|
Activity Number:
|
176
- Contributed Poster Presentations: Section on Statistics and the Environment
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 31, 2017 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
|
Abstract #324821
|
|
|
Title:
|
Antarctic Subglacial Lake Detection Via a Discrete State-Switching Stochastic Volatility Model
|
|
Author(s):
|
Alyssa Jones* and Wesley Tansey and James Scott and Jamin Greenbaum
|
|
Companies:
|
The University of Texas at Austin and University of Texas at Austin and The University of Texas at Austin and The University of Texas at Austin
|
|
Keywords:
|
glaciology ;
geology ;
planetary science ;
climate ;
stochastic volatility
|
|
Abstract:
|
Discovering subglacial lakes is a crucial part of modeling ice melt speed and is consequently vital for the accuracy of downstream tasks in climate modeling. Current practice is to have experts manually analyze radar data and make educated guesses about both the presence of subglacial lakes and the lake boundary points. We present a hierarchical Bayesian model of the underlying basal reflectivity that probabilistically detects the presence or absence of subglacial lakes. We show that radar reflectivity volatility maps well to a discrete-state switching AR(1) process and observe that subglacial lakes present as areas of substantially lower innovation variance in comparison to grounded ice. To the best of our knowledge our approach represents the first principled statistical method for sublacial lake detection in aerial ice-penetrating radargrams.
|
Authors who are presenting talks have a * after their name.
