Diagnóstico de áreas de pastagem degradada e correlação com a qualidade física do solo no município de Veríssimo-MG
Carregando...
Arquivos
Data
2016-07-29
Autores
Rosa, Aline Rombega Tito [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
As pastagens no Brasil ocupam mais de 172 milhões de hectares, no qual 50% deste total já se encontram em alguma fase de degradação, com perda de matéria orgânica do solo e consequente elevação das emissões de carbono para a atmosfera. Neste sentido, informações quantitativas dos níveis de degradação das pastagens e do ambiente podem apoiar políticas de gestão ambiental. Este estudo teve o intuito de avaliar a aplicabilidade e eficácia do uso do sensoriamento remoto e geoprocessamento no diagnóstico de áreas de pastagens degradadas e sua correlação com a qualidade física do solo, na bacia do rio Uberaba, Veríssimo – MG. Os dados de sensoriamento remoto foram analisados utilizando três modelos matemáticos de regressão, sendo dois não lineares com distribuição ondulatória e um linear. O modelo de regressão cúbica, a partir das séries temporais do Moderate Resolution Imaging Spectroradiometer (MODIS), descreveu melhor o comportamento médio do pixel ao longo do ano, elucidando em média 75,7% dos dados válidos, apresentando inexistência de perda amostral em decorrência dos efeitos atmosféricos; principalmente presença de nuvens. Em todas as áreas de pastagens avaliadas foi observada densidade do solo acima dos valores críticos descritos na literatura, o que confere condições restritivas ao crescimento e desenvolvimento do sistema radicular, à infiltração e ao transporte de água, bem como às trocas gasosas entre o solo e a atmosfera. Ainda, foi evidenciada redução da proporção de macro microporos, comprometendo a adequada aeração, permeabilidade e armazenamento de água. Portanto, os resultados apontam que o método de diagnóstico das áreas de pastagens a partir de séries temporais do MODIS MOD13Q1 foi mais adequado na determinação do padrão de comportamento do pixel ao longo do ano; as áreas de pastagens do município de Veríssimo-MG apresentaram padrões espectrais distintos, permitindo a classificação e correlação com os níveis de degradação; as análises dos atributos físicos revelam que todas as áreas avaliadas apresentam sinais de degradação, inclusive as áreas com presença de pasto sadio; os solos das diferentes fisionomias de pastagens apresentaram intenso processo de compactação, com redução da porosidade, decorrente principalmente do efeito do pisoteio do gado.
Pastures in Brazil occupy more than 172 million hectares, in which 50% are already in some stage of degradation with loss of soil organic matter and the consequent rise in carbon emissions to the atmosphere. Thus, quantitative information of the degradation levels of pastures and the environment can support environmental management policies. This study aimed to evaluate the applicability and effectiveness of remote sensing and GIS use in the diagnosis of degraded pastures areas and its correlation with the soil physical quality in the basin of Uberaba river, Veríssimo - MG. The remote sensing data were analyzed using three mathematical regression models, two nonlinear wave with a linear distribution and linear .The cubic regression model, from temporal series of Moderate Resolution Imaging Spectroradiometer (MODIS), best described the average behavior of the pixel throughout the year, elucidating on average 75.7% of the valid data, showing absence of sample loss due to atmospheric effects; mainly the presence of clouds. In all areas evaluated pasture was observed soil density above the critical values reported in the literature, which gives restrictive conditions for growth and root development, infiltration and water transport and gas exchange between soil and atmosphere. Furthermore, there was a decrease in the proportion of macro micropores, compromising the proper aeration, water permeability and storage. Therefore, the results indicate that the diagnostic method of pasturelands from MODIS MOD13Q1 the time series was more appropriate in determining the pixel pattern of behavior throughout the year; the areas of Veríssimo-MG pastures showed distinct spectral patterns, allowing classification and correlation with the levels of degradation; the analysis of physical attributes reveal that all evaluated areas show signs of degradation, including areas with presence of healthy pasture; the soils of different pastures physiognomies showed intense process of compaction, reducing the porosity, mainly due to the effect of cattle trampling.
Pastures in Brazil occupy more than 172 million hectares, in which 50% are already in some stage of degradation with loss of soil organic matter and the consequent rise in carbon emissions to the atmosphere. Thus, quantitative information of the degradation levels of pastures and the environment can support environmental management policies. This study aimed to evaluate the applicability and effectiveness of remote sensing and GIS use in the diagnosis of degraded pastures areas and its correlation with the soil physical quality in the basin of Uberaba river, Veríssimo - MG. The remote sensing data were analyzed using three mathematical regression models, two nonlinear wave with a linear distribution and linear .The cubic regression model, from temporal series of Moderate Resolution Imaging Spectroradiometer (MODIS), best described the average behavior of the pixel throughout the year, elucidating on average 75.7% of the valid data, showing absence of sample loss due to atmospheric effects; mainly the presence of clouds. In all areas evaluated pasture was observed soil density above the critical values reported in the literature, which gives restrictive conditions for growth and root development, infiltration and water transport and gas exchange between soil and atmosphere. Furthermore, there was a decrease in the proportion of macro micropores, compromising the proper aeration, water permeability and storage. Therefore, the results indicate that the diagnostic method of pasturelands from MODIS MOD13Q1 the time series was more appropriate in determining the pixel pattern of behavior throughout the year; the areas of Veríssimo-MG pastures showed distinct spectral patterns, allowing classification and correlation with the levels of degradation; the analysis of physical attributes reveal that all evaluated areas show signs of degradation, including areas with presence of healthy pasture; the soils of different pastures physiognomies showed intense process of compaction, reducing the porosity, mainly due to the effect of cattle trampling.
Descrição
Palavras-chave
Degradação, NDVI, Sensoriamento Remoto, Degradation, Remoto Sensing