Correlações de longo alcance em séries temporais de focos de calor nos biomas brasileiros

• Ano de defesa: 2014
• Autor(a) principal: COSTA, Simara Lúcia Lemos da
• Orientador(a): STOSIC, Tatijana
• Banca de defesa: CUNHA FILHO, Moacyr, FIGUEIRÊDO, Pedro Hugo de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal Rural de Pernambuco
• Programa de Pós-Graduação: Programa de Pós-Graduação em Biometria e Estatística Aplicada
• Departamento: Departamento de Estatística e Informática
• País: Brasil
• Palavras-chave em Português: Bioma; Queimada; Focos de calor; Vegetation fire; Hotspots
• Palavras-chave em Inglês: Detrended Fluctuation Analysis; Biome
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::PROBABILIDADE E ESTATISTICA
• Link de acesso: http://www.tede2.ufrpe.br:8080/tede2/handle/tede2/5349

Preserving the environment is a difficult task and depends on appropriate global and regional protectionist policies and characteristics of each region. Vegetation fires, controlled or not, cause direct damage to the environment, with environmental, social, and economic consequences. The detection of hot-pixels in Brazil was initiated by the National Institute for Space Research (INPE) in 1987, being a pioneering and the most comprehensive such initiative in the world, making use of a large number of satellites that generate hundreds of daily images used for mapping the fire outbreaks. In this work we apply the Detrendend Fluctuation Analisys (DFA) method to study correlations in daily temporal series of hot pixels detected in biomes Amazon, Caatinga, Cerrado, Atlantic Forest, Pantanal and Pampa, collected by satellites NOAA-12 and AQUA_M-T during the period 1999-2012. For all the biomes hot pixel series are found to be characterized by persistent long-term correlations, where the Pampa biome demonstrates the weakest correlation. For the Amazonas biome the entire series was analyzed in consecutive six-month periods corresponding to dry season and rainfall season for each year. The dry season (July to December) is characterized by a higher number of fires, with scaling exponent > 0;5 indicating persistent long-term correlations, while the rainy season (January to June) shows fewer fires and weaker persistency in hot-pixels time series.