Monitoring winter vegetation cover using multitemporal modis data

Image time series from the medium resolution sensor EOS/MODIS are frequently used to monitor vegetation phenology at regional or global scales. Such data have not previously been assessed to follow vegetation cover in winter in cropping areas, although leaving soils bare lead to environmental degradation. In this case study we used MODIS imagery from four consecutive winter seasons (2000/01-2003/04) in an intensive farming region to highlight areas with recurrent patches of bare soils, which can be considered as "hot spots" from an environmental point of view. We evaluated the estimation of bare soil surfaces from MODIS data in using SPOT HRVIR images and ground truth observations on four pilot sites selected in the study region. High correlations were observed between bare soil proportions estimated with MODIS data and SPOT HRVIR images. Furthermore, first results have highlighted the ability of MODIS to detect bare soil areas depending on the dispersion of bare soil fields in the landscape. Hence, this study demonstrated the usefulness of MODIS data for detecting and monitoring bare soil areas during winter at a regional scale. It is believed that this approach of using medium resolution data to derive bare soil maps, when validated with reference data sets, would help to change land management practices.