Atmosphere Ocean Science Colloquium
Historical and Future Climate Impacts: From quantitative approaches to process understanding
Speaker: Frances Davenport, Stanford University
Date: Wednesday, January 26, 2022, 3:30 p.m.
Extreme climate events like severe precipitation, extreme heat, drought, and wildfires, have a disproportionate impact on society and ecosystems, but are challenging to model and predict due to their complexity and rarity. As a result, there are many unresolved questions about the characteristics of extreme events in a changing climate. In this talk, I will highlight examples from my research to understand climate extremes and their impacts. First, I will present research using quantitative, causal inference methods to understand the economic and societal impacts of climate extremes, illustrating an example of precipitation and flooding in the U.S. I find that more than one third of recent U.S. flood damages are due to historical changes in precipitation. I also analyze precipitation change simulated in an ensemble of global climate models and find that the pattern of observed change is consistent with human-caused climate change. In the second part of the talk, I will present research to understand the physical processes leading to changes in extreme precipitation and flooding. One effect of global warming is that warmer winter temperatures cause a shift from winter snow to winter rain. Using causal inference regression techniques to analyze data from over 400 watersheds in the western U.S., I show that this shift from snow to rain leads to non-linear increases in flood size due to larger rain-driven floods compared to snowmelt. Additionally, I use deep learning to understand the processes that lead to increases in extreme precipitation occurrence. Using the U.S. Midwest region as a case study, I find that large-scale atmospheric circulation conditions associated with extreme precipitation have become more frequent over the past 20 years. These increases in frequency have co-occurred with increases in atmospheric moisture flux to the Midwest region. As a result, these conditions are also more likely to result in extreme precipitation now than in the past. Combined, this research (i) provides new empirical evidence quantifying the impacts of climate change, which can inform policy decisions and climate adaptation, and (ii) demonstrates new methodological approaches to develop process-based understanding of climate extremes that can be extended broadly to study other hazards and regions.