Biogeochemical and Biophysical Consequences of Disturbances in Forests of the Western United States

Forest disturbances are accelerating in magnitude, frequency and extent in the United States, driven by anthropogenic land management practices, as well as climatic factors associated with elevated temperatures and increase in incidence of droughts. These disturbances lead to changes in ecosystem properties associated with forest structure, composition and function. They can have significant consequences for ecosystem services, and can subsequently alter plant biomass accumulation and storage, water availability, and land surface temperatures. The complete carbon, water, and energy consequences of disturbances and their parameterization in ecosystem models is clearly lacking, with need for improved representations of mechanisms related to direct emissions, heterotrophic respiration of disturbance killed biomass, and post-disturbance recovery.;The broad scientific purpose of this study is to forge new understanding of the post-disturbance dynamics of carbon, water and energy balance at a range of spatial (local to regional) and temporal (past, present and future) scales. The study integrates multiple datasets and methodological approaches linked to eddy covariance and micro-meteorological observations, forest inventories, process based biophysical and biogeochemical models, statistical analysis, geographical information science (GISci) and remote sensing derived process drivers and disturbance products as well as more recently developed Bayesian-based Markov Chain Monte Carlo (MCMC) model-data fusion techniques.;The first paper is entitled "Fire induced carbon emissions and regrowth uptake in western United States forests: Documenting variation across forest types, fire severity, and climate regions". This paper is motivated by the need to account for the fire severity effects on all the components of carbon balance including direct and indirect emissions of carbon pools and regrowth sinks at regional scales. The results demonstrate that ecosystem recovery varies greatly by forest type as well as severity, with the maximum net ecosystem productivity (NEP) and its timing, minimum NEP, and NEP crossover time (number of years before NEP > 0) all increasing with higher fire severity. Scaling to the region, under most intensive biomass combustion, western United States (US) fires between 1984 and 2008 imposed a net source of 12.3 TgC yr -1 in 2008.;The second paper is entitled "Carbon consequences of bark beetle outbreaks across the western United States". This study is motivated by the need to understand the carbon consequences of bark beetle outbreaks, and associated recovery across larger spatial scales. This study determines that the best estimates of biomass killed ranges from 3.69 to 12.3 Tg C yr-1, and NEP reduction ranges from 5.02 to 14.46 Tg C yr-1 during 2000 to 2009 bark beetle outbreak period after accounting for data and model uncertainties. The carbon flux legacy of 2000-2009 outbreaks will continue decades into the future (e.g. 2040-2060) as committed emissions from heterotrophic respiration of beetle-killed biomass are partially balanced by forest regrowth and carbon accumulation.;The third paper is entitled "On model-data fusion to determine optimal parameter sets for a recently clearcut, regenerating temperate forest: Managing equifinality and uncertainty to quantify inter-annual variability". This study is driven by the need to develop simple and intuitive statistical approaches to quantify equifinality (i.e., different parameter values leading to the same solution). The study demonstrates that equifinality can vary depending on the scale of the parameter space analyzed, and that parameters controlling biological processes have larger equifinality than physical parameters. After accounting for equifinality, this paper goes on to show that parameters related to vegetation structure have larger year-to-year trends than those related to physical function.