Any chance I can implement the following code but using getDownloadURL() or getThumbURL()? Raise a similar question in ee Community but no response. #2179

suredream · 2024-11-22T22:05:13Z

suredream
Nov 22, 2024

As below. Tks

import ee
import rasterio as rio

import geedim as gd

# Initialize the Earth Engine and geedim
gd.Initialize()

# Define your Area of Interest (AOI)
aoi = ee.Geometry.Polygon(
    [[[-122.5155, 37.7033], [-122.5155, 37.8333], [-122.3845, 37.8333], [-122.3845, 37.7033]]]
)

# Search for NAIP images in the specified date range and region
naip_collection = (
    ee.ImageCollection("USDA/NAIP/DOQQ")
    .filterBounds(aoi)
    .filterDate('2020-01-01', '2021-01-01')
    .select(['R', 'G', 'B'])
)

# Use geedim to select and download the most recent image (or modify as needed)
naip_image = gd.download.BaseImage(naip_collection.first())

# Download the image as GeoTIFF in the original projection
naip_file = 'naip_rgb.tif'
naip_image.download(filename=naip_file, scale=1, overwrite=True)

lst_image = gd.download.BaseImage(
    ee.ImageCollection("LANDSAT/LC08/C02/T1_TOA")
    .filterBounds(aoi)
    .filterDate('2020-01-01', '2021-01-01')
    .select(['B2', 'B3', 'B4'])
    .mosaic()
)

lst_file = 'lst_composite.tif'
with rio.open(naip_file, 'r') as naip_ds:
    lst_image.download(
        filename=lst_file,
        crs=naip_ds.crs.to_string(),
        crs_transform=naip_ds.transform,
        shape=naip_ds.shape,
        overwrite=True,
    )

# inspect pixel grids
with rio.open(naip_file, 'r') as naip_ds, rio.open(lst_file, 'r') as lst_ds:
    assert naip_ds.transform == lst_ds.transform
    assert naip_ds.shape == lst_ds.shape