Estimativa do estoque de carbono de florestas a partir de abordagem multissensor

Abstract

The Amazon rainforest heterogeneity effects on its own dynamics and structure presents a characterization and quantification challenge for forest attributes such as biomass and carbon stocks. This condition implies the necessity for more precise attributes estimation, with a higher number of in situ surveys, turning it on a more cost-intensive evaluation. These limitations prompt the search for more accurate carbon quantities estimate alternatives, supported by remote sensing tools, considering the large aboveground live biomass (AGLB) stocks presence in humid tropical forest regions. This study evaluates the remote sensing (RS) data use in carbon stocks estimative for AGLB in primary and secondary forest areas in Amazonas State. The data integration at different scales possibilities are investigated to develop a carbon stocks estimation method, subsequently comparing it with data obtained through non-invasive methods, such as GEDI missions data usage and aerial photogrammetry. This study also evaluates to which level these methods may overestimate or underestimate carbon stocks in two Amazon rainforest different physiognomies. It is expected that through RS, the AGLB carbon stocks measurement in different physiognomies can be effective and accurate for the Amazon rainforest considering its heterogeneity and regardless of the sampling intensity. Furthermore, it is anticipated that this alternative will, over its use, be validated and improved aiming to reduce data processing limitations in native forests, as current estimation techniques primarily focus on planted forests and agronomic crops as main case study.