Despite the importance of forests in regulating global climate, they are under an increasing threat by human activities. Active action against deforestation and forest degradation is critical, and is currently supported by mechanisms such as Reduced Emission from Deforestation and forest Degradation (REDD+). The implementation and effectiveness of such mechanisms relies partially on continuous observations of forests using satellite technology and partially on ground-based measurements of forest aboveground volume/biomass (AGV/AGB), carbon density and changes therein. Together, these means of forest monitoring enable the development of policies and measures to alter current trends in global forest and biodiversity loss. This thesis investigates the use of long wavelength (~23 cm, L-band) spaceborne radar, which has all-weather and canopy-penetration capabilities, acquired by the Advanced Land Observing Satellite (ALOS) for forest monitoring. Using a combination of local expert knowledge, plot inventories, and data from lidar and optical sensors, it aims to understand (1) whether forest disturbance dynamics may be detected with radar, and (2) what physical and macroecological properties influence the radar backscatter and forest AGV/AGB relation. The papers in the thesis show that radar is able to pick up forest disturbances to larger extent than traditional optical-based detection approaches, the radar to AGV/AGB relation is strongly driven by spatial scale of assessments and age- and management-related development of forest structure, and significant research towards integrating radar with other remote sensing data is urgently required. The results are important to reducing global forest loss with the best available remote sensing technologies, hence addressing a critical aspect of global environmental change.