60155

Ondaleta Coerência aplicada em série temporal NDVI/MODIS para avaliação de fitofisionomias campestres

Favoritar este trabalho

In this paper, the technique of Wavelet Coherence was used to identify patterns in time series of NDVI/MODIS for different grassland typologies and the relationship with meteorological variables, such as rainfall and air temperature. The life cycle of plant species is sensitive to variations in weather conditions and, for a climate change scenario, is important for studying the variability of climate factors in local, regional and global scales. The study covered 10 grassland typologies in Rio Grande do Sul (RS) state, Southern Brazil and a NDVI/MODIS time series with a spatial resolution of 250 m, from 2000 to 2014. The Wavelet Coherence was applied in the NDVI/MODIS data and rainfall and air temperature to check for consistency. Two well-marked cycles were seen: an annual cycle, ranging between 1 and 23 observations and an interannual cycle that varies between 92 and 184 observations. These cycles were defined and represented by the y axis in the power spectrum and were distributed throughout the series (2000-2014) on the x-axis. In the humid subtropical climate conditions prevalent in the RS, the application of Wavelet Coherence was effective in showing where and when changes in the phenological pattern of grassland vegetation occurred, as well as to point out the intensity (correlation) between NDVI data and variability of weather conditions. In the annual cycle the association between the NDVI and air temperature (high frequency) was dominant. For the interannual cycles, events such as El Niño and La Niña, which impact strongly the precipitation in RS, caused the high correlation observed between NDVI data and rainfall.