Published March 31, 2026 | Version v20260331
Software Open

Point to area comparison using Jensen-Shannon disimilarity index

Description

OEMC project task: 4.9 Integration and harmonization between in-situ and gridded data

Computing the potential impact of the spatial mismatch between in-situ observations and remote sensing products is a fundamental step to improve the representation of the ecosystem states at a global scale. In this project task, we explore the use of the Jensen-Shannon distance as an indicator of the mismatch between eddy covariance towers and different potential remote sensing areas. We expect this will help to improve the upscaling in situ estimates of Sun-Induced Fluorescence and Gross Primary Productivity to variables measured by satellite missions. Nevertheless, the framework is generic enough that the Jensen-Shannon can be computed and applied for the upscaling of other biophysical variables.

The following repository contains three tutorials to compute the Jensen-Shannon distance (dissimilarity metric) using the Copernicus Data Space Ecosystem. Specifically uses the SentinelHub backend through the xcube package and the plugin xcube-sh

The notebooks cover several potential use cases:

  1. Final_product_1.ipynb: One point to radius time series.
  2. Final_product_2.ipynb: Multiple area sensitivity, based on cloud free composite.
  3. Final_product_3.ipynb: Two polygons comparison, using user-defined polygons.

You can run the notebooks locally on your computer or directly in the Copernicus Data Space Ecosystem Jupyterlab You only need to drag and drop the notebooks in your mystorage folder. Be sure to select Sentinel Hub python kernel before running the notebooks.

Acknowledgement

This project has received funding from the Open-Earth-Monitor Cyberinfrastructure project that is part of European Union's Horizon Europe research and innovation programme under grant 101059548.

Funding awards

Additional details

See also

Created:
September 16, 2026
Modified:
September 28, 2026