Thirty-eight oligotrophic lakes located around the Reservoir Gouin in central Quebec (48°N, 75°W) were sampled over three years to test the hypothesis that forest clearcutting and fires should be reflected in both higher nutrient export rates and ultimately in greater areal hypolimnetic oxygen deficit rates (AHOD). Significant differences in estimated total phosphorus export rates across treatments were found. However, no effect of clearcutting or forest fire on hypolimnetic oxygen consumption rates could be demonstrated as the result of a much greater and confounding variation in the effect of lake morphometry and the absence of information on the role of catchment-derived organic matter on the AHOD. Consequently, only lake morphometry (hypolimnetic volume to hypolimnetic surface area ratio) served as a predictor of the AHOD. Covariation of mean hypolimnetic water temperature with morphometric variables underlines the influence of lake morphometry on heat dynamics and hypolimnetic respiration rates in these lakes. / This research made considerable use of specialized data manipulation techniques involving a relational database management system, owing to the size of the dataset used (114 lake-years of data). The specific approach used in this thesis is presented in an appendix.

No description available.

Fires are the primary method of deforestation and agricultural management in the tropics, but associated emissions such as aerosols, ozone, and carbon monoxide can have negative impacts on ecosystems, climate, and public health. Recent advances in satellite monitoring of fire activity, including using thermal anomalies for active fire detections and burn scar mapping of post-fire effects, have offered an unprecedented level of detail in understanding the magnitude and extent of fire activity. This dissertation aims to quantify the human health impact across populations in tropical regions by determining which areas are the most susceptible to transported fire emissions and how this exposure varies over time. The following chapters can be used to highlight critical conservation regions, not only for conserving ecosystems for biodiversity and climate benefits, but also for protecting public health. To address how fire emissions can affect regional populations, satellite observations of fire activity are combined with models of how tropical fire emissions are transported in the atmosphere. Satellites provide two primary pieces of information for this approach: 1) measurements of the distribution and magnitude of fire activity, and 2) categorization of fire types (such as agricultural burning or deforestation) by overlaying observed fire patterns on land use maps. Atmospheric models perform the crucial step of simulating how emissions evolve and where they are transported after release into the atmosphere. The following dissertation chapters are linked through exploration of fire emissions impacts from continental to local scales, including implementing fire emissions inventories into atmospheric models, quantifying population exposure to fire activity in Equatorial Asia, and projecting fire emissions associated with various future land use scenarios in Sumatra. Model estimates of aerosol concentrations are more influenced than trace gases by using finer temporal resolution fire emissions, due to interactions between emissions and modeled meteorology and transport. This in turn can impact air quality estimates by permitting higher peak concentrations. In addition, model results show that population exposure to fire emissions in Equatorial Asia is highly variable over time depending on the phase of the El Niño cycle; strong El Niño years can have fire contributions to fine particulate matter of up to 200 µg/m³ near fire sources, corresponding to 200 additional days per year over the World Health Organization 50 µg/m³ 24-hour fine particulate matter air quality target. These risks are not confined to people living near fire sources, but expose broad regional populations due to the atmospheric transport of emissions. Health impacts also depend on underlying fuel characteristics, with the future magnitude of Equatorial Asian fire emissions estimated to be strongly dependent on the level of protection given to fuel-rich peatswamp forests (contributing 33-48% of future emissions in the absence of protection). Collectively, these chapters emphasize variability in how tropical fire emissions affect regional population exposures to outdoor air pollution, and the need to consider the dependence of this public health effect on different fuel types and year-to-year variations in climate. The results described in this dissertation quantify direct benefits of conservation for people living near fire areas.

This study provides an analysis of watershed response to climate change and forest fire impacts, to better understand the hydrologic budget and inform water management decisions for present and future needs. The study site is 2,365 km2, located in the upper Umatilla River Basin (URB) in northeastern Oregon. The Precipitation Runoff Modeling System, a distributed-parameter, physical-process watershed model, was used in this study. Model calibration yielded a Nash Sutcliffe Model Efficiency of 0.73 for both calibration (1995-2010) and validation (2010-2014) of daily streamflow. Ten Global Climate Models using Coupled Model Intercomparison Project Phase 5 experiments with Representative Concentration Pathways 4.5 and 8.5 (RCP), were used to observe hydrologic regime shifts in the 2020s, 2050s, and 2080s. Mean center timing of flow occurs earlier in the year in both pre- and post-fire conditions, where there are increased winter flows and decreased summer flows throughout the 21st century. Change in temperature and percent change in precipitation is more variable in the summer than winter increasing over time, with a slight decrease in winter precipitation in the 2080s in RCP 8.5. Temperature increases 1.6°C in RCP 4.5 and 3.3°C in RCP 8.5 by the end of the 21st century. The ratio of Snow Water Equivalent to Precipitation decreases 96% in the 2080s in RCP 8.5 before forest cover reduction, and decreases 90-99% after forest cover reduction. Potential basin recharge and the base-flow index are both sustained throughout the 21st century with slight declines before forest cover reduction, with an increase in basin recharge and increase in base-flows in the 2080s after fire-burns. However, the simulated sustained base-flows and area-weighted basin recharge in this study, do not take into account the complex geologic structure of the Columbia River Basalt Group (CRBG). A more robust characterization and simulation of URB aquifer recharge would involve coupling the PRMS model with a groundwater model in a future study. Although groundwater recharge in the CRBG in the URB is not well understood, the long-term decline of groundwater storage presents a serious environmental challenge for the Confederated Tribes of the Umatilla Indian Reservation and communities in the URB.

Forested watersheds in western North America are subject to significant change from natural and anthropogenic disturbance, including wildfire. Forest canopy changes have subsequent impacts on sub-canopy snow processes. A simple, process-based point energy balance model was developed to quantify differences in energy balance characteristics between a burned and a healthy forest stand. Potential model uncertainties were identified using sensitivity analyses. Simulated snowmelt accurately recreated measured snowmelt, providing confidence in the model’s ability to simulate energy balance processes in subcanopy environments where wind redistribution and sublimation are not major drivers of the local snowmelt energy balance. In the burned stand, sub-canopy snow accumulation was greater but melted more rapidly than in the healthy stand. The removal of forest canopy resulted in more energy available for snowmelt, including higher short-wave and lower long-wave radiation, and increased turbulent fluxes. Burned stands should be considered a separate land cover type in larger scale watershed models. / xii, 129 leaves : ill,, map ; 29 cm