Determinación de la mejor clasificación de coberturas obtenidas con imágenes satelitales de radar, ópticas y su sinergia en el periodo 2015 – 2020 e implementación de un modelo de predicción de la pérdida de bosque al año 2030 en seis municipios de la Amazonia colombiana

Abstract

In order to determine the areas affected by the phenomenon of deforestation and changes in land use, it is possible to use some tools that allow the development of these and other use, it is possible to make use of some tools that allow the development of these and other types of studies without the strict need of field reconnaissance, through techniques and studies without the strict need of field reconnaissance, by means of techniques and digital image digital image processing, obtaining very accurate results that reflect reality. In this project, by means of a multitemporal analysis of satellite images of two different types of remote sensing sensors sensors such as optical (passive), radar (active) and an optical-radar synergy, it is identified optical-radar synergy, the deforestation and erosion zones of the Colombian Amazon region are identified using classifications supervised by the Random Forest algorithm and following the Corine Land classification following the Corine Land Cover classification methodology. In addition, the results of these classifications are the results of these classifications with the Kappa index, thus establishing which of the three methods of capturing information is more efficient in this which is more efficient in this specific application. Affected areas and soil changes are determined in the municipalities of La Macarena in the department of Meta, San Vicente del Meta and San Vicente del Meta in the department of Meta. of the department of Meta, San Vicente del Caguán and Cartagena del Chairá of the department of Caquetá, San José del Guanajuato, San Vicente del Caguán and Cartagena del Chairá. Caquetá, San José del Guaviare and Calamar in the department of Guaviare. Validating the results information taken from official and specialized sources in the Amazon region. Finally, after obtaining the results, a cellular automaton mathematical model is applied, taking into account significant variables from model is applied taking into account significant variables from official sources. the Dinamica Ego tool, with transition matrices and weights of evidence that support the multitemporal support the multitemporal analysis and allow the identification of possible zones at risk of being affected by deforestation. with the phenomenon of deforestation. Translated with www.DeepL.com/Translator (free version)