Modelagem espacial da eros?o em entressulcos e taxa de infiltra??o est?vel

• Ano de defesa: 2018
• Autor(a) principal: Moraes, Andr? Geraldo de Lima
• Orientador(a): Carvalho, Daniel Fonseca de
• Banca de defesa: Carvalho, Daniel Fonseca de, Schultz, Nivaldo, Antunes, Mauro Antonio Homem, Alves Sobrinho, Teodorico, Carvalho Junior, Waldir de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal Rural do Rio de Janeiro
• Programa de Pós-Graduação: Programa de P?s-Gradua??o em Agronomia - Ci?ncia do Solo
• Departamento: Instituto de Agronomia
• País: Brasil
• Palavras-chave em Português: Sensoriamento remoto; Simulador de chuva; Pastagens degradadas; Degraded pastures
• Palavras-chave em Inglês: Remote sensing; Rainfall simulator
• Área do conhecimento CNPq: Agronomia
• Link de acesso: https://tede.ufrrj.br/jspui/handle/jspui/4488

The M?dio Para?ba do Sul region has a long history of soil degradation due to the association of incorrect crop management, mainly coffee from the sec. XIX, pastures, rains of high erosivity and rugged relief. To guide intervention measures in the management and conservation of soil and water, it is important to gather data and information to direct the necessary efforts. In this context, erosion prediction models can be used, and despite the existence of dozens of these models, much is discussed about their predictability in different regions from which they were created, as well as the difficulty of obtaining the input data to feed them. This work aims to: characterize soil loss and steady infiltration rate considering different types of vegetation cover and soils; to evaluate if there is a relation between data obtained by remote sensing with the soil loss and the steady infiltration rate observed in the field and; to develop spatial prediction models of soil loss and steady infiltration rate from easy-to-acquire input data. 71 points were sampled with simulated rainfall in order to evaluate the variability of cover types, soils and relief of the region. As a covariate, three different sensors (RapidEye REIs, Sentinel 2A MSI and Landsat 8 OLI) were tested for remote sensing data (NDVI, SAVI, EVI, EVI2 and fraction images generated by linear spectral mixture analysis), a soil class map, chemical and physical soil attributes maps and terrain information derived from digital elevation models. Soil loss and steady infiltration rate are affected by the type of cover and soils. However, soil types affect more the steady infiltration rate than soil loss. Remote sensing data show a strong correlation with soil loss and steady infiltration rate, highlighting, among the tested, the NDVI with the stable infiltration rate and the EVI2 with the soil loss. That made possible the creation of a new soil cover factor (CEVI2 Factor). Interill erosion spatial prediction models using easy-to-acquire data (remote sensing data) present similar results to models that use data that are difficult to obtain (soil data). Steady infiltration rate spatial models using difficult-to-acquire data present the best results. However, models that use easily accessible input data have satisfactory results and have the potential to be used by users who do not have soil data.