I am trying to use Empirical Bayesian Kriging interpolation method to interpolate some point values. I want the cells of the output raster (Raster B) to align perfectly with another raster (Raster A) layer. Because later I need to do some map algebra between these two rasters. I did some configurations in Environments under Geoprocessing as below:
- Processing Extent: Extent -- same as Raster A; Snap raster -- Raster A.
- Raster Analysis: Cell size -- same as Raster A; Mask -- extent polygon of Raster A.
However, the interpolated Raster B didn't snap to the edges of the cells of Raster A as I expected. I further converted Raster B to points so I can still somehow attach the values of Raster B to Raster A. However, I am confused about how the snap raster function works. If it takes the extent, cell size and snap options as I specified above, shouldn't the cells of Raster B align with the cells of Raster A? Below is a screenshot of the three layers(Raster A: discrete color coded cells, Raster B: the transparent white layer, and Black points: points converted from Raster B.)
From the graph, the points converted from Raster B align right at the centers of cells in Raster A. One of my guesses is environmental settings worked. However, snap raster operation only snaps to the center of the cells and use those locations to create new raster and then add a little bit margin to satisfy the cell size specified? But the explanation for the function says "the lower left corner of the extent is snapped to a cell corner of the snap raster... ". Form my output, it didn't snap to a cell corner. Did I miss something in the configuration?
By the way, all original points fall inside of Raster A. But the extent of all the points slightly smaller than the extent of Raster A. Thanks!