Avaliação da série temporal de dados de temperatura da superfície derivadas do sensor MODIS
This paper describes a construction of time series using land surface temperature product (MYD11A2) from MODIS sensor aboard the Aqua platform. The main problem encountered was the many failures that occur even in the eight day compositions. Gaps occur much more in nighttime than in daytime temperature products. Then several interpolation methods were tested to fill these gaps. The best way to fill the gaps was to first build the 16-day compositions similar to the existing EVI2 time series in LAF website and the most appropriate interpolation would be linear interpolation. It was also analyzed some samples of different types of land use and land cover, such as water, urban area, forest, agriculture, cerrado, wetland and deforestation using the time series of temperature constructed in this work. As expected, the variations between daytime and nighttime temperatures are lower in water and higher in agricultural and urban areas. Regarding the analysis of the samples was found a tendency of temperature increase during the period analyzed.