Modeling Extremal Streamflow using Deep Learning Approximations and a Flexible Spatial Process

Presented by Reetam Majumder, North Carolina State University

Thursday, January 18 2024
3:30 PM-4:30 PM ET
AUST 105
Webex Meeting Link
Coffee will be served at 3:00 pm in the Noether Lounge (AUST 326)

Quantifying changes in the probability and magnitude of extreme flooding events is key to mitigating their impacts. While hydrodynamic data are inherently spatially dependent, traditional spatial models such as Gaussian processes are poorly suited for modeling extreme events. Spatial extreme value models with more realistic tail dependence characteristics are under active development. While theoretically justified, they have intractable likelihoods - this makes computation challenging for small datasets and prohibitive for continental-scale studies. We propose a process mixture model (PMM) which specifies spatial dependence in extreme values as a convex combination of a Gaussian process and a max-stable process, yielding desirable tail dependence properties but intractable likelihoods. To address this, we employ a unique computational strategy where a feed-forward neural network is embedded in a density regression model to approximate a surrogate likelihood for the PMM. This approach facilitates full Bayesian posterior inference, and can provide future projections of streamflow. The PMM is used to analyze changes in annual maximum streamflow within the US over the last 50 years, and is able to detect areas which show increases in extreme streamflow over time.

Speaker Bio:

Reetam is a postdoc at NC State University, and a member of the inaugural USGS Climate Adaptation Postdoctoral Fellows cohort conducting interdisciplinary research on the future of fire in the US. He is also involved in several related projects including modeling extreme streamflow and wildfire risk, downscaling climate model data, and the detection and attribution of climate change. Reetam completed his PhD in Statistics from the University of Maryland Baltimore County in 2021, and his research is at an intersection of Bayesian inference, computational statistics, and deep learning.