3.4 - Working with Raster Data

Geospatial data consists fundamentally of either raster (gridded) or vector (feature-based) representations. The Sentinel-2 satellite imagery utilized here is stored as GeoTIFFs - raster datasets encoding Earth observation data and digital elevation models. Rasterio provides a Python library for programmatic manipulation of such gridded geospatial data. By reading Sentinel-2 scenes into raster array structures via Rasterio, the red, green, and blue spectral bands can be combined into composite true color images for interpretation and analysis. Working natively with the raster arrays facilitates customizable image processing, indexing to areas of interest, application of filters, etc. The tight integration of Rasterio functionality into the Python geospatial software ecosystem allows researchers to focus more fully on their Earth observation data science tasks rather than data wrangling challenges. Overall, Rasterio enables flexible, reproducible research workflows essential for the advancement of open geospatial techniques applied to urgent sustainability issues.
from rasterio import plot
from rasterio.plot import show
from rasterio.session import AWSSession
from rasterio.windows import Window