I am trying to get the ground level altitude from either DSM or PC. I am new to GIS and Python and still learning, any pointer or source?

I have a latitude and longitude of the point, for the same I want altitude from DSM or PC.

1 Answer 1


Usually DSM or DEM rasters are georeferenced Tiff files. You can use GDAL or rasterio library to load them into Numpy Array.

    import numpy as np
    from osgeo import gdal

    raster = gdal.Open("dem.tif")
    elevation_array = np.array(raster.GetRasterBand(1).ReadAsArray())

The georeferenced Tiff files will have information about corner coordinates of the 2D elevation raster, width and height per pixel, row rotation and column rotation as part of its GeoTransform. You can obtain this GeoTransfrom using following method.

transform = raster.GetGeoTransform()

It should print a tuple containing 6 values like

(x-coordinate of the upper-left corner of the upper-left pixel, west-east pixel resolution / pixel width, row rotation (typically zero), y-coordinate of the upper-left corner of the upper-left pixel, column rotation (typically zero),north-south pixel resolution / pixel height (negative value for a north-up image)

Refer documentation of GetGeoTransfrom

Once you have this transform values, you can use it to convert between row,col of the numpy array of the raster to lat,lon and reverse.

So given lon, lat (or x,y)

# x,y to col,row
col = int((x - transform[0]) / transform[1]) 
row = int((y - transform[3]) / transform[5])

now look up the elevation_array with col and row by

elevation = elevation_array[row][col]

this gives you elevation at that location.

  • Also one more issue how can we load such data from S3 for example , the DSM files are huge in size. In this example entire file is loaded in memory. Jun 3, 2022 at 9:58
  • You do one time processing of these rasters and break them down to level 12 Mercator tiles. Index them by Mercator tile id's. During query first identify which mercator tile the query lon lat is falling into and accordingly load a level 12 raster. This will occupy lesser memory. Other way could be loading it into a database like postGIS which allows to query rasters.
    – Abhilshit
    Jun 3, 2022 at 20:36
  • You can also host up a geoserver to query the DEM rasters.
    – Abhilshit
    Jun 3, 2022 at 20:39

Not the answer you're looking for? Browse other questions tagged or ask your own question.