Predicting Forest Stand Height and Canopy Cover from LANDSAT and LIDAR data using decision trees

With 57,4 % of the national territory covered by forests[7], Slovenia is one of the most forested countries of the European Union. Forests represent the most important CO2 sink for Slovenia in the framework of the Kyoto protocol to the United Nations convention on climate change, predominantly due to accumulation of forest biomass. Only one quarter of the annually accumulated aboveground biomass is harvested in the form of timber, the rest represents a CO2 sink according to the articles 3.3 and 3.4 of the Kyoto protocol. Furthermore, there is also a noticeable process of abandonment of arable land and pastures going on in Slovenia, due to the depopulation of rural areas. This leads to 0,4 % of annual forest cover increase, due to the spontaneous afforestation of abandoned agricultural areas ([2]). The forest biomass accumulation and the enlargement of forest areas are not only crucial in the global Slovenian CO2 budget, but also also important items in trading of the CO2 emission quotas. Furthermore, the accumulated forest biomass is also an important factor in potential risk of forest fire outbreaks and in forest fire behavior. Airborne laser scanning (ALS), also termed airborne LIDAR (Light Detection And Ranging), is one of many laser remote sensing techniques [5]. By measuring the round trip time of an emitted laser pulse from the sensor to a reflecting surface and back again, the distance from the sensor to the surface