A New Method for the Prediction of Carbon Sequestration in Reforested Areas Using a Fuzzy-ART-BP Neural Network

摘要

Global emissions of carbon dioxide the main cause of greenhouse effects and, as consequence, global warming, endanger the lives of all living species on the planet. Thus, it becomes essential to adopt measures to reduce carbon emissions, aiming to environmental sustainability and also the development of efficient methods for quantifying the flow of carbon into the atmosphere. Therefore, this paper presents an intelligent system to quantify emissions and carbon sequestration in reforestation areas, medium and long term. The proposed system consists of a combination fuzzy-ART neural network architecture and a multilayer feedforward training based on the backpropagation algorithm. Aiming to test the proposed system, we present an application in-an area located in a reforested area in the Amazon region of Mato Grosso-Brazil, on a farm with land area of approximately 8939 hectare.