USO DE IMAGENS DE RADAR NA ESPACIALIZAÇÃO DA TEMPERATURA DO AR

Abstract

The objective of this work was to develop mathematical models to estimate air temperatures (minimum, medium and maximum), monthly and annual in the state of Espírito Santo, Brazil, and spatializing information based on a digital elevation model (MDE), obtained through radar data from SRTM (Shuttle Radar Topography Mission). The equations were developed with the use of multiple regression analysis, adopting the linear model, which had as independent variables the altitude, latitude and longitude as the dependent variable and the air temperature. The coefficients of the equations were determined using as the base the monthly average air temperatures for the period from 1977 to 2006, obtained on fourteen meteorological stations distributed throughout the state. The SRTM radar image of the original mission of mapping the topography land developed by NASA (National Aeronautics and Space Administration) and NGA (National Geospatial-Intelligence Agency) of the United States in 2000, allows for the provision of a digital elevation model (MDE)with spatial resolution of 90 m to the area of coverage of the study. The images SRTM were mosaics, resulting in a single image with average values of altitude and the geographic coordinates for the entire state. The statistical analysis of regression have shown the level of adjustment of the data obtained by the linear statistical model, which showed adjusted coefficient of determination (R2) ranging between 0.94 to 0.97, with good performance to estimate the air temperature in places where it is not measured. The use of SRTM identified very sensitively the variation of relief, demonstrating efficiency and estimate the results of spatial air temperature, helping to improve the results in relation to simple interpolation of the measured data on meteorological stations, especially in regions of bumpy relief.