EPSG:4326 and EPSG:6933 raster handling #918
Description
Comparing between projections for GFW 2024 Fiji Data, raster in 4326 and raster downloaded from SeaSketch after processing. My main concern is generating unreal hours in the data through the resample. I ran a couple test cases through gp, qgis in both projections, and the ouputs from seasketch.
QGIS (4326) total sum of raster layer (bounding box around Fiji): 133205.79012040235
SeaSketch (6933) sum of raster layer (bounding box around Fiji): 439611.48988441564
GP (4326) sum within EEZ: 85494.260149
QGIS (4326) sum within EEZ: 85538.02015047893
QGIS (6933) sum within EEZ: 105867.89012767002
SeaSketch (6933) sum within EEZ: 318902.7268747569
GP (4326) sum within AOI polygon: 7677.960212
QGIS (6933) sum within AOI polygon: 7677.96021201089
QGIS (6933) sum within AOI polygon: 5823.060114627704
SeaSketch (6933) sum within AOI polygon: 6759.702502192929
GP (4326) sum within mpa polygon: 607.62
QGIS (6933) sum within mpa polygon: 605.5199997741729
QGIS (6933) sum within mpa polygon: 734.9899999592453
SeaSketch (6933) sum within mpa polygon: 2577.96397217
GFW data: gfw_2024.tif.zip
AOI polygon (high-value area, no antimeridian cross): aoi.geojson.json.zip
mpa polygon (low-value area, antimeridian crossing): mpa.geojson.json.zip