When you download a large image from Google Earth Engine, it is in multiple small images. Each sub-image name contains the x_min and y_min coordinates of that image.
It is described here - https://developers.google.com/earth-engine/image_upload#tfrecord
How to join all these .tif files into one .tif file in right order by using a Python script.
Merging tiles can be accomplished using:
It is not necessary to order the tiles in a specific way... they are non-overlapping.
If you want to order the tif files by coordinate pair this will do it.
from osgeo import gdal
tiffiles=["basefilename-99-100.tif","basefilename-98.7-200.tif","basefilename-101-300.tif","basefilename-100.25-400.tif","basefilename-99.925-500.tif","basefilename-97.552-100.tif"]
coordarray=[]
tifdict={}
for tf in tiffiles:
ds = gdal.Open(tf)
w = ds.RasterXSize
h = ds.RasterYSize
gtrans = ds.GetGeoTransform()
minx = gtrans[0]
miny = gtrans[3] + w*gtrans[4] + h*gtrans[5]
# splittifname=tf.split("-")
coords=(minx,miny)
coordarray.append(coords)
tifdict[coords]=tf
coordarray=sorted(coordarray,key=lambda k:(k[0],k[1]))
for coord in coordarray:
### DO STUFF TO COMBINE THE RASTER TO A STACK
print(tifdict[coord])
I don't think I explain myself clearly, this will put your tif in order by coordinate from min x min y of the set, then you can merge or place the file names in order on .txt file and batch process. I change the code to use the coord value in the tif file instead of value in file name but I have not had a chance to test you may need to declare the spatial reference somewhere.
