Preprints
https://doi.org/10.5194/amt-2023-224
https://doi.org/10.5194/amt-2023-224
20 Nov 2023
 | 20 Nov 2023
Status: a revised version of this preprint was accepted for the journal AMT and is expected to appear here in due course.

Fast retrieval of XCO2 over East Asia based on the OCO-2 spectral measurements

Fengxin Xie, Tao Ren, Changying Zhao, Yuan Wen, Yilei Gu, Minqiang Zhou, Pucai Wang, Kei Shiomi, and Isamu Morino

Abstract. The increase in greenhouse gas concentrations, particularly CO2, has significant implications for global climate patterns and various aspects of human life. Spaceborne remote sensing satellites play a crucial role in high-resolution monitoring of atmospheric CO2. However, the next generation of greenhouse gas monitoring satellites is expected to face challenges such as low retrieval efficiency and insufficient retrieval accuracy. To address these challenges, this study focuses on enhancing the retrieval of column-averaged dry air mole fraction of carbon dioxide (XCO2) using spectral data from the OCO-2 satellite. A novel approach based on neural network (NN) models is proposed to tackle the nonlinear inversion problems associated with XCO2 retrieval. The study employs a data-driven supervised learning method and explores two distinct training strategies. Firstly, training is conducted using experimental data obtained from the inversion of traditional optimization models, which are released as the OCO-2 satellite products. Secondly, training is performed using a simulated dataset generated by an accurate forward calculation model. The inversion and prediction performance of the machine learning model for XCO2 is compared, analyzed, and discussed for the observed region. The results demonstrate that the model trained on simulated data accurately predicts XCO2 in the target area. Furthermore, when compared to OCO-2 satellite product data, the developed XCO2 retrieval model achieves rapid predictions (<1 ms) with high precision (2 ppm or approximately 0.5 %). The accuracy of the machine learning model's retrieval results is validated against reliable data from TCCON sites, demonstrating its capability to capture CO2 seasonal variations and annual growth trends effectively.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
Fengxin Xie, Tao Ren, Changying Zhao, Yuan Wen, Yilei Gu, Minqiang Zhou, Pucai Wang, Kei Shiomi, and Isamu Morino

Status: closed

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on amt-2023-224', Steffen Mauceri, 14 Dec 2023
    • AC1: 'Reply on RC1', Tao Ren, 28 Jan 2024
  • RC2: 'Comment on amt-2023-224', Anonymous Referee #2, 03 Jan 2024
    • AC2: 'Reply on RC2', Tao Ren, 28 Jan 2024

Status: closed

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on amt-2023-224', Steffen Mauceri, 14 Dec 2023
    • AC1: 'Reply on RC1', Tao Ren, 28 Jan 2024
  • RC2: 'Comment on amt-2023-224', Anonymous Referee #2, 03 Jan 2024
    • AC2: 'Reply on RC2', Tao Ren, 28 Jan 2024
Fengxin Xie, Tao Ren, Changying Zhao, Yuan Wen, Yilei Gu, Minqiang Zhou, Pucai Wang, Kei Shiomi, and Isamu Morino
Fengxin Xie, Tao Ren, Changying Zhao, Yuan Wen, Yilei Gu, Minqiang Zhou, Pucai Wang, Kei Shiomi, and Isamu Morino

Viewed

Total article views: 415 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
265 125 25 415 17 13
  • HTML: 265
  • PDF: 125
  • XML: 25
  • Total: 415
  • BibTeX: 17
  • EndNote: 13
Views and downloads (calculated since 20 Nov 2023)
Cumulative views and downloads (calculated since 20 Nov 2023)

Viewed (geographical distribution)

Total article views: 420 (including HTML, PDF, and XML) Thereof 420 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 23 May 2024
Download
Short summary
This study demonstrates a new machine learning approach to efficiently and accurately estimate atmospheric carbon dioxide levels from satellite data. Rather than traditional complex physics-based retrieval methods, neural network models are trained on simulated data to rapidly predict CO2 concentrations directly from satellite spectral measurements.