Paper accepted by IEEE J-STARS: Mapping annual land use and land cover changes using MODIS time series. Doctoral researcher He Yin et.al., geomatics Lab.
This manuscript aims to characterize changes between land cover/use types based on annual time series of machine learning derived land cover probabilities. We used the temporal segmentation algorithm MODTrendr to identify trends and changes in the probability time series that were associated with land cover/use conversions. Four land conversion types were mapped and tested: conversions between forest and grassland, conversions between cropland and grassland. The results reveal good performance of our approach - though the mapping accuracy varies among different change classes. We presented for the first time an approach that combines probabilities derived from machine learning with time series based analysis for land cover/use change analysis at MODIS scale.