Using gdal warp to clip raster

Question

I am trying to clip a raster data with GDAL warp. I want data values of raster within the buffer.

import rasterio as rio
import shapely as shp
import geopandas as gp
import numpy as np

raster=rio.open('C:/Users/schol/montreal_500m.tif')
band=raster.read()

y=45.508888
x=-73.561668

poly=shp.geometry.Point(x,y).buffer(0.1)
crs='EPSG:4326'
my_poly=[poly]
mydat=gp.GeoDataFrame(crs=crs,geometry=my_poly)
mydat.to_file("C:/Users/schol/mypo.shp")

import gdal

ras_in='C:/Users/schol/montreal_500m.tif'
shp_in="C:/Users/schol/mypo.shp"
ras_out='C:/Users/schol/montreal_clip.tif'

result=gdal.Warp(ras_out,ras_in,cutlineDSName=shp_in,dstNodata=np.nan) 

When I get the minimum and maximum value - minimum says -41.46 and the maximum says nan. When seeing the clipped raster array it has a maximum value of 200. Also the input raster and the clipped raster both have same dimension. How is this happening and where am I doing it wrong?

Also, when I use this line below, I am not getting the clipped raster as output at all. And I need to get the clipped output like this with gdal.

result=gdal.Warp(ras_out,ras_in,cutlineDSName=shp_in,cropToCutline=True,dstNodata=np.nan)

Answer 1

Have a look at the gdalwarp documentation https://gdal.org/programs/gdalwarp.html. It is not so clearly expressed but the cutline makes only a mask

cutline

Enable use of a blend cutline from the name OGR support datasource.

It seems that you want to crop the image by the same and for doing that you must use also another option

-crop_to_cutline

Crop the extent of the target dataset to the extent of the cutline.

For the reference how it goes with gdalwarp from the command line. I created and saved the buffer geometry into GeoJSON. As WKT it is.

POLYGON (( -73.461668 45.508888, -73.46358947195968 45.48937896779839, -73.46928004674886 45.47061965676349, -73.47852103876974 45.45333097669804, -73.49095732188134 45.43817732188134, -73.50611097669804 45.42574103876974, -73.52339965676349 45.41650004674887, -73.54215896779839 45.41080947195967, -73.561668 45.408888, -73.58117703220161 45.41080947195967, -73.59993634323651 45.41650004674887, -73.61722502330196 45.42574103876974, -73.63237867811866 45.43817732188134, -73.64481496123025 45.45333097669804, -73.65405595325113 45.47061965676349, -73.65974652804032 45.48937896779839, -73.66166799999999 45.508888, -73.65974652804032 45.52839703220161, -73.65405595325113 45.54715634323651, -73.64481496123025 45.56444502330196, -73.63237867811866 45.57959867811866, -73.61722502330196 45.592034961230254, -73.59993634323651 45.601275953251125, -73.58117703220161 45.606966528040324, -73.561668 45.608888, -73.54215896779839 45.606966528040324, -73.52339965676349 45.601275953251125, -73.50611097669804 45.592034961230254, -73.49095732188134 45.57959867811866, -73.47852103876974 45.56444502330196, -73.46928004674886 45.54715634323651, -73.46358947195968 45.52839703220161, -73.461668 45.508888 ))

Command 1:

gdalwarp -cutline -crop_to_cutline Montreal_buffer.json montreal_500m.tif cropped.tif

Command 2:

gdalwarp -cutline Montreal_buffer.json -crop_to_cutline montreal_500m.tif cropped.tif

Result 1:

Result 2: