FAIRiCUBE / data-requests

Request data to be made available within FAIRiCUBE HUB
2 stars 0 forks source link

ChangeType.update stac_dist/dominant_leaf_type_10m/dominant_leaf_type_10m.json #303

Open fairicube-data opened 2 months ago

fairicube-data commented 2 months ago

{"filename": "dominant_leaf_type_10m/dominant_leaf_type_10m.json", "item_type": "stac_dist", "change_type": "Update", "user": "FAiRICUBE", "data_owner": true}

misev commented 2 months ago

Some minor corrections needed. Most importantly, the extent does not look right. In EEA SDI spatial catalog is

The extent that you referenced is for a WGS84 bbox. But this data is in native CRS EPSG:3035, and has coordinates in metres and not degrees.

mari-s4e commented 2 months ago

Some minor corrections needed. Most importantly, the extent does not look right. In EEA SDI spatial catalog is

The extent that you referenced is for a WGS84 bbox. But this data is in native CRS EPSG:3035, and has coordinates in metres and not degrees.

But these

  "bbox": [
    -56.5051419017044,
    24.2841770079211,
    72.90613675900907,
    58.95275109308057
  ],

are values in degrees, aren't they?

misev commented 2 months ago

But these

  "bbox": [
    -56.5051419017044,
    24.2841770079211,
    72.90613675900907,
    58.95275109308057
  ],

are values in degrees, aren't they?

@mari-s4e I think you discovered a bug in the catalog-editor (cc @baloola). This WGS84 bbox was generated by the catalog-editor, I only entered the EPSG:3035 coordinates in it (the "x" and "y" in the json).

Our WMS service also reports a WGS84 bbox which matches the one you showed in the previous comment:

                <EX_GeographicBoundingBox>
                    <westBoundLongitude>-56.50514190170437</westBoundLongitude>
                    <eastBoundLongitude>72.9061049341568029565</eastBoundLongitude>
                    <southBoundLatitude>24.28417068794855834328</southBoundLatitude>
                    <northBoundLatitude>72.66326966834436</northBoundLatitude>
                </EX_GeographicBoundingBox>
baloola commented 2 months ago

This is not a bug, catalog editor reprojects the X & Y dimensions extnets to WGS84 and generates the bbox, according to stac item spec (same goes for the geometry as well)

misev commented 2 months ago

@baloola there is a bug because the generated northBoundLatitude is 58.95275109308057, but it should be 72.66.

The problem is likely that you transform individual coordinates which is incorrect (cc @bangph for pointing this out). This can be verified with gdaltransform:

$ gdaltransform -s_srs EPSG:3035 -t_srs EPSG:4326
Enter X Y [Z [T]] values separated by space, and press Return.
7400000 5500000
72.906136759009 58.9527510923263 0

It gives 58.952 like the catalog-editor and not 72.66.

The correct way is to transform the whole bbox, e.g. with gdalwarp:

  1. Create a bbox_epsg3035.vrt file with content
    <VRTDataset rasterXSize="650000" rasterYSize="460000">
    <SRS>EPSG:3035</SRS>
    <GeoTransform>900000, 10, 0, 5500000, 0, -10</GeoTransform>
    <VRTRasterBand dataType="Byte" band="1">
    <NoDataValue>0</NoDataValue>
    <SimpleSource>
      <SourceFilename relativeToVRT="1">dummy.tif</SourceFilename>
      <SourceBand>1</SourceBand>
      <SourceProperties RasterXSize="1" RasterYSize="1" DataType="Byte" BlockXSize="1" BlockYSize="1"/>
      <SrcRect xOff="0" yOff="0" xSize="1" ySize="1"/>
      <DstRect xOff="0" yOff="0" xSize="1" ySize="1"/>
    </SimpleSource>
    </VRTRasterBand>
    </VRTDataset>
  2. put whatever dummy.tif file in the same directory
  3. check the bbox with gdalinfo:
    Origin = (900000.000000000000000,5500000.000000000000000)
    Pixel Size = (10.000000000000000,-10.000000000000000)
    Corner Coordinates:
    Upper Left  (  900000.000, 5500000.000) ( 56d30'18.51"W, 56d29' 4.75"N)
    Lower Left  (  900000.000,  900000.000) ( 23d49'33.58"W, 24d17' 3.04"N)
    Upper Right ( 7400000.000, 5500000.000) ( 72d54'22.09"E, 58d57' 9.90"N)
    Lower Right ( 7400000.000,  900000.000) ( 40d39'45.75"E, 25d32'40.96"N)
  4. reproject the file:
    gdalwarp -t_srs EPSG:4326 -of VRT bbox_epsg3035.vrt bbox_epsg4326.vrt
  5. check the result bbox with gdalinfo bbox_epsg4326.vrt:
    Origin = (-56.505141901704370,72.663269668344356)
    Pixel Size = (0.000128060211603,-0.000128060211603)
    Corner Coordinates:
    Upper Left  ( -56.5051419,  72.6632697) ( 56d30'18.51"W, 72d39'47.77"N)
    Lower Left  ( -56.5051419,  24.2841707) ( 56d30'18.51"W, 24d17' 3.01"N)
    Upper Right (  72.9061049,  72.6632697) ( 72d54'21.98"E, 72d39'47.77"N)
    Lower Right (  72.9061049,  24.2841707) ( 72d54'21.98"E, 24d17' 3.01"N)
misev commented 2 months ago

@mari-s4e FYI I created a separate issue for the bbox problem.

I don't think it's practical that I go through some 30 jsons and fix these bboxes manually - I expect this to be fixed in the catalog separately, as it probably affects already published catalog entries as well. So I suggest to ignore the bbox in this and other reviews until that issue is solved.