Jesus Lopez
MIT Department: Earth, Planetary and Atmospheric Sciences
Faculty Mentor: Prof. Paul O’Gorman
Research Supervisor: Justin Finkel
Undergraduate Institution: Texas A&M University, San Antonio
Hometown: San Antonio, Texas
Website: LinkedIn
Biography
Jesus Lopez is a rising senior at Texas A&M University-San Antonio, double majoring in Biology and Computer Science. He has always possessed a great devotion to environmental preservation, one that has become coupled with an interest in computer science and mathematics. His research focuses on computational applications in urban ecology and environmental science. This summer, he is working at MIT to utilize and optimize atmospheric models to predict extreme precipitation events. Last summer, he interned at Harvard Systems Biology, analyzing chronic pollution effects on Zebrafish behavior. At his home university, he is developing an agent-based model to simulate amphibian populations, aiding educators, researchers, and wildlife managers. He aims to pursue a Ph.D. in computational modeling for urban ecology and environmental science, hoping to work in the federal government to contribute to environmental regulations and legislation. Offline, Jesus enjoys spending time with loved ones, baking, biking, and promoting sustainability in his city.
Abstract
Analyzing Idealized Atmospheric Models to Understand the Spatiotemporal Dynamics of Extreme Precipitation Events.
Jesus A. Lopez1, Justin Finkel2, Paul O’Gorman2
1Department of Natural Sciences, Department of Computational, Engineering, and Mathematical Science, Texas A&M University-San Antonio
2Department of Earth, Atmospheric, and Planetary Sciences, Massachusetts Institute of Technology
Climate change has increased the frequency and severity of extreme weather events, significantly impacting vulnerable populations. These extreme events can be simulated numerically, however, producing enough samples to quantify their probabilities is challenging, especially when considering changing climate conditions. Conceptual and computational difficulty arises from the interacting components of Earth system models, whose effects are hard to disentangle. To make conceptual progress, we use simplified models representing water vapor as tracers: substances moved by wind. Literature suggests that this assumption explains empirical tracer characteristics, like heavy-tailed distributions, and can be modeled using prescribed velocity fields. However, past studies mostly focus on spatially and temporally aggregated statistics. Here we use an idealized flow to explore the dynamics of temporally intermittent, spatially localized events for regional risk assessment. First, we characterize spatially dependent climatological statistics of tracer concentrations, assessing skewness and other non-Gaussian characteristics. Second, we identify local spikes of tracer concentration and perturb previous conditions at several lead times to re-simulate more intense events. This reveals timescales where the spread of extreme event magnitudes is maximal. Our work makes conceptual progress on the causality for individual weather events, while also informing future studies using rare event algorithms based on trajectory splitting.
