Modelling the effect of fire, insect, and logging disturbances on climate and vegetation across various spatial and temporal scales

Fire, insect, and logging disturbances affect ecosystems worldwide and often lead to spectacular local changes in vegetation characteristics, as well as biogeochemical and biogeophysical fluxes. Despite a growing body of empirical and modelling studies, many key questions remain unanswered, and even unasked, about the effect of these disturbances on climate and vegetation. In this thesis, I explored some of these questions at different spatial and temporal scales, and mostly from a modelling perspective. In Chapter 2, I simulated various future (2015–2300) fire regimes in the University of Victoria Earth System Climate Model and estimated the resulting impacts on global carbon stocks and surface temperature, with and without the effect of the main fire-related non-CO2 emissions. This global-scale study in a fully coupled climate–carbon model allowed me to differentiate the net from the gross carbon emissions in response to changes in fire frequency, a crucial distinction that is sometimes overlooked in the literature, and to also highlight the possibly dominant effect of fire-emitted aerosols on the net climatic impact from non-deforestation fires. In Chapter 3, I investigated, with a new model developed for this purpose, how different approaches to represent fire and logging in climate models affected the simulation of albedo over boreal forests. This methodological study showed that the simplest approach, which was the one applied in Chapter 2, noticeably underestimates the long-term fire-induced albedo increase over boreal forests, but that substantial improvements can be obtained without undue additional computing requirements. Having focussed on stand-clearing disturbances in the previous two chapters, I looked at the role of insect outbreaks in Chapter 4. I assessed, with a modified version of the Integrated BIosphere Simulator dynamic vegetation–land surface model, the long-term impacts (up to >1,000 years) from recurrent mountain pine beetle (MPB) outbreaks at three different locations in British Columbia on different vegetation- and climate-related variables, namely merchantable biomass, ecosystem carbon, albedo, and net radiative forcing. This study emphasized the major role of the non-target vegetation in MPB-induced changes and illustrated various non-linearities in the responses to recurrent outbreaks. In Chapter 5, I finally performed a critical review of the literature relevant to the climate–forest–disturbances triad in Canada. More precisely, I proposed five principles relevant for the management of Canadian forests in the context of carbon cycling, climate regulation, and disturbances, and then applied these principles to address four questions of current interest. One conclusion from this study was the need to perform >100-year analyses of disturbances in Canadian forests, as I did in Chapter 4. Overall, suggesting that fire or insect disturbances seem to generally have less impact per event as they become more frequent (i.e., sub-linear scaling) is the most important scholarly contribution from this thesis. Another major outcome consists of identifying sources of uncertainty that prevent sound conclusions on the net warming or cooling impact from natural disturbances, namely the strength of cloud-mediated aerosol effects in the case of fire and the response from the non-target vegetation in the case of MPB outbreaks. Other contributions to knowledge include the possible influence of fire-emitted aerosols on land–atmosphere and ocean–atmosphere carbon exchanges, the impact of dead standing trees on the post-disturbance carbon response due to their interactions with energy and water exchanges, and the idea that the spatial distribution of trees killed by insects can modulate the resulting biogeophysical and biogeochemical consequences. Finally, this thesis illustrates that fire, insect, and logging disturbances lead to both large and small impacts, depending upon the specific scale and element considered.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.