Abstract
This paper aims to assess vegetation changes on the forest canopies of Spain over the period 2000- 2009. A multi-resolution analysis (MRA) based on the wavelet transform (WT) has been implemented to gross primary production (GPP) and NDVI and EVI MODIS data sets. This analysis decomposes a non-stationary signal in different temporal scales. In order to derive significant changes, the Mann-Kendall non-parametric test and the Sen's method have been applied to the trend component derived from the MRA. These methods provide the magnitude and direction of the vegetation changes observed in the temporal series. Comparing the results with recent findings it is shown that vegetation change detection from temporal series analysis is highly dependent on the studied period and the starting date.
