Análise de dados orbitais e suborbitais no diagnóstico e mapeamento de áreas prioritárias para conservação no corredor da Chapada dos Guimarães

Abstract

Cerrado areas are gradually being reduced by several factors, whether due to agricultural production, biological invasions and even due to the suppression of fire. In view of the spatiotemporal changes in the landscape, the use of remote sensing is widely processed in the discrimination of plant types, and allied to different classification techniques provide results that enable public policies for recovery and extensive monitoring of plant areas. The mapping of plant formations is limited as sensor solutions, whereas RPAS technologies are defined as more advantageous, seen as scales of detail when compared to traditional and manned airborne platforms. The general mapping by Digital Image Processing (PDI) is carried out by classifying algorithms “pixel by pixel” or by “regions”, obtaining the best results via application of classifiers by regions, where the Analysis method stands out Object-Based Image Analysis (OBIA). Therefore, this work aimed to integrate the mapping techniques via orbital data and Remote Piloted Aircraft Systems (RPAS) supported by field data collections. Chapter 1 covers the theoretical foundation addressing concepts regarding the importance of Wetlands (AUs) in the Cerrado as priority areas for the conservation and applicability of sensors to this study. Chapter 2 deals with the classification of three Key Biodiversity Areas (KBAs) in the discrimination of footpath and other vegetation types in the Cerrado, aided by Remote Sensing techniques and Geographic Information Systems, integrated with Boolean with different information plans. The “vereda” class was manually delimited in the “Wetland” typology from field recognition, totaling 157.39 ha. The classification obtained reached a Kappa index of 0.92. It is concluded that work was able to identify possible places of presence of habitats of interfluvial footpaths, however, it needs to improve the criteria for crossing information plans, covering as types of existing footpaths. Chapter 3 offers a focal study of a 20 ha footpath inserted in the Chapada dos Guimarães National Park submitted to Integrated Fire Management (MIF) being monitored by a multi-engine RPAS, recovered after the data processing stage obtained in the missions, three orthomosaic and digital models that were obtained in the laboratory and validated in the field. The products generated with the data obtained by the RPAS made it possible to diagnose the native and invasive vegetation of the area as to the response to the prescribed burning, and it can be inferred that the fire acted as a renewer of the vegetation and inhibited the growth of invasive fern species in the same place. The phytosociological survey that the footpath is in the process of maturation due to the ecological group as to which the inventoried species belong. Kappa greater than 0.8 was obtained for all classifications as generated classifications. The location of AUs and later of paths, are financing management tools for institutions such as CEPF and research for organizations such as INAU where they develop technical recommendations that guide the installation of enterprises and licensing for rural regularization.