I am trying to work out the relationship between the 'processing cell size' and the result from Tabulate Areas. If one uses the same as the raster cell size, and this is the same for both rasters, the area output is correct.
However if one uses another value, for example use a processing size of 50 to a cell size of 25×25. The result is roughly four times higher. This suggests it is counting the overlapping cells and calculating the area by multiplying it with the 'processing cell size' value, over riding the rasters own cell size.
Can this be right? I would have expected the processing cell to size to be the unit of measurement, e.g. to allow adjustment of the resolution of the area calc – if for example one uses different resolution rasters or a vector set.
Any clarification would be welcome.
Best Answer
The cell size is not used to allow adjustment. You have to manually adjust the resolution of the coarser raster to make it match the finer raster (or set the Cell size raster analysis environment to Minimum of Inputs).
https://desktop.arcgis.com/en/desktop/latest/tools/spatial-analyst-toolbox/h-working-with-tabulate-area.htm
"Some of the area values in the output table may be smaller than what you may have expected.
This is typically related to NoData cells in the input with a higher resolution becoming a larger area of NoData after it is resampled to match the other, coarser input.
To avoid this situation, use either Resample on the coarser input raster to make it match the resolution of the finer input raster, or set the Cell size raster analysis environment to Minimum of Input"