Comparação entre as redes neurais artificiais e o método de interpolação krigagem aplicados à pesquisa agronômica

Pós-graduação em Agronomia (Energia na Agricultura) - FCA

Interpolação geoespacial da evapotranspiração de referência (ETo) em regiões com escassez de dados: estudo de caso no Sul de Minas Gerais, Brasil

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Spatial-temporal distribution of fire-protected savanna physiognomies in Southeastern Brazil

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Batimetria fluvial estimada com dados orbitais: Estudo de caso no alto curso do rio Paraguai com o sensor ASTER

The Paraguay River is the main tributary of the Paraná River and has an extension of 1.693 km in Brazilian territory. The navigability conditions are very important for the regional economy because most of the central-west Brazilian agricultural and mineral production is transported by the Paraguay waterway. Increased sedimentation along the channel requires continuous dredging to waterway maintenance. Systematic bathymetric surveys are periodically carried out in order to check depth condition along the channel using echo-sounding devices. In this paper, digital image processing and geostatistical analysis methods were used to analyze the applicability of the ASTER sensor to estimate channel depths in a segment of the upper Paraguay River. The results were compared with field data in order to choose the band with better adjustment and to evaluate the standard deviation. Comparing the VNIR bands, the best fit was presented by the red wavelength (band 2; 0,63 - 0,69 μm), showing a good representation of the channel depths shallow than 1,7 m. Applying geostatistical methods, the model accuracy was enhanced from 43 cm to 36 cm and undesired components were slacked. It was concluded that the digital number of band 2, converted to bathymetry information allows a good estimation of river depths and channel morphology.

Validação de sistema de amostragem para monitoramento de formigas cortadeiras atta spp. (Hymenoptera: formicidae) em plantações de eucalipto

Pós-graduação em Ciência Florestal - FCA

Modelagem da incerteza da distribuição espacial de fósforo e potássio do solo sob a cultura canavieira

Pós-graduação em Agronomia (Ciência do Solo) - FCAV

Análise multivariada de dados espaciais na classificação interpretativa de solos

Pós-graduação em Agronomia (Energia na Agricultura) - FCA

Suscetibilidade magnética e sua associação com atributos do solo e do relevo em uma topossequência de Latossolos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Mapeamento geoestatístico da vulnerabilidade do Aquífero Guarani no município de Ribeirão Bonito - SP

The Guaraní aquifer has relevant importance both as a source of water for several urban centres and the development of agriculture and livestock. In recharge areas the aquifer is free and, therefore, subject to contamination of effluents and tailings deposited on soils that cover it. Thus, it becomes crucial not only its protection at all levels, as the knowledge of its degree of natural vulnerability. The present work used geostatistics modeling techniques to study the natural vulnerability of the Guaraní aquifer in the city of Rio Bonito, State of São Paulo, southeastern Brazil, where the Guarani aquifer is exposed. These techniques, extensively used in evaluation studies of mineral deposits and oil tanks, can be adapted to produce a spatial classification or a regionalisation of probabilistic indices of vulnerability. By ordinary kriging method maps of vulnerability classification were obtained. To determine the vulnerability of the aquifer was employed the Aquifer Vulnerability Index (AVI), which requires knowledge of unsaturated zone thickness and permeability. The final product was a map with probabilistic index of vulnerability of the Guaraní aquifer, which presented values between 0 to 0.33 years, framing the area studied in AVI class extremely high vulnerability

Mapeamento geoestatístico da permeabilidade superficial no entorno do Ribeirão Claro: e simulação da possibilidade de sua contaminação

The evaluation of soil permeability throughout the weathering profile is one of the most important features to be considered in environmental studies. This study, developed from field testing and analysis of data obtained by geostatistical methods, aims at mapping the permeability around the Ribeirão Claro river. The intent is to simulate an accident with toxic liquids where soil permeability is of fundamental importance. Another purpose of the research was to determine the minimum time that, in the event of an accident, a possible contaminant to reach the water table level and be routed to the nearest drain, in this case, the Ribeirao V Claro river constitutes fundamental information. The studied area of approximately 4 km² is located within the UNESP-Rio Claro campus, consisting of colluvial soil from Fm. Rio Claro superimposed on residual soil of Fm. Corumbataí. The method used to determine the permeability is the concentric cylinders performed on a sampling grid with 64 points containing spacing of 5 meters EW and 10 meters NS. In the places of permeability tests were collected samples for laboratory determination of the percentage of fines. From particle size analysis was performed and analysis statistical and geostatistical on this data. The histogram was based on the statistical studies, and the semivariograms were based on geostatistical estimation methods. Based on the comparison between the maps and the data obtained, it was determined that the percentage of fines in colluvial surface soil has little influence on permeability, which the proximity to the Ribeirao Claro river, the eastern portion, a factor that influences the distribution of permeability values