Study area: Belgian Inland Waters (with extension to global for external collaborators)

Programme: STEREO 4

Contract: SR/00/406

Start - End: 1 december 2022 - 31 juli 2025

Project type: Exploration

Context and objectives

Recent improvements in optical remote sensing through the launch of high quality decametre and metre scale sensors, i.e. the Operational Land Imagers on board Landsat 8/9 (2013/2022) and the MultiSpectral Instrument on board of Sentinel-2A/B/C (2015/2017/2024), and commercial cubesat imagers such as those on PlanetScope/SuperDove, has increased interest in remote sensing of narrow and small water bodies. At short spatial length scales the high resolution imagery is impacted by adjacency effects, which may dominate the signal in certain bands, and render defaults products unusable for water quality applications. The RAdCor project developed a generic and fast correction algorithm for these effects, and made it available in the open source ACOLITE processor. The applicability of the processor was demonstrated using in situ measurements in Belgian lakes, and a contributed global dataset.

Project outcome

The RAdCor adjacency correction algorithm providing improved remote sensing reflectance retrievals over narrow and small inland water bodies is available in ACOLITE. These improved reflectance retrievals will lead to improved remote sensing derived products such as pigment concentration and water turbidity. The processor was validated over Belgian lakes using measurements collected during PONDER and HYPERMAQ projects. An external matchup database was also constructed using contributed data from 18 Early Adopter teams, and an intercomparison exercise using the Early Adopter data and two additional adjacency correction methods was performed.

ACOLITE/RAdCor is available to the community and allow broad application of the project results to any region. The processor was developed and tested for Landsat (5/7/8/9), Sentinel-2 (A/B/C) and SuperDove satellite sensors, and experimental code for a series of other satellites is made available (WorldView, Pléiades, MERIS, OLCI). Training resources and examples are available from the RAdCor website.