Terra-i detects land-cover changes resulting from human activities in near real-time, producing updates every 16 days.Currently it works for the whole of Latin America and the tropics. Terra-i is a collaboration between the International Center for Tropical Agriculture (CIAT - DAPA, based in Colombia), The program on Forestry, Trees and Agroforestry (FTA) ,The Nature Conservancy (TNC, global environmental organization), the School of Business and Engineering (HEIG-VD, based in Switzerland) and King’s College London (KCL, based in the UK). The system is based on the premise that natural vegetation follows a predictable pattern of changes in greenness from one date to the next brought about by site-specific land and climatic conditions over the same period. A so-called computational neural network is ‘trained’ to understand the normal pattern of changes in vegetation greenness in relation to terrain and rainfall for a site and then marks areas as changed where the greenness suddenly changes well beyond these normal limits. Running on many computers this analysis is refreshed with new imagery every 16 days and for every 250m square of land.
Terra-i only provides data about vegetation change (gain or loss). Secondary data (i.e. spatial distribution of natural coverage, forest plantations, crop areas, among others) must be used to get more proximate results about which sort of vegetation changed or which land-use activity could be associated with the change.