Evaluating Uncertainty In Model Representations Of Land-Atmosphere Carbon Exchange And Atmosphere-Watershed Interactions Toward Informed Climate Change Impact Planning

Abstract

Investigating uncertainty within Earth’s complex climate system, as well as within tools used to represent system interactions, can contribute to multidisciplinary projects such as model benchmarking and scenario planning. In this dissertation, three case studies are presented that focus on climate system connections to land management and water quality. First, global soil heterotrophic respiration (HR) is a component land flux that drives net carbon exchange between the atmosphere and terrestrial ecosystems. Due to observational limitations, it is difficult to quantify HR or to evaluate it in global-scale model simulations, leading to error ranges with comparable magnitude to recent fossil fuel burning. We analyze three soil model configurations that simulate HR fluxes within a biogeochemical testbed and subsequently quantify the effects of variation in carbon fluxes on atmospheric CO2 distributions using a three-dimensional atmospheric tracer transport model. We show that atmospheric CO2 can provide a useful constraint on large-scale patterns of soil heterotrophic respiration. Second, Lake Erie has experienced a resurgence of harmful algal blooms (HABs) that is attributed to agricultural practices and fertilizer run-off exacerbated by spring rain events. We examine extreme precipitation events for the Great Lakes Basin and the Western Lake Erie Basin subregion for historical and mid-21st century periods by quantifying changes in precipitation seasonality and daily intensity. We utilize three model ensembles that cover a range of spatial scales and future emissions scenarios to evaluate the roles of model method and grid resolution within the projection output. Third, climate change influences regional drivers of HABs, emphasizing the need for usable information in planning and policy. HAB modeling and mitigation actions have focused on the agricultural sector to reduce land run-off and nutrient loading in Western Lake Erie. Other regional stakeholders, such as drinking water managers, beach managers, marina operators, and recreational land owners, respond and adapt to algal blooms once they form, representing different data needs from those in the mitigation space. We define sources of uncertainty in climate, watershed, and HAB modeling and present four pathways for conceptualizing uncertainty across a modeling chain. We further discuss how scenario planning can incorporate model uncertainty information and stakeholder knowledge for HAB decision-making.