Temporally-dense multi-source satellite remote sensing for advancing the monitoring and characterization of tropical forest disturbances

Summary

Tropical forests are pivotal for climate regulation, biodiversity, water regulation, and soil conservation. Despite their critical importance, tropical forests have faced high rates of forest loss, a pattern associated with increased natural and human-induced disturbances over the past decades. Accurate and timely large-scale information on forest disturbances is crucial for local governments and local communities in tropical regions, seeking to protect forests and reduce illegal and unsustainable activities. Remote sensing has been established as a prime tool for monitoring large-scale forest disturbances in the tropics. This thesis investigates the potential of temporally-dense multi-source remote sensing data to improve the detection and characterization of tropical forest disturbances. This work explores the benefits of SAR-based textural features, the combination of multi-wavelength SAR data, the characterization of fire-related disturbances using optical and SAR data, and the correlation between forest disturbances and precipitation dynamics.