Using GeoPandas with Rasterio to sample point data¶

This example shows how to use GeoPandas with Rasterio. Rasterio is a package for reading and writing raster data.

In this example a set of vector points is used to sample raster data at those points.

The raster data used is Copernicus Sentinel data 2018 for Sentinel data.

In [ ]:

import geopandas
import rasterio
import matplotlib.pyplot as plt
from shapely.geometry import Point

Create example vector data¶

Generate a geodataframe from a set of points

In [ ]:

# Create sampling points
points = [
    Point(625466, 5621289),
    Point(626082, 5621627),
    Point(627116, 5621680),
    Point(625095, 5622358),
]
gdf = geopandas.GeoDataFrame([1, 2, 3, 4], geometry=points, crs=32630)

The GeoDataFrame looks like this:

In [ ]:

gdf.head()

Open the raster data¶

Use rasterio to open the raster data to be sampled

In [ ]:

src = rasterio.open("s2a_l2a_fishbourne.tif")

Let's see the raster data with the point data overlaid.

In [ ]:

from rasterio.plot import show

fig, ax = plt.subplots()

# transform rasterio plot to real world coords
extent = [src.bounds[0], src.bounds[2], src.bounds[1], src.bounds[3]]
ax = rasterio.plot.show(src, extent=extent, ax=ax, cmap="pink")

gdf.plot(ax=ax)

Sampling the data¶

Rasterio requires a list of the coordinates in x,y format rather than as the points that are in the geomentry column.

This can be achieved using the code below

In [ ]:

coord_list = [(x, y) for x, y in zip(gdf["geometry"].x, gdf["geometry"].y)]

Carry out the sampling of the data and store the results in a new column called value. Note that if the image has more than one band, a value is returned for each band.

In [ ]:

gdf["value"] = [x for x in src.sample(coord_list)]
gdf.head()