Insights in climate variability and impacts on ecology

Abstract

Climate models predict significant changes in temperature variability over the course of the 21st century. Shifts in the temporal properties of temperature fluctuations can affect the performance of all organisms, but ectotherms, which comprise most of the biodiversity on land, are expected to be particularly impacted because they lack many of the key physiological mechanisms required to regulate thermal stress. In this dissertation, the latest generation of climate projections from Coupled Model Intercomparison Project Phase 6 (CMIP6) and dynamical population models are used to document changes in the spatiotemporal distribution of temperature over the next one hundred years and demonstrate how complex temperature changes are expected to impact the performance of 38 ectothermic species around the globe. Our results illustrate the spread of ecological response driven by inter-climate model variability and highlight the cascade of model-based uncertainty to impact studies. This dissertation further develops a framework for deep learning emulation of physics-based models in the earth sciences, which may help to accelerate the pace of climate model development and characterize model variability. Insights from this research could improve understanding of the larger scale changes in physical climate processes associated with changes in the temporal dependence structure of air temperature. Further, this work could advance strategic understanding of the effects of temporal variability of temperature of population-level dynamics with implications for understanding and mitigating the effects of climate change on ecosystems.--Author's abstract