Artigo recebido para publicação em 28/02/2010 e aceito para publicação em 01/06/2010 RESUMO:Este estudo avaliou a variação espaço-temporal da precipitação de 1970 a 2000
This study evaluates erosivity, surface runoff generation, and soil erosion rates for Mamuaba catchment, sub-catchment of Gramame River basin (Brazil) by using the ArcView Soil and Water Assessment Tool (AvSWAT) model. Calibration and validation of the model was performed on monthly basis, and it could simulate surface runoff and soil erosion to a good level of accuracy. Daily rainfall data between 1969 and 1989 from six rain gauges were used, and the monthly rainfall erosivity of each station was computed for all the studied years. In order to evaluate the calibration and validation of the model, monthly runoff data between January 1978 and April 1982 from one runoff gauge were used as well. The estimated soil loss rates were also realistic when compared to what can be observed in the field and to results from previous studies around of catchment. The long-term average soil loss was estimated at 9.4 t ha(-1) year(-1); most of the area of the catchment (60%) was predicted to suffer from a low- to moderate-erosion risk (<6 t ha(-1) year(-1)) and, in 20% of the catchment, the soil erosion was estimated to exceed > 12 t ha(-1) year(-1). Expectedly, estimated soil loss was significantly correlated with measured rainfall and simulated surface runoff. Based on the estimated soil loss rates, the catchment was divided into four priority categories (low, moderate, high and very high) for conservation intervention. The study demonstrates that the AvSWAT model provides a useful tool for soil erosion assessment from catchments and facilitates the planning for a sustainable land management in northeastern Brazil.
Abstract:Environmental degradation, and specifically erosion, is a serious and extensive problem in many areas in Brazil. Prediction of runoff and erosion in ungauged basins is one of the most challenging tasks anywhere and it is especially a very difficult one in developing countries where monitoring and continuous measurements of these quantities are carried out in very few basins either due to the costs involved or due to the lack of trained personnel. The erosion processes and land use in the Guaraíra River Experimental Basin, located in Paraíba state, Brazil, are evaluated using remote sensing and a runoff-erosion model. WEPP is a process-based continuous simulation erosion model that can be applied to hillslope profiles and small watersheds. WEPP erosion model have been compared in numerous studies to observed values for soil loss and sediment delivery from cropland plots, forest roads, irrigated lands and small watersheds. A number of different techniques for evaluating WEPP have been used, including one recently developed in which the ability of WEPP to accurately predict soil erosion can be compared to the accuracy of replicated plots to predict soil erosion. WEPP was calibrated with daily rainfall data from five rain gauges for the period of 2003 to 2005. The obtained results showed the susceptible areas to the erosion process within Guaraíra river basin, and that the mean sediment yield could be in the order of 3.0 ton/ha/year (in an area of 5.84 ha).
MODELING AND TEMPORAL ANALYSIS OF DYNAMICS OF SOIL USE AND OCCUPATION IN THE CUIÁ-PB RIVER BASINMODELACIÓN Y ANÁLISIS TEMPORAL DE LA DINÁMICA DEL USO Y OCUPACIÓN DEL SUELO EN LA CUENCA DEL RÍO CUIÁ-PBRESUMONas últimas duas décadas o crescimento da população acelerou o processo de ocupação humana nessas áreas ao longo dos rios, criando diversos problemas ambientais. Assim, este trabalho teve por objetivo analisar as mudanças no uso e ocupação do solo na bacia do rio Cuiá e estimar o uso e ocupação do solo para a bacia da Cuiá para o ano de 2030, utilizando algoritmo de Redes Neurais artificiais: Rede Neural Multi-Layer Perceptron. Dessa forma, foi realizado o processamento de imagens de satélites dos anos de 1998, 2001 e 2005, realizada a predição do uso do solo para o ano de 2001 baseado em Redes Neurais para a validação. Em seguida, o resultado da predição do uso do solo foi analisado pelo do índice Kappa e após obtenção do índice Kappa com um resultado de acordo com a classificação prevista por Landis & Koch (1977) foi realizada a predição do uso e ocupação do solo para 2030. Os resultados mostraram um aumento da classe Ocupadas entre os anos analisados. A modelagem dinâmica do uso do solo baseada em Rede Neural mostrou resultados satisfatórios para a bacia do rio Cuiá com acurácia de 98,64%, após 10.000 iterações e índice Kappa igual a 0,94, classificado como excelente. A previsão do uso do solo para 2030 apresentou aumento da área da classe Ocupada e uma diminuição da área com expansão.Palavras-chave: Uso do solo. Redes neurais. Geoprocessamento.ABSTRACTIn the last two decades the population growth accelerated the process of human occupation in these areas along the rivers, creating several environmental problems. Thus this study aimed to analyze the changes in land use and occupation in the Cuiá river basin and to estimate the land use and occupation for the Cuiá basin for the year 2030 using an artificial neural network algorithm: Multi-Layer Perceptron Neural Network. In this way, the satellite image processing of the years of 1998, 2001 and 2005 was carried out, with the prediction of the use of the ground for the year 2001 based on neural networks for the validation, followed by the prediction of the land use was analyzed through the Kappa index and after obtaining the Kappa with a result according to the classification predicted by Landis & Koch (1977) the use and occupation of the soil was predicted by 2030. The results showed an increase of the Occupied class among the analyzed years. The dynamic modeling of soil use based on Neural Network showed satisfactory results for the Cuiá River basin with accuracy of 98.64%, after 10,000 iterations and Kappa equal to 0.94, classified as excellent. Prediction of land use for 2030 showed an increase in the area of the occupied class and a decrease of the area with expansion.Keywords: Soil use. Neural network. Geoprocessing.RESUMENEn las últimas dos décadas el crecimiento de la población aceleró el proceso de ocupación humana en esas áreas a lo largo de los ríos, creando diversos problemas ambientales. Entonces este trabajo tuvo por objetivo analizar los cambios en el uso y ocupación del suelo en la cuenca del Río Cuiá y estimar el uso y ocupación del suelo para la cuenca de Cuiá para el año 2030 utilizando algoritmo de Redes Neurales Artificiales: Red Neural Multi-Layer Perceptron. De esta forma, se realizó el procesamiento de imágenes de satélites de los años 1998, 2001 y 2005, realizada la predicción del uso del suelo para el año 2001 basado en Redes Neurales para la validación, luego el resultado de la predicción del uso del suelo se analizó por el índice Kappa y tras la obtención del índice Kappa con un resultado de acuerdo con la clasificación prevista por Landis & Koch (1977) se realizó la predicción del uso y ocupación del suelo para 2030. Los resultados mostraron un aumento de la clase Ocupados entre los años analizados. El modelado dinámico del uso del suelo basado en Red Neural mostró resultados satisfactorios para la cuenca del Río Cuiá con exactitud del 98,64%, después de 10.000 iteraciones e índice Kappa igual a 0,94, clasificado como excelente. La previsión del uso del suelo para 2030 presentó aumento del área de la clase Ocupada y una disminución del área con expansión.Palabras clave: Uso del suelo. Red neurales; Geoprocesamiento.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.