Activity Number:
|
165
- SPEED: Environmetrics: Spatio-Temporal and Other Models
|
Type:
|
Contributed
|
Date/Time:
|
Monday, July 30, 2018 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
Abstract #329620
|
|
Title:
|
Tools for Simulation-Based Uncertainty Quantification in Remote Sensing Inverse Problems
|
Author(s):
|
Jonathan Hobbs* and Amy Braverman and Ali Behrangi and Sandy Burden and Eric Fetzer and Kyo Lee and Hai Nguyen
|
Companies:
|
Jet Propulsion Laboratory and Jet Propulsion Laboratory and University of Arizona and University of Wollongong and Jet Propulsion Laboratory and Jet Propulsion Laboratory and Jet Propulsion Laboratory
|
Keywords:
|
uncertainty quantification;
inverse problem;
remote sensing;
satellite;
hierarchical model;
carbon cycle
|
Abstract:
|
Earth-orbiting satellites provide substantial information on numerous atmospheric and surface quantities of interest. The scientific data products produced from these remote sensing instruments are estimates based on complex retrieval algorithms. Each retrieval strategy presents unique challenges, but most methods involve a physical and/or statistical model with a computational inverse method. This work develops tools for executing and summarizing Monte Carlo experiments for atmospheric remote sensing retrievals, with an emphasis on uncertainty quantification. We illustrate these tools with two remote sensing instruments that use different retrieval methods. The Orbiting Carbon Observatory-2 (OCO-2) provides high-resolution estimates of carbon dioxide using a Bayesian hierarchical model. The Atmospheric Infrared Sounder (AIRS) provides estimates of atmospheric temperature and humidity profiles using a nonlinear regression in combination with a cloud-clearing procedure.
|
Authors who are presenting talks have a * after their name.