Improving Environmental Monitoring Data Collection in the Amazon by Remote Sensing: A Comparative Analysis of Cloud Detection Models for Sentinel-2

Remote sensing is essential for environmental monitoring in the Amazon, playing a crucial role in conservation initiatives in forested areas such as REDD+, especially in the process of monitoring, reporting, and verification (MRV) of vegetation cover and CO2 emissions. However, the presence of clouds in areas of interest has impacted the effectiveness of this tool. As a response to cloud cover challenges, several cloud detection models have been developed, aiming to improve the efficiency and quality of information regarding CO2 emissions and deforestation in these regions. This article presents a comparative analysis of four cloud detection models applied to imagery acquired by the Sentinel-2 satellite in the Amazon region. The methodologies, strengths, limitations of each model, and the accuracy in the tested images are discussed, aiming to promote greater precision and reliability of the environmental monitoring strategy through remote sensing in the Amazon region.