Preliminary Result of Synergy between Optical Satellite and GNSS-R Technique to Retrieve Vegetation Parameters

Due to the climate change crisis and a growing global population, natural resources and ecosystem stability face significant stress. To assess and manage these challenges, continuous monitoring of vegetation conditions at fine spatial resolution is essential. Leaf Area Index (LAI) is a key biophysical parameter for determining vegetation status. The Sentinel-2 (S2) optical satellites offer a great source for LAI retrieval, with five-day revisit time and fine spatial resolution of 10 meters. However, optical observations are frequently hindered by clouds which limit continuous global coverage. To overcome this limitation, spaceborne Global Navigation Satellite System Reflectometry (GNSS-R) technology offers an all-weather complementary source as an alternative. GNSS-R is an emerging remote sensing technique involving a bistatic radar configuration that continuously collects surface-reflected signals regardless of weather conditions. The objective of this study is to explore the synergy between Cyclone Global Navigation Satellite System (CYGNSS) science data and S2 to retrieve a continuous LAI product within a machine learning framework. We utilized CYGNSS L1 v3.2 science data from low-Earth orbits, covering a latitudinal range of ±38° over the two-year period of 2022–2023, with 18 months allocated for model training and 6 months for independent testing. After masking the impact of open water, a machine learning model was developed to integrate CYGNSS-derived observables with auxiliary data to retrieve LAI. This approach leverages the high temporal density and all-weather capabilities of CYGNSS to fill gaps in S2-derived LAI, leading to improved spatiotemporal continuity in vegetation monitoring.

Keywords: GNSS-R, Sentinel-2, LAI, Vegetation, Monitoring, Machine Learning