Uncertainty‐Aware Machine Learning Bias Correction and Filtering for OCO‐2: 1

Abstract The Orbiting Carbon Observatory‐2 (OCO‐2) makes space‐based radiance measurements of reflected sunlight. Using a physics‐based retrieval algorithm, these measurements are inverted to estimate column‐averaged atmospheric carbon dioxide dry‐air mole fractions (XCO2). However, biases are prese...

Full description

Saved in:
Bibliographic Details
Main Authors: Steffen Mauceri, William Keely, Josh Laughner, Christopher W. O’Dell, Steven Massie, Robert Nelson, David Baker, Matthäus Kiel, Otto Lamminpää, Jonathan Hobbs, Abhishek Chatterjee, Tommy Taylor, Paul Wennberg, Sean Crowell, Britton Stephens, Vivienne H. Payne
Format: Article
Language:English
Published: American Geophysical Union (AGU) 2025-07-01
Series:Earth and Space Science
Online Access:https://doi.org/10.1029/2025EA004328
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850078852272881664
author Steffen Mauceri
William Keely
Josh Laughner
Christopher W. O’Dell
Steven Massie
Robert Nelson
David Baker
Matthäus Kiel
Otto Lamminpää
Jonathan Hobbs
Abhishek Chatterjee
Tommy Taylor
Paul Wennberg
Sean Crowell
Britton Stephens
Vivienne H. Payne
author_facet Steffen Mauceri
William Keely
Josh Laughner
Christopher W. O’Dell
Steven Massie
Robert Nelson
David Baker
Matthäus Kiel
Otto Lamminpää
Jonathan Hobbs
Abhishek Chatterjee
Tommy Taylor
Paul Wennberg
Sean Crowell
Britton Stephens
Vivienne H. Payne
author_sort Steffen Mauceri
collection DOAJ
description Abstract The Orbiting Carbon Observatory‐2 (OCO‐2) makes space‐based radiance measurements of reflected sunlight. Using a physics‐based retrieval algorithm, these measurements are inverted to estimate column‐averaged atmospheric carbon dioxide dry‐air mole fractions (XCO2). However, biases are present in the retrieved XCO2 due to sensor calibration errors and discrepancies between the physics‐based retrieval and nature. We propose a Random Forest (RF), a non‐linear, interpretable machine learning (ML) technique, to correct these biases. The approach is rigorously validated, comes with quantified uncertainties, and is derived independent of carbon flux models. Compared to the operational approach, our method reduces unphysical variability over land and ocean and shows closer agreement with independent ground‐based observations from the Total Carbon Column Observing Network. The RF‐bias correction is suitable for integration into the operational processing pipeline for the next version of OCO‐2 products, pending additional testing and validation. It is inherently generalizable to other existing and planned greenhouse gas monitoring missions. This paper (Part 1) describes the RF bias correction, while a second paper (Part 2) describes the development of a data filtering strategy specifically designed for a subset of retrievals exhibiting irreducible errors that remain inadequately corrected by the ML bias correction.
format Article
id doaj-art-a82d2ee958c84b6f905bf74f308c1234
institution DOAJ
issn 2333-5084
language English
publishDate 2025-07-01
publisher American Geophysical Union (AGU)
record_format Article
series Earth and Space Science
spelling doaj-art-a82d2ee958c84b6f905bf74f308c12342025-08-20T02:45:27ZengAmerican Geophysical Union (AGU)Earth and Space Science2333-50842025-07-01127n/an/a10.1029/2025EA004328Uncertainty‐Aware Machine Learning Bias Correction and Filtering for OCO‐2: 1Steffen Mauceri0William Keely1Josh Laughner2Christopher W. O’Dell3Steven Massie4Robert Nelson5David Baker6Matthäus Kiel7Otto Lamminpää8Jonathan Hobbs9Abhishek Chatterjee10Tommy Taylor11Paul Wennberg12Sean Crowell13Britton Stephens14Vivienne H. Payne15Jet Propulsion Laboratory California Institute of Technology Pasadena CA USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USACooperative Institute for Research in the Atmosphere Colorado State University Fort Collins CO USALaboratory for Atmospheric and Space Physics University of Colorado Boulder CO USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USACooperative Institute for Research in the Atmosphere Colorado State University Fort Collins CO USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USACooperative Institute for Research in the Atmosphere Colorado State University Fort Collins CO USACalifornia Institute of Technology Pasadena CA USALumenUs Scientific LLC Oklahoma City OK USANSF National Center for Atmospheric Research Boulder CO USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USAAbstract The Orbiting Carbon Observatory‐2 (OCO‐2) makes space‐based radiance measurements of reflected sunlight. Using a physics‐based retrieval algorithm, these measurements are inverted to estimate column‐averaged atmospheric carbon dioxide dry‐air mole fractions (XCO2). However, biases are present in the retrieved XCO2 due to sensor calibration errors and discrepancies between the physics‐based retrieval and nature. We propose a Random Forest (RF), a non‐linear, interpretable machine learning (ML) technique, to correct these biases. The approach is rigorously validated, comes with quantified uncertainties, and is derived independent of carbon flux models. Compared to the operational approach, our method reduces unphysical variability over land and ocean and shows closer agreement with independent ground‐based observations from the Total Carbon Column Observing Network. The RF‐bias correction is suitable for integration into the operational processing pipeline for the next version of OCO‐2 products, pending additional testing and validation. It is inherently generalizable to other existing and planned greenhouse gas monitoring missions. This paper (Part 1) describes the RF bias correction, while a second paper (Part 2) describes the development of a data filtering strategy specifically designed for a subset of retrievals exhibiting irreducible errors that remain inadequately corrected by the ML bias correction.https://doi.org/10.1029/2025EA004328
spellingShingle Steffen Mauceri
William Keely
Josh Laughner
Christopher W. O’Dell
Steven Massie
Robert Nelson
David Baker
Matthäus Kiel
Otto Lamminpää
Jonathan Hobbs
Abhishek Chatterjee
Tommy Taylor
Paul Wennberg
Sean Crowell
Britton Stephens
Vivienne H. Payne
Uncertainty‐Aware Machine Learning Bias Correction and Filtering for OCO‐2: 1
Earth and Space Science
title Uncertainty‐Aware Machine Learning Bias Correction and Filtering for OCO‐2: 1
title_full Uncertainty‐Aware Machine Learning Bias Correction and Filtering for OCO‐2: 1
title_fullStr Uncertainty‐Aware Machine Learning Bias Correction and Filtering for OCO‐2: 1
title_full_unstemmed Uncertainty‐Aware Machine Learning Bias Correction and Filtering for OCO‐2: 1
title_short Uncertainty‐Aware Machine Learning Bias Correction and Filtering for OCO‐2: 1
title_sort uncertainty aware machine learning bias correction and filtering for oco 2 1
url https://doi.org/10.1029/2025EA004328
work_keys_str_mv AT steffenmauceri uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT williamkeely uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT joshlaughner uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT christopherwodell uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT stevenmassie uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT robertnelson uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT davidbaker uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT matthauskiel uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT ottolamminpaa uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT jonathanhobbs uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT abhishekchatterjee uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT tommytaylor uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT paulwennberg uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT seancrowell uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT brittonstephens uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21
AT viviennehpayne uncertaintyawaremachinelearningbiascorrectionandfilteringforoco21