Overall objective of this study is to generate an up-to-date land-use/ land-cover map for Jambi Province, Indonesia. Prior attempts are either outdated or cover only small extents. Based on a large amount of recent satellite data, freely available in online archives, a map with 10m spatial resolution should be produced for 2018/19.
Due to the demanding biophysical conditions of the study region (e.g. dense cloud cover and rough mountainous terrain), neither optical nor radar images provide the ideal solution. Hence, a combination of Sentinel-1 and Sentinel-2 data is used to mitigate these problems. In combination with high resolution reference data a model is trained on Sentinel data to predict land-use/ land-cover classes. To achieve this an enormous of image data is processed and combined on a cloud computing platform (Google Earth Engine).
In a second step this model should be applied on images from previous years. This will allow to track changes in land-use composition, giving ideas e.g. about regional trends on land-use intensification.
An up-to-date land-use/ land-cover map with appropriate resolution and accuracy could then be used in other modules for further research in combination with additional data.