Open Collections

Wang, Weiwei

University of British Columbia

The frequency of extreme wildfires has been increasing across Canada in recent history, and the situation is expected to be worsened under the changing climate. Quantifying fire effects and changes is critical to confront these challenges. Burn severity and burned area are two important aspects of fire activity that are directly connected to substantial ecological research and forest management practices, and are the main focuses of this dissertation. The primary objectives are therefore to (1) estimate burn severity and its key drivers at landscape and sub-continental scales, and (2) model climate-driven changes in forest fire areas by incorporating fire-fuel feedbacks in Canada. Burn severity was estimated at a heterogeneous landscape in the Rocky Mountains, which typically experienced mixed-severity fires, at 30-m resolution at first, and was then extended to the sub-continental scale, the Canadian ecozones, at 0.25° (~27.8 km) resolution. Random forest model and multinomial logistic regression model were used to estimate burn severity and identify its key drivers. For changes in burned areas, generalized additive models (GAMs) were built to capture the non-linear fire-climate relationships in the ecozones under four climate change scenarios. Fire-fuel feedbacks, which were built to quantify potential fuel constraints from previous fires, were then incorporated to the GAMs to assess their effects. Results demonstrated that topography, vegetation, and climate/weather contributed equally to burn severity in the Rocky Mountains. Fuel aridity measures were the most influential drivers to burn severity at the larger scale of Canadian Ecozone. Based on the 40-year (1981-2020) estimated daily burn severity in the ecozones, about 6% (0.54-14.64%) of modeled areas show significant increases in the number of high burn severity days, most of which were detected in the 2001-2020 period. The western and central uplands and eastern regions with extensive coniferous trees were shown to be most influenced by high burn severity primarily in summer, while broadleaf trees exhibited high severe-burning potential in spring. Regarding the near-future burned areas, although fire–fuel feedbacks did not change the overall trend of climate-driven fire increases (four- to five-fold), it could rectify the overestimated increases by ~27%, at the Canadian ecozone level.

Text

2025-01-31

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/86731

Forestry

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/