EGU24-21701, updated on 11 Mar 2024
EGU General Assembly 2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Why it's time to switch your EO data processing and analysis to the cloud 

William Ray
William Ray
  • Sinergise Solutions, Graz, Austria

Traditionally, processing and analysing satellite imagery has not been a straightforward task, with a lot of specialist knowledge required to make your data analysis ready. In addition, limited computing power made it only feasible to process and analyse individual satellite image acquisitions. 

This was not an issue as data even 10 years ago was not as numerous as it is today. For example, back in 2013, you would likely use Landsat 8 which has a 16 day revisit time at 30m resolution. In 2021 though, we now have access to Sentinel-2 with a revisit time of 5 days at 10m resolution. And if a scientist wished to develop any multitemporal application then they would likely rely on super-computers funded and managed by their institution. 

However, with the advances in computing power and the improvements in satellite revisit times meaning there is more data available now means that processing imagery on your desktop/laptop is no longer an effective approach.

The introduction of the Copernicus Data Space Ecosystem is now making cloud processing methods accessible and open to earth observation scientists across Europe and the world. With its scalability, ease of use, and powerful data processing capabilities, the Copernicus Data Space Ecosystem is designed to be flexible and adaptable to the needs of all types of users, from researchers to intergovernmental institutions. 

In this talk, we will run a comparison exercise generating a multi-temporal satellite imagery derived product running a traditional workflow, before demonstrating an optimized version of the workflow using cloud processing APIs. A use case will be presented; where a simple harvest detection map will be derived using the Normalized Difference Vegetation Index (NDVI) and the Bare Soil Index (BSI). Using the difference in NDVI between the most recent and least recent timestamp, the BSI in the most recent timestamp, and thresholding these values, we classify pixels as either harvested or not harvested.

The same workflow will then be run using the Sentinel Hub APIs that are available to all users of the Copernicus Data Space Ecosystem. The lines of code, amount of data downloaded and the time taken to process the data will then be compared demonstrating the efficiencies that can be gained from moving your satellite imagery processing to the cloud.


How to cite: Ray, W.: Why it's time to switch your EO data processing and analysis to the cloud , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-21701,, 2024.