Join us for a virtual seminar by Dr. Yixin ‘Berry’ Wen, Department of Geography, University of Florida. Her talk is titled "Creating synergy between satellite and radar precipitation measurements aided by symbolic deep learning."
Date and Time: Thursday, March 14, 2024 at 11:00am EST
Join us via Teams.
Abstract:
"Ground-based polarimetric radars - WSR-88D network in the US - underpin nationwide precipitation observations and estimates with advanced Doppler capability. Alternatively, spaceborne radar - GPM DPR (Global Precipitation Measures Dual-frequency Precipitation Radar) - has an unprohibited downward view and higher vertical resolution, relative to ground-based radars. The synergy between space-borne and ground-based radars combined with AI/ML creates promising opportunities for observing and probing three-dimensional clouds and precipitation structures. However, it is a prerequisite to fully understand and quantify the similarities and differences between the WSR-88D and GPM DPR.
The first half of the talk will focus on the nationwide comparisons at 140 WSR-88D radar sites from 2014 to 2020. Systematic differences are found across the ground radar sites with an average of ~ 2 dB. The recently updated DPR version 7 product improves rain detection and attenuation corrections, effectively reducing the overall average WSR-88D and DPR reflectivity differences to 1.0 dB.
The second half of the talk will focus on polarimetric radar quantitative estimation (QPE) using AI/ML methods. Conventional parametric radar QPE algorithms such as the radar reflectivity and rain rate relations cannot fully represent clouds and precipitation dynamics. Recent years have seen the application of Deep Learning methods to improve QPE products. However, the 'black box' nature of deep learning poses challenges in understanding the underlying physical meaning. In contrast, symbolic learning, a branch of artificial intelligence, utilizes symbols to represent knowledge in a clear and interpretable manner. The presentation will showcase our preliminary findings from our exploration of neural network and and symbolic learning approaches for polarimetric radar QPE."
Biography:
Dr. Berry Wen is an Assistant Professor of the Department of Geography at the University of Florida. Berry received her PhD degree from the School of Meteorology, University of Oklahoma. She further honed her skills during postdoctoral research at the Jet Propulsion Laboratory before returning to Oklahoma as a research scientist at CIWRO. In this role, she was affiliated with Radar Division in National Severe Storms Laboratory.
Berry is a member of the User Working Group with NASA/Goddard Earth Sciences Data and Information Services Center (GES DISC) as an expert in AI and data fusion. Additionally, she serves as an editor of the AGU’s new journal, JGR: Machine Learning and Computation.
Berry's research areas of focus include radar and satellite remote sensing, machine learning, and deep learning. Additionally, she engages in long-term climate data analysis, studying extreme events and water-related natural hazards. Beyond her scientific pursuits, Dr. Wen is committed to working with Indigenous Peoples and Native Nations for Environmental Justice and Climate Justice.
For more information on the GESTAR II Seminar Series, click here.