This is a past event.

Please join ATM graduate students, Sabrina Li and Geeta Nain for their presentations on Monday, April 7 at 4 PM - Fisher Hall 125.

Sabrina Li (Advisor: Xin Xi)

Understanding the Dust Emission Dynamics using Explainable Artificial Intelligence (XAI)

Wind-blown dust from global dryland areas is an essential element of the Earth’s biogeochemical cycle, and produces a wide range of impacts on the climate system, environmental health, and socioeconomic activities. Yet, climate model representations of dust emission processes remain highly uncertain, with the sources of model disagreement and errors still largely unknown due to limited observational constraints. In this work, we explore the use of a newly developed ground-based dust event record and explainable artificial intelligence techniques to advance our understanding of dust emission dynamics, and characterization of dust event prediction model errors under different hydroclimate regimes. In this presentation, I will present preliminary results on the dust event prediction errors in the MERRA2 reanalysis over Asia.

Geeta Nain (Advisor: Pengfei Xue)

Hurricane induced hazard with coupled and parametric modeling

Hurricane-induced hazards, including extreme winds, waves, and storm surge, pose significant risks to offshore wind farms, coastal infrastructure, and communities. The increasing frequency and intensity of hurricanes, exacerbated by rapid coastal development and rising sea levels, further amplify these threats. Accurate assessment of hurricane wind structures is crucial for offshore wind farm site characterization, while high-resolution onshore surface wind data is essential for storm surge modeling. Parametric wind models, which approximate surface wind fields, offer computational efficiency for risk assessment but are often limited by their simplified representation of wind structures. This study investigates hurricane boundary layer wind structures using both parametric and coupled modeling approaches. For coupled modeling, a fully integrated system—C-WFS (WRF-FVCOM-SWAN; Jung et al 2025) is employed, combining the Weather Research and Forecasting (WRF) model for atmospheric dynamics, the Finite Volume Community Ocean Model (FVCOM) for ocean circulation, and the Simulating Waves Nearshore (SWAN) model for wave dynamics. Beyond analyzing hurricane wind structures, this study also compares three widely used parametric wind models for surface wind estimation and evaluates their impact on storm surge simulations. The findings aim to enhance the accuracy of wind field representation and improve storm surge modeling for coastal resilience planning.