Illicit Crop Monitoring based on Sentinel-2 data
The third challenge identified for the Geospatial Intelligence domain (driven by the EU Satellite Centre, SatCen) focuses on the analysis of multispectral satellite image time series, with the aim of identifying spectral signatures of various types of land cover classes, particularly vegetation types. The main application domain is the remote sensing and long-term monitoring of illicit crop production, considered as a cheaper but effective alternative to on-field surveying.
Vegetation cuts for coca cultivation, identified in false-colour infrared from satellite imagery ahead of ulterior field evidence in August 2018
The main data sources considered are the Sentinel-2 constellation (provided by SatCen) which records land surface conditions over wide regions and with a high frequency (5-10 days). To solve this challenge BETTER will deliver time series data obtained from the entire available Sentinel-2 archive on selected areas of interest. Analysis-ready time series for illicit crop monitoring is of value for governmental institutions and international organizations (e.g., the United Nations) that make efforts to counter illicit crop cultivation. These efforts require a constant flow of information, reliable monitoring, and the possibility of extracting periodic reports from measurements and observations. The solution implemented by BETTER can also benefit a wider range of entities that are interested in monitoring vegetation changes based on time series extracted from satellite images for other applications.