The first in a new monthly series to highlight a data project that fuels innovative research and discoveries. 

NASA’s Harmonized Landsat and Sentinel-2 (HLS) project provides new surface reflectance data for all global land masses, except Antarctica, every two to three days. The data come from the combined efforts of four satellites: Landsat 8 and Landsat 9, which are jointly run by NASA and the United States Geological Survey (USGS), and the European Space Agency (ESA) Sentinel-2A and Sentinel-2B.

The satellites collect data in different spectral bands with different methods, but NASA’s Interagency Implementation and Advanced Concepts Team (IMPACT) engineered the solutions that make the HLS project a success. Images from each satellite are processed through an algorithm and corrected to account for the conditions during each observation, making a seamless, compatible dataset ready for analysis. 

The possibilities for using HLS data are vast. The project allows researchers to monitor and track everything from deforestation to natural disasters to crop yields. Inline with NASA’s commitment to open science, HLS data are freely available for anyone to use. 

As scientists leverage artificial intelligence and machine learning techniques for greater discovery, the vast repository of HLS images has become a dataset of choice for Earth researchers looking to train AI models. A team of experts from NASA and IBM Research recently launched an AI foundation model for Earth scientists, trained entirely on HLS data, to help fuel AI studies of the planet. Just like the HLS data, the model is free and open for anyone to use. 

To learn more about the HLS project and see more examples of science powered by HLS data, watch NASA Goddard’s Data in Harmony video feature.