Análise estatística e geotecnologias no estudo dos indicadores de desertificação nos “Cariris Velhos”-PB

• Ano de defesa: 2015
• Autor(a) principal: Macêdo, Mônica Larissa Aires de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Geografia; Programa de Pós-Graduação em Geografia; UFPB
• Palavras-chave em Português: Análise Multivariada; Imagens Landsat TM; Vegetação; Temperatura da Superfície; Caatinga; Degradação; Multitemporal Change; Vegetation; Surface Temperature.; Degradation; Multivariate Analysis; CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/tede/9807

The practices of soil usage and occupation generate severe deterioration processes and in semiarid areas may result in desertification. In Paraiba, one of the areas representing this process is the one named “Cariris Velhos”. Through a multitemporal analysis, aided by statistical techniques and geotechnologies, it was possible to carry out some analysis and to diagnose the desertification state. Descriptive analysis, non-parametric test and multivariate statistics techniques were used. The precipitation was used and revealed three important patterns that explain about 70% of data variability, exposing how the rainfall occurred in the years of the images, making visible the rainy occurred, drought and transitioning periods. It was possible to confirm the edaphoclimatic frailty existent and the strong correlation with the climate anomalies (El Niño and La Niña) caused by meteorological systems. The second aspect to be analyzed was the vegetation coverage, defined in different strata and surface temperature. Three timed clippings were selected to compose these aspects: images from the 80’s (1989, 1985), from the 90’s (1990 and 1995) and of the 2000 years (2004,2005 and 2007). The images processing was done with part of algorithm SEBAL which were determined SAVI, IAF and Surface Temperature. SAVI and IAF were obtained, however, owing few studies with IAF for semiarid was decided to used. The results have shown that the vegetation strata diminished gradually, from a moderate stage to a severe one, with coefficient of variation (56%, 78,75% e 84%) unrepresentative, proving the diversity of landscapes. A significant amount of municipalities, on the last clipping, already presented, in its majority, more than 50% of values of the pixels presented in the classes of exposed soil, that is, with accentuated deterioration, proving strong evidence of desertification. With the deforestation, the surface temperature increased significantly configuring a higher number of pixels presented in classes superior to 30ºC. This factor, allied to the rainfall, hampers the development and spreading of species, since seed dispersal is by anemochory, diminishing the biodiversity of the caatinga. Concerning the agricultural, economic and social aspects, three profiles were revealed for each clipping, showing the most important indicators on each one. With this, it was observed that between 1980 and 1995, the population was more dependent of the land, in a way that when there was a worsening in the physical plan, a diminishing in the rainfall, the production also diminished, directly affecting social and economically the population. Between 2000 and 2010, the governmental incentives contribute to an improvement of these aspects so that they did not contribute, not directly, for the process of desertification, configuring a smaller impact (weight) on the analysis. In this case, the deterioration is originated from the agricultural aspect, being an example the incentive to the goatrearing and wood extraction to the plaster polo of Pernambuco in some municipalities of the “Cariris Velhos” region.