We proposed a new methodology based on NDVI time series autocorrelation values and machine learning algorithms to assess fallowing temporal patterns across rainfed agricultural areas that was tested in mainland Spain. We found that approximately half of rainfed agricultural areas were Regularly Cultivated (RC) and distributed mainly in the north- western Spain whereas lands with crop-fallow rotation patterns every two (CF-2) and three years (CF-3) were distributed across northeast, center and south of Spain. You can see the spatial distribution in the map viewer below. Results were published in Recuero et al., 2019.
Category Archives: Maps
How are vegetation growing seasons around the world?
We identified the most relevant vegetation seasonal patterns based on spectral analysis of NDVI time series. Most of global land surface (93.18%) showed one intra-annual growing season whereas bi- and trimodal patterns occupy lower area (5.58%), mainly in tropical and arid regions along with agricultural areas. The highest values of amplitude and stability were found at Northern Hemisphere high latitudes whereas lowest values were located in tropical and arid regions. High presence of pluri-annual cycles was found in water-limited areas of Australia. See more in this article.
