Articles | Volume 10, issue 4
Research article
20 Apr 2017
Research article |  | 20 Apr 2017

Intercomparability of XCO2 and XCH4 from the United States TCCON sites

Jacob K. Hedelius, Harrison Parker, Debra Wunch, Coleen M. Roehl, Camille Viatte, Sally Newman, Geoffrey C. Toon, James R. Podolske, Patrick W. Hillyard, Laura T. Iraci, Manvendra K. Dubey, and Paul O. Wennberg

Abstract. The Total Carbon Column Observing Network (TCCON) has become the standard for long-term column-averaged measurements of CO2 and CH4. Here, we use a pair of portable spectrometers to test for intra-network bias among the four currently operating TCCON sites in the United States (US). A previous analytical error analysis has suggested that the maximum 2σ site-to-site relative (absolute) bias of TCCON should be less than 0.2 % (0.8 ppm) in XCO2 and 0.4 % (7 ppb) in XCH4. We find here experimentally that the 95 % confidence intervals for maximum pairwise site-to-site bias among the four US TCCON sites are 0.05–0.14 % for XCO2 and 0.08–0.24 % for XCH4. This is close to the limit of the bias we can detect using this methodology.

Short summary
Two portable spectrometers, assumed to be internally precise, were taken to four different sites with (stationary) TCCON spectrometers. Biases of column averaged CO2 and CH4 measured among the TCCON sites were estimated experimentally. Results suggest that maximum (95 % confidence interval) bias among sites is less than what was estimated from a previous analytical error analysis.