Articles | Volume 14, issue 12
Research article
03 Dec 2021
Research article |  | 03 Dec 2021

Assessing the feasibility of using a neural network to filter Orbiting Carbon Observatory 2 (OCO-2) retrievals at northern high latitudes

Joseph Mendonca, Ray Nassar, Christopher W. O'Dell, Rigel Kivi, Isamu Morino, Justus Notholt, Christof Petri, Kimberly Strong, and Debra Wunch

Data sets


TCCON data from Bremen (DE) J. Notholt, C. Petri, T. Warneke, N. M. Deutscher, M. Palm, M. Buschmann, C. Weinzierl, R. C. Macatangay, and P. Grupe

TCCON data from Bialystok (PL) N. M. Deutscher, J. Notholt, J. Messerschmidt, C. Weinzierl, T. Warneke, C. Petri, and P. Grupe

TCCON data from Sodankylä (FI) R. Kivi, P. Heikkinen, and E. Kyrö

TCCON data from Ny Ålesund, Spitsbergen (NO) J. Notholt, T. Warneke, C. Petri, N. M. Deutscher, C. Weinzierl, M. Palm, and M. Buschmann

TCCON data from East Trout Lake, SK (CA) D. Wunch, J. Mendonca, O. Colebatch, N. T. Allen, J.-F. Blavier, S. Roche, J. Hedelius, G. Neufeld, S. Springett, D. Worthy, R. Kessler, and K. Strong

TCCON data from Eureka (CA) K. Strong, S. Roche, J. E. Franklin, J. Mendonca, E. Lutsch, D. Weaver, P. F. Fogal, J. R. Drummond, R., Batchelor, and R. Lindenmaier

TCCON data from Park Falls (US) P. O. Wennberg, C. M. Roehl, D. Wunch, G. C. Toon, J.-F. Blavier, R. Washenfelder, G. Keppel-Aleks, N. T. Allen, and J. Ayers

TCCON data from Rikubetsu (JP) I. Morino, N. Yokozeki, T. Matsuzaki, and M. Horikawa

Short summary
Machine learning has become an important tool for pattern recognition in many applications. In this study, we used a neural network to improve the data quality of OCO-2 measurements made at northern high latitudes. The neural network was trained and used as a binary classifier to filter out bad OCO-2 measurements in order to increase the accuracy and precision of OCO-2 XCO2 measurements in the Boreal and Arctic regions.