Remote sensing involves collecting information about the Earth’s surface using satellites or airborne sensors.

These datasets are widely used in:

  • Climate science
  • Agriculture
  • Forestry
  • Urban monitoring
  • Environmental change detection

Python provides powerful tools for analyzing satellite imagery and raster datasets.

Key Python Libraries

Library Purpose
rasterio Reading and writing raster datasets
xarray Multidimensional data arrays
rioxarray Geospatial raster operations
numpy Numerical processing
matplotlib Visualization

Install packages:

pip install rasterio xarray rioxarray matplotlib numpy

Import libraries:

import rasterio
import xarray as xr
import numpy as np
import matplotlib.pyplot as plt

Understanding Raster Data

Satellite images are stored as rasters.

A raster contains:

  • rows and columns
  • pixels representing measurements
  • geographic coordinate metadata

Examples:

Pixel Value Meaning
0 water
1 vegetation
2 urban

Opening Satellite Data

Load a raster dataset:

import rasterio

dataset = rasterio.open("satellite_image.tif")

Inspect metadata:

print(dataset.meta)

Read raster band:

band = dataset.read(1)

Visualizing Satellite Imagery

Display image:

plt.imshow(band)
plt.colorbar()
plt.title("Satellite Image")
plt.show()

Working with Multi-band Imagery

Satellite images often contain multiple spectral bands.

Example bands:

Band Wavelength Use
Blue visible water detection
Red visible vegetation
NIR near infrared plant health

Load multiple bands:

red = dataset.read(3)
nir = dataset.read(4)

Calculating Vegetation Indices

A common remote sensing index is NDVI (Normalized Difference Vegetation Index).

NDVI measures vegetation health.

Formula:

NDVI = (NIR - Red) / (NIR + Red)

Calculate NDVI:

ndvi = (nir - red) / (nir + red)

Visualize NDVI:

plt.imshow(ndvi)
plt.colorbar()
plt.title("NDVI Vegetation Index")
plt.show()

Working with xarray

Satellite datasets are often large and multidimensional.

xarray provides labeled arrays for working with these datasets.

Load raster using rioxarray:

import rioxarray

data = rioxarray.open_rasterio("satellite_image.tif")

Inspect dataset:

print(data)

Time Series Satellite Data

Many Earth observation datasets contain temporal data.

Example:

  • monthly vegetation index
  • daily temperature
  • yearly land cover

xarray handles time dimensions easily.

Example:

dataset.mean(dim="time")

Cropping Satellite Images

Extract specific geographic regions.

from rasterio.mask import mask

cropped, transform = mask(dataset, shapes, crop=True)

Change Detection

Remote sensing is often used to detect environmental change.

Example:

  • deforestation
  • urban expansion
  • glacier retreat

Calculate difference between images:

change = image_2024 - image_2015

Visualize change:

plt.imshow(change)
plt.colorbar()

Large Satellite Datasets

Earth observation datasets can be extremely large.

Strategies for handling large data:

  • chunked processing
  • lazy loading
  • cloud optimized GeoTIFFs
  • using xarray + dask

Example:

data = xr.open_dataset("large_satellite_dataset.nc")

Example Workflow

Typical remote sensing workflow:

import rasterio
import numpy as np

dataset = rasterio.open("image.tif")

red = dataset.read(3)
nir = dataset.read(4)

ndvi = (nir - red) / (nir + red)

Applications of Remote Sensing

Satellite analysis is used for:

  • monitoring forests
  • tracking crop health
  • detecting wildfires
  • mapping land use change
  • monitoring climate change

Python tools make these analyses accessible for research and environmental monitoring.

Summary

Python provides powerful tools for analyzing satellite data.

Key capabilities include:

  • reading raster datasets
  • analyzing spectral bands
  • calculating vegetation indices
  • detecting environmental change
  • working with large Earth observation datasets

These tools allow scientists and analysts to study the Earth at global scale.

Next Steps

Advanced remote sensing topics include:

  • satellite image classification
  • deep learning for imagery
  • hyperspectral analysis
  • cloud-based Earth observation platforms
  • real-time environmental monitoring
« Previous: Geospatial Machine Learning