Numerous Earth-observing satellites provide high-resolution and high-volume data that facilitate scientific inference on physical and environmental processes. Most remote sensing data products used for scientific investigations are often subject to multiple stages of processing before they are widely used, and the scientific utility of these data products critically depends on a comprehensive assessment of the sources of uncertainty encountered in these stages of processing. One key stage involves the use of a retrieval algorithm to infer a geophysical quantity of interest from a satellite’s observed intensity of radiation.
The retrieval is an inverse problem that has been implemented mathematically and computationally in numerous ways for different satellite missions. Several of the presentations in this mini-symposium will each highlight an individual Earth-observing satellite and its retrieval methodology, emphasizing important contributions to uncertainty in retrieval data products. Methodological developments that interrogate the joint distribution of true geophysical states, retrieved states, and observed satellite spectra will be introduced. The presentations will span multiple Earth science applications, including weather and climate, the carbon cycle, air quality, atmospheric chemistry, and ecosystem health.
08:30
Uncertainty quantification for temperature and humidity retrievals from atmospheric sounders
Jonathan Hobbs | Jet Propulsion Laboratory, California Institute of Technology | United States
Show details
Authors:
Jonathan Hobbs | Jet Propulsion Laboratory, California Institute of Technology | United States
Amy Braverman | Jet Propulsion Laboratory, California Institute of Technology | United States
Eric Fetzer | Jet Propulsion Laboratory, California Institute of Technology | United States
Kyo Lee | Jet Propulsion Laboratory, California Institute of Technology | United States
Hai Nguyen | Jet Propulsion Laboratory, California Institute of Technology | United States
Joaquim Teixeira | Jet Propulsion Laboratory, California Institute of Technology | United States
A multi-decade record of satellite observations from the Atmospheric Infrared Sounder (AIRS) has found utility in numerous Earth science investigations. The instrument's spectral range includes sensitivity to atmospheric temperature and clouds at different vertical levels, as well as to abundance of several trace gases, including water vapor. These important weather and climate variables can be inferred from AIRS spectral observations as part of an inverse problem known as a retrieval. The retrieval involves a physical and/or statistical model with a computational inverse method. In this presentation, we use simulation to examine the distribution of the retrieval error distribution of these quantities for the AIRS retrieval, which uses a nonlinear regression in combination with a cloud-clearing procedure. We illustrate the impacts of the retrieval uncertainty on derived quantities of interest in several weather and climate applications and discuss the implications for follow-on sounding missions such as the Cross-track Infrared Sounder (CrIS).
09:00
Multivariate spatial statistical modeling and applications in simulation-based uncertainty experiments for remote sensing
Amy Braverman | Jet Propulson Laboratory, California Institute of Technology | United States
Show details
Authors:
Emily Kang | University of Cincinnati | United States
Amy Braverman | Jet Propulson Laboratory, California Institute of Technology | United States
Uncertainty quantification for remote sensing usually require simulation-based experiments. The retrievals are evaluated at varying instrument configurations and input parameters. Some inputs often present not only spatial dependence but also inter-variable dependence. Such features should be maintained through the simulations in order to assess and quantify the uncertainty in output accurately and realistically. We propose a multivariate spatial statistical model that possesses the flexibility to characterize nonstationary and asymmetric cross-covariance structure but also allows for efficient computation. Numerical examples and an application to inputs for NASA’s ECOSTRESS are used to illustrate the advantages of the proposed model.
09:30
Assessing the impact of uncertainty in the Orbiting Carbon Observatory-2 estimates of CO2 concentration
Joaquim Teixeira | Jet Propulsion Laboratory, California Institute of Technology | United States
Show details
Authors:
Joaquim Teixeira | Jet Propulsion Laboratory, California Institute of Technology | United States
Amy Braverman | Jet Propulsion Laboratory, California Institute of Technology | United States
Jonathan Hobbs | Jet Propulsion Laboratory, California Institute of Technology | United States
Satellites that track atmospheric greenhouse gases, such as NASA’s Orbiting Carbon Observatory-2 (OCO-2), measure spectral radiances at fine spatial and temporal resolutions. Retrieval algorithms are designed to estimate the atmospheric constituent of interest, such as carbon dioxide (CO2), by modelling the process of radiative transfer through the atmosphere. The retrieved atmospheric states inform key hypotheses in carbon cycle science, but these inferences require an assessment of the sources of uncertainty in the retrieval process, including algorithm parameter choices and the inherent variability of the atmospheric states and measured radiances. One of the fundamental uses of OCO-2’s data product is carbon flux inversion modelling, which relies on the satellite’s global coverage to produce detailed maps of CO2 sources and sinks. We present the results of a series of simulation experiments and spatial statistical analyses designed to quantify the uncertainty for individual retrievals and aggregate summaries, assessing the impact of rigorous uncertainty quantification on global flux inversion modelling.
10:00
An uncertainty quantification framework for optimal estimation retrieval of aerosols
Meredith Franklin | University of Southern California | United States
Show details
Author:
Meredith Franklin | University of Southern California | United States
Although satellite-derived aerosol properties including aerosol optical depth (AOD) are retrieved quantities, rarely are uncertainty estimates provided as part of the released product. Retrieval algorithms are complex and have many approximations and assumptions in addition to unresolvable spatial dependencies and sensor noise that can produce errors in estimated quantities, which can impact conclusions and decisions based on the retrieved data. In preparation for the launch of NASA’s MultiAngle Imager for Aerosols (MAIA), we have created a testbed for quantifying uncertainties associated with AOD retrieved by optimal estimation (OE). MAIA is a multi-angle, spectro-polarimetric imager that will estimate ground-level particulate matter (PM) air pollution based on OE-retrieved top of atmosphere aerosol properties including AOD. We apply OE through the General Retrieval of Aerosol and Surface Properties (GRASP) algorithm. Several scenarios are tested by systematically perturbing a selected set of initial parameters in the inversion model. Under this ensemble approach we produce a scenario-based description of the retrieval system. We evaluate the retrieved estimates against aerosol robotic network (AERONET) measurements. Our approach provides a much needed framework for deriving uncertainties for aerosol retrievals and emphasizes the importance of uncertainty quantification in complex, multi-stage satellite product development.