Detection of Climate Extremes and Their Effects on Hydrology and Ecosystems: Case Studies in California and the Great Plains, US

Future climate extremes are projected to increase with global climate change. However, the effects of climate extremes on hydrological and ecosystem cycling remain unclear and there is uncertainty in the detection of climate extremes. The first study uses existing extreme climate indices to study the interactions between hydrological cycling and extreme precipitation. Most current studies focus on the effects of long-term climate variability on groundwater recharge but ignore the effects of extremes. Through use of a soil water balance model, my results show that extreme precipitation plays a significant role in determining recharge rates in the Northern Great Plains. The second study focuses on the impacts of drought, increased precipitation, and warming on ecosystem level nutrient allocation within grasslands in the Southern Great Plains. Our results show that these factors have significant influences on how much carbon and nitrogen are allocated to roots vs. shoots. This study also shows that an updated two-layer-soil model with an "S" curve approach to improve root water access can produce reasonable discharge rates and ecosystem productivity. The final study introduces two functional data analytic approaches, functional principle component analysis (FPCA) and functional boxplot, and applies them to imputing missing meteorological data and to detect extreme temperature events. Most existing imputation methods are based on linear regression models, which are highly limited by data quality from surrounding weather stations. The FPCA solves this problem by utilizing the most available data from additional weather stations in the same climate zone. Our results demonstrate that the number of the weather stations showing extreme daytime temperatures has increased in California.