Reliable data on deforestation patterns and their spatial and temporal variability are of highest importance as basis for the development of efficient and precise counter-measures for affected regions and countries. In this context, revealing direct and underlying causes of deforestation are a fundamental prerequisite for the implementation of REDD+.
While the development of reference scenarios for deforestation is currently pushed by several countries, there is still enormous need for revealing causes of deforestation on subnational and local level in relation to current and historical land use and related socio-economic and institutional stakeholders (LAMBIN et al. 2001).
Further on, many countries with net-deforestation on national level frequently exhibit strata with increasing forest cover on subnational levels, which have to be considered in integrated and effective policy-approaches. The project contributes improving efficiency of global and national forest related policies and helps identifying potential leakage-effects.
For this purpose, we aim at combining historical time series of satellite images of lower spatial resolution with recent high resolution images, in order to distinguish the most important forestry, agroforestry and agricultural land-use types within their landscape context.