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.
16:30
satGP: Efficient Gaussian process regression for massive remote sensing data
Jouni Susiluoto | Jet Propulsion Laboratory, California Institute of Technology | United States
Show details
Authors:
Jouni Susiluoto | Jet Propulsion Laboratory, California Institute of Technology | United States
Heikki Haario | Lappeenranta University of Technology | United States
Youssef Marzouk | Massachusetts Institute of Technology | United States
We describe recent developments in satGP, which is a general purpose Gaussian process software for remote sensing data. The program is able to make use of huge numbers of observations, describe spatio-temporal multi-scale features in the data, learn model parameters, utilize wind information for improving the quality of the inferred fields, and learn the mean behavior of the quantity of interest. We describe the underlying methodology, how the software can used, and what caveats remain. The features are demonstrated by applying the software to real-world remote sensing products.
17:00
Model discrepancy in CO2 retrievals from the OCO-2 satellite
Jenný Brynjarsdóttir | Case Western Reserve University | United States
Show details
Authors:
Jenný Brynjarsdóttir | Case Western Reserve University | United States
Jonathan Hobbs | Jet Propulsion Laboratory, California Institute of Technology | United States
Amy Braverman | Jet Propulsion Laboratory, California Institute of Technology | United States
The Orbiting Carbon Observatory 2 (OCO-2) collects space-based measurements of atmospheric CO2. The CO2 measurements are indirect since the instrument observes radiances (reflected sunlight) over a range of wavelengths and a physical model is inverted, via Bayes Theorem, to estimate CO2 concentration in the atmosphere. This inference is in fact an estimation of physical parameters (an inverse problem) which can be both biased and over-confident when model error is present but not accounted for. The OCO-2 mission addresses this problem in a few different ways, e.g. with a post-inference bias correction procedure based on ground measurements. This talk will discuss methods to account for structured and informative model error directly in the inversion procedure to lessen bias and provide more reliable uncertainty estimates.
17:30
Uncertainty quantification of ACE-FTS data products
Patrick Sheese | University of Toronto | Canada
Show details
Authors:
Patrick Sheese | University of Toronto | Canada
Kaley Walker | University of Toronto | Canada
Chris Boone | University of Waterloo | Canada
The ACE-FTS (Atmospheric Chemistry Experiment–Fourier Transform Spectrometer) instrument is a limb sounding spectrometer on the Canadian SciSat satellite. It was launched into orbit in August 2003 and has been making measurements of atmospheric temperature and trace species concentrations since February 2004. However, none of the 70+ ACE-FTS data products have a fully characterized uncertainty budget. Over the past couple years, a consortium of atmospheric retrieval experts has gotten together to make recommendations on how uncertainty budgets for atmospheric profilers should be created and reported. This group, known as TUNER (Towards Unification of Error Reporting), is currently in the process of creating a "TUNER framework" that aims to formalize and codify how uncertainty budgets for atmospheric retrievals are made and reported. This study will discuss how the ACE-FTS uncertainty budget will be evaluated and reported under the auspices of the TUNER framework and will highlight some preliminary uncertainty quantification results.
18:00
Emissions from individual power plants by satellite data
Heikki Haario | Lappeenranta University of Technology | Finland
Show details
Authors:
Heikki Haario | Lappeenranta University of Technology | Finland
Teemu Härkönen | Lappeenranta University of Technology | Finland
Esa Vakkilainen | Lappeenranta University of Technology | Finland
Janne Hakkarainen | Finnish Meteorological Institute | Finland
Emission estimates from single industrial sources of various sites are produced by Tropomi satellite data and compared to available in-situ data. The primary focus is NO2, but knowledge of the plant operation allows the estimation of CO2 emissions by the "stoichiometry" of the plant as well. The 2D satellite emission data is modelled as a snapshot surface, and integrated to give a 1D concentration data for the estimation of the emission together with simplified decay chemistry. UQ is considered at every step to get the estimates as Bayesian distributions. The final results are compared with reported in-situ emissions or concentration measurements. The main sources of uncertainties are discussed.