Why am I creating an extra row and column when I use rasterio.mask to clip a raster image?

Question

I am trying to clip Raster2 to the extent of Raster1. The clip works successfully, but it I am getting an extra row and column in the output raster.

Raster1 is a raster to the extent of the state of Maine. Raster2 is the NLCD country-wide dataset. I am getting the bounding box of Raster1 and clipping Raster2 with rasterio.mask to create Raster3. However, Raster3's shape has an additional row and column when compared to the original shape of Raster1/

src = rasterio.open(f'Raster1.tif')
idArray = src.read()[0]
idArray.shape
#(45006, 41551)

geometry = box(*src.bounds)

with rasterio.open(r'Raster2') as src2:
    NLCD, NLCD_transform = rasterio.mask.mask(src2, [geometry], crop=True)
    out_meta = src2.meta

NLCD[0].shape
#(45007, 41552)

The two datasets are in the same coordinate system, but the only thing I can think of is that their grids aren't perfectly aligned, but I can't figure how that would give me an extra row and column. Any insight here would be great.

Answer 1

It could have something to do with the precision of the resolution of the two rasters. If the precisions (and thus the effective resolutions) of the two rasters are different, the same bounding box can hold a slightly different amount of rows and columns in the second raster.