WorldPop — multi-country mosaic (live)¶

A bbox over the Rwanda–Burundi border spans several countries. WorldPop is per-country, so earthlens downloads each country's raster and mosaics + crops to the AOI. Uses 1 km (wpic1km) to keep it small.

What this notebook does¶

1. Download — build a WorldPop request for a cross-border AOI, see which countries it spans, and download + mosaic + crop each country's 1 km population raster to the bbox.
2. Inspect — read the mosaicked GeoTIFF back with pyramids' Dataset to confirm its CRS and shape.

WorldPop is open (CC-BY-4.0) and needs no credentials.

Setup¶

The imports: tempfile for a throwaway download directory, earthlens for the unified EarthLens entry point, and pyramids' Dataset for reading the result back. We also make a temporary directory to hold the downloaded rasters.

In [ ]:

import tempfile

from pyramids.dataset import Dataset

from earthlens import EarthLens

download_dir = tempfile.mkdtemp()

import tempfile from pyramids.dataset import Dataset from earthlens import EarthLens download_dir = tempfile.mkdtemp()

1 · Download the multi-country mosaic¶

WorldPop serves population as one raster per country, so we point a single AOI at the Rwanda–Burundi border and let earthlens fetch every country it touches.

Build the request¶

We ask for the pop (population count) variable for 2020 at 1 km resolution (wpic1km) over a bbox straddling the border. The request is built first, on its own, so it's easy to read and tweak.

In [ ]:

backend = EarthLens(
    data_source="worldpop",
    variables=["pop"],
    start="2020",
    end="2020",
    fmt="%Y",
    aoi=[29.0, -2.9, 30.0, -2.3],
    path=download_dir,
    resolution="1km",
)

backend = EarthLens( data_source="worldpop", variables=["pop"], start="2020", end="2020", fmt="%Y", aoi=[29.0, -2.9, 30.0, -2.3], path=download_dir, resolution="1km", )

Which countries does the AOI span?¶

_iso3s lists the ISO3 country codes the bounding box overlaps — these are the per-country rasters earthlens will fetch and mosaic together.

In [ ]:

print("countries in AOI:", backend._iso3s)

print("countries in AOI:", backend._iso3s)

Download, mosaic and crop¶

download() fetches each country's raster, mosaics them, and crops the result to the AOI, returning the list of written file paths.

In [ ]:

paths = backend.download(progress_bar=False)
print("written:", [p.name for p in paths])

paths = backend.download(progress_bar=False) print("written:", [p.name for p in paths])

2 · Inspect the result¶

Reading the mosaicked GeoTIFF back with pyramids' Dataset confirms the output CRS and array shape. We grab the first written path, then read it.

In [ ]:

mosaic_path = paths[0]
ds = Dataset.read_file(str(mosaic_path))
print("mosaicked + cropped EPSG:", ds.epsg, "shape:", ds.shape)

mosaic_path = paths[0] ds = Dataset.read_file(str(mosaic_path)) print("mosaicked + cropped EPSG:", ds.epsg, "shape:", ds.shape)