunatic request?
-Jukka Rahkonen-
--
Sent from: http://osgeo-org.1560.x6.nabble.com/GDAL-Dev-f3742093.html
_______________________________________________
gdal-dev mailing list
gdal-dev@lists.osgeo.org <mailto:gdal-dev@lists.osgeo.org>
https://lists.osgeo.org/mailman/listinfo/gdal-dev
_______________________________________________
gdal-dev mailing list
gdal-dev@lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/gdal-dev
_______________________________________________
gdal-dev mailing list
gdal-dev@lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/gdal-dev
