Modelos estocásticos da dinâmica da paisagem florestal e simulação de cenários para o estado do Rio Grande do Sul no período de 1988 a 2020

• Autor(a) principal: Hendges, Elvis Rabuske
• Data de Publicação: 2007
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Manancial - Repositório Digital da UFSM
• dARK ID: ark:/26339/001300000jwnz
• Texto Completo: http://repositorio.ufsm.br/handle/1/3709
• topic: Sensoriamento remoto; Uso da terra; Sistema de informações geográficas; Remote sensing; Land use; Geographic information system; CNPQ::CIENCIAS AGRARIAS::RECURSOS FLORESTAIS E ENGENHARIA FLORESTAL

Land use patterns and the potential careless use of our planet s natural resources have recently called the attention of an increasing number of studies in the area of scenario simulation in several parts of the world. The present research aims at simulating the forest scenario of Rio Grande do Sul, Brazil, in order to identify and quantify in a map the forest s evolution in this state for the year 2020. Since 2003 Rio Grande do Sul has been encouraging forestation practices trough its State Forestation Program as a new economic alternative; but the program s development has also raised controversies over the possibility of an environmental and economic unbalance caused by the excess of forested areas. On the other hand, data collected in Forest Inventories had already revealed an increase in native and exotic forest areas before the program was created and, in this sense, the projection proposed in this study is based on the development of a stochastic simulation model of Rio Grande do Sul s forest scenario without considering the forested areas that emerged from the forestation program. Land use patters will be classified based on images from the LANDSAT 5 and LANDSAT 7 satellites referring to the years 1988 and 1998 as well as images from the TERRA satellite referring to the year 2007. Based on a cross-check of forest use patterns (native forest, bush land and reforestation with exotic species) against the other types of land use (cropped farmland, grassland, water, dune, rocky outcrop, wetland) identified in the images, a new map was generated showing the forest s evolution in Rio Grande do Sul over the last 20 years. This map was then cross-checked with several thematic maps containing environmental, economic and social information. The forest evolution map and its cross-check with the thematic maps defined the Markov transition matrix and the weights of evidence, respectively, both requirements for the calibration of the stochastic model. As a result, the simulation of Rio Grande do Sul s forest landscape for the year 2020 showed an increase in almost all regions of the state of the bush land, native forest and reforestation with exotic species areas, given that the latter two increased by 5,113 km2 and 4.036 km2 respectively. From this growth the most significative is the exotic species one, which represents 72% in relation do 2007, even without taking into account and simulating those areas boosted by the State Forestation Program.