Uso do pacote R sits no mapeamento de uso e cobertura da terra da Reserva Extrativista Marinha Araí-Peroba/Pa

Abstract

The Marine Extractive Reserve RESEX Araí-Peroba, located in the municipality of Augusto Corrêa, in the state of Pará, is a conservation unit that shelters traditional communities engaged in extractive activities, agriculture, and small animal husbandry for subsistence. With the aim of spatially analyzing land cover and the evaluation of R sits in the reserve to assist in monitoring and maintaining the local flora, this work generated a land use classification of the RESEX, through the creation of a cube of orbital images regularized by geoprocessing techniques and R programming language. To do so, orbital images from the sensor on the Sentinel satellite Level 2A were used, along with wide use of spectral bands and spectral indices (NDVI and EVI) and Linear Spectral Mixture Model (MLME) to aid in the identification of features on the ground. In order to carry out the task, the generation of a supervised classification was performed through the creation of a regularized image cube, utilizing the R sits package implemented in the R software. Based on the generated data and conducted analyses, a substantial amount of transition areas totaling approximately 5.5% of the mapped area of the RESEX was observed. The R sits package becomes a promising tool as it brings updates to image catalogs and deep learning algorithms for the analysis of time series of images, as well as specific parallel processing methods for data cubes that guarantee efficient performance, denoting flexible processing with shorter processing time, enabling statistical and spectral analyses for product generation. Therefore, geoprocessing techniques serve as significant tools in assisting land observation, environmental monitoring, and the REDD+ policy through carbon credit.