Modelling patterns and drivers of post-fire forest effects through a remote sensing approach
Forests play a significant role in the global carbon budget as they are a major carbon sink. In addition to the deforestation caused by human activities, some forest ecosystems are experiencing detrimental changes in both quantity and quality due to wildfires and climate change that lead to the heterogeneity of forest landscapes. However, forest fires also play an ecological role in the process of forming and functioning of forest ecosystems by determining the rates and direction of forest stand recovery. This process is strongly associated with various biotic and abiotic factors such as: the disturbance regimes, the soil and vegetation properties, the topography, and the regional climatic conditions. However, the factors that influence forest-recovery patterns after a wildfire are poorly understood, especially at broad scales of the boreal forest ecosystems. The study purpose of this research is to use remote sensing approaches to model and evaluate forest patterns affected by fire regimes under various environmental and climatic conditions after wildfires. We hypothesized that the forest regeneration patterns and their driving factors after a fire can be measured using remote sensing approaches. The research focused on the post-fire environment and responses of a Siberian boreal larch (Larix sibirica) forest ecosystem. The integration of different remotely sensed data with field-based investigations permitted the analysis of the fire regime (e.g. burn area and burn severity), the forest recovery trajectory as well as the factors that control this process with multi temporal and spatial dimensions. Results show that the monitoring of post-fire effects of the burn area and burn severity can be conducted accurately by using the multi temporal MODIS and Landsat imagery. The mapping algorithms of burn area and burn severity not only overcome data limitations in remote and vast regions of the boreal forests but also account for the ecological aspects of fire regimes and vegetation responses to the fire disturbances. The remote sensing models of vegetation recovery trajectory and its driving factors reveal the key control of burn severity on the spatiotemporal patterns in a post-fire larch forest. The highest rate of larch forest recruitment can be found in the sites of moderate burn severity. However, a more severe burn is the preferable condition for the area occupied quickly by vegetation in an early successional stage including the shrubs, grasses, conifer and broadleaf trees (e.g. Betula platyphylla, Populus tremula, Salix spp., Picea obovata, Larix sibirica). In addition, the local landscape variables, water availability, solar insolation and pre-fire condition are also important factors controlling the process of post-fire larch forest recovery. The sites close to the water bodies, received higher amounts of solar energy during the growing season and a higher pre-fire normalized difference vegetation index (NDVI) showed higher regrowth rates of the larch forest. This suggests the importance of seed source and water-energy availability for the seed germination and growth in the post-fire larch forest. An understanding of the fire regimes, forest-recovery patterns and post-wildfire forest-regeneration driving factors will improve the management of sustainable forests by accelerating the process of forest resilience.
DegreeDoctor of Philosophy (Ph.D.)
DepartmentGeography and Planning
CommitteeGuo, Xulin; Wilmshurst, John; Zeng, Winston; Tanino, Karen; Takeda, Kazuo
Copyright DateOctober 2015
boreal larch forest