Multi-attribute decision-making model based on Remote Sensing to enhance the environmental auditing of Sustainable Forest Management in the Brazilian Amazon forest

Abstract

Sustainable Forest Management (SFM) plans are distributed across 2% of the Brazilian Legal Amazon (BLA) territory. Although these enterprises are authorized through the national system (Sinaflor), there is a lack of a computational platform capable of processing, analyzing, and correlating monitoring data from remote sensing techniques with the production data declared in the system. By combining multi-attribute decision-making (MADM) methods with geographic information system (GIS) data and DETEX satellite image Linear Spectral Mixture Model (LSMM) processing on Google Earth Engine (GEE), we developed a semi-automatic system that calculates an Environmental Pressure Index (EPI). This index, composed of categories of cost and benefit indicators, measures the environmental impacts of SFM operations and the surrounding land use and land cover (LULC) dynamics that can externally stress these enterprises. To test this framework, we evaluated 15 SFMs in operation between 2022 and 2023 in the highly environmentally stressed AMACRO region of the BLA. The evaluation was performed in three tiers, proving that cost indicators developed to measure environmental impacts caused by SFMs are more consistent for the selection of enterprises to be audited than LULC dynamics indicators, as they combine planned and declared production data with the classification of intervention shown by DETEX images. The developed semi-automatic system have the potential to be used by environmental agencies as a decision-making tool for selecting SFMs to be audited, as it provides a quantitative approach based on index calculations and can be easily adapted for specific auditing purposes.