the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Sep 2025
Annual Growth Rates of Column-Averaged CO2 Inferred from Total Carbon Column Observing Network (TCCON)
Abstract. Monitoring annual atmospheric CO2 growth rates is a key constraint on assessing the long-term effectiveness of emission reduction strategies. We analyzed annual growth rates of column-averaged dry-air mole fractions of CO2 (XCO2) using long-term data from 12 sites within the Total Carbon Column Observing Network (TCCON), spanning four regions: the Arctic, two Northern Hemisphere midlatitude bands (40–50° N and 30–40° N), and the Southern Hemisphere. While in situ ground-based measurements provide detailed records of near-surface CO2 concentrations, XCO2 reflects the column-averaged abundance across the entire atmosphere, offering a complementary perspective.
We compared TCCON-derived growth rates with ground-based in situ observations from the Mauna Loa Observatory (MLO). Three calculation methods—Monthly Mean (MM), Fourier Fit residuals (FF), and Dynamic Linear Model (DLM)—were evaluated, with particular attention to the Eureka site, where polar night introduces substantial data gaps. In addition, the Copernicus Atmosphere Monitoring Service (CAMS) reanalysis product was used to assess consistency with TCCON-based growth rates and to evaluate each method’s robustness to missing data. Among the methods tested, the DLM approach proved most resilient to data gaps.
Regionally averaged CO2 growth rates, calculated from 2010 or from the earliest available data through 2024, ranged from approximately 2.33 to 2.40 ppm per year. The most prominent signal was associated with the 2015–2016 El Niño–Southern Oscillation (ENSO) event, during which growth rates increased by up to 1.7 ppm per year. The impact of COVID-19-related emission reductions in 2020 was also examined: a decline of 0.4 ppm per year was observed in the 30–40° N region, whereas other regions showed no significant decline. Correlation analysis between growth rates and ENSO strength revealed significant relationships in the Southern Hemisphere and at Mauna Loa, but not in northern mid- or high-latitude regions.
- Preprint
(10058 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 05 Nov 2025)
- RC1: 'Comment on egusphere-2025-4080', Anonymous Referee #1, 03 Nov 2025 reply
Data sets
2020 TCCON Data Release (Version GGG2020) Total Carbon Column Observing Network (TCCON) Team https://doi.org/10.14291/TCCON.GGG2020
Model code and software
JonasHach/dlmhelper: Pre-release of v1.0.0 Jonas Hachmeister https://doi.org/10.5281/zenodo.14772372
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 640 | 44 | 10 | 694 | 13 | 10 |
- HTML: 640
- PDF: 44
- XML: 10
- Total: 694
- BibTeX: 13
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This study provides an analysis of annual growth rates of atmospheric carbon dioxide, derived from long term (more than ten years) observations of column-averaged dry-air mole fractions of CO2 (hereafter XCO2) from 12 sites, covering four latitudinal regions: the Arctic, two midlatitude bands in the Northern Hemisphere, and the Southern Hemisphere. These measurement sites form part of the Total Carbon Column Observing Network (TCCON) of ground-based remote sensing measurements of greenhouse gases; the column concentrations are obtained from high spectral resolution Fourier Transform Infrared (FTIR) spectrometer observations of direct sunlight, by considering how much light is absorbed by the molecules of interest at specific wavelengths as it passes down through the atmosphere, and into the FTIR spectrometer via a solar tracker. The column-averaged abundances provided by TCCON are complementary to the near-surface concentrations of CO2 measured by in-situ air sampling instruments (e.g. from the NOAA Global Monitoring Laboratory network), which benefit from being able to acquire data through day and night, albeit with greater sensitivity to any local emissions sources. The XCO2 data obtained by TCCON are also widely used for validation of satellite data. All three of these data sources (in situ, TCCON, and satellite -- indirectly, via assimilation into the CAMS reanalysis) are considered by this study in their analysis of long term CO2 growth rates.
For the analysis a subset of the TCCON sites are chosen, covering four different latitudinal bands: two in Northern mid-latitudes, where the greatest impact of fossil fuel emissions is expected; one in the high Northern latitudes, to investigate CO2 growth rate in the Arctic; and one covering the whole of the Southern Hemisphere, where the greatest influence on CO2 concentration growth is expected to be transport from the Northern Hemisphere. Three sites are selected for each of these regions, with the choice determined by the availability of data over at least a five year period leading up to 2020 whilst avoiding sites located in highly urbanised regions. The decision to define the whole Southern Hemisphere as a latitudinal band is justified by the relative lack of spatial variability in CO2 in this area compared with that seen in the Northern Hemisphere. As well as the TCCON data, data from a single NOAA in situ site (Mauna Loa, commonly used for CO2 growth rate studies owing to its geographically isolated high altitude location) and the CAMS reanalysis total column CO2 product are also considered. The continuous year-round availability of the CAMS data is used in this study to assess the impacts of gaps in the TCCON data record on the calculated CO2 growth rates, particularly in the case of the Northern high latitude sites where the low solar elevation angle during winter prevents the acquisition of solar spectra for a significant fraction of the year. Three different methods for calculating the CO2 growth rate are presented and evaluated -- the Dynamic Linear Model (DLM) method is shown to be the least sensitive to gaps in the time series, as demonstrated by comparing growth rates calculated from CAMS data that has been down-sampled to the TCCON data availability for Eureka, a Northern high latitude site. This is a key finding for this study, as it demonstrates the best method to use when investigating growth rates from irregularly sampled time series data, such as that provided by satellites and other remote sensing techniques over regions of variable cloud cover.
The final part of the analysis concerns the interannual variations in CO2 growth rates (calculated solely using the DLM method, based on the conclusions of the previous section), and how these compare between the different latitudinal regions. Overall, there is broad agreement in CO2 growth rate between the four regions, despite year-to-year variability. The results show the significant impact of ENSO conditions on growth rate in all regions, with the strongest correlation between growth rate and ENSO strength occurring in the Southern Hemisphere TCCON and the Mauna Loa Observatory data. The spatial variability in correlation is likely due to differences in biospheric sensitivity to ENSO events in different parts of the world. The authors also note that the analysis is able to identify shifts in anthropogenic emissions, most notably in 2020 as a result of COVID-19 related reductions which were significant in the densely urbanised 30 to 40°N latitude band. The EDGAR emissions map show in Figure A5 confirms that the majority of anthropogenic CO2 emissions occur in this band, demonstrating that CO2 growth rates in the 30-40°N region are particularly sensitive to fossil fuel activity. The relatively small impact of the change in anthropogenic emissions during the COVID-19 affected period, compared with the influence of the strong and very strong El Nino and La Nina events, is an interesting outcome of this work which I think warrants further study.
Overall, this study provides a timely analysis of the now long-term XCO2 data available from TCCON, which is only possible now thanks to the continuous dedicated effort over the past decade and more of those operating the sites and processing the data. The study provides a robust method for calculating growth rates from irregularly sampled time series data, the selection of which is well supported by thorough testing of different methods on a down-sampled reanalysis product, which can be used with confidence on other long-term atmospheric concentration datasets where it would be of interest to determine the growth rate. As more data -- both ground-based and satellite -- is acquired over longer time periods, it will be very interesting to see similar studies applying this methodology to different regions of scientific interest. I am happy for this paper to be published, subject to the following minor technical corrections and suggestions:
Line 27: I'm not sure "monotonically" is the best word to use here, given that it implies little-to-no variation in the rate of increase of CO2 emissions -- I would suggest removing the word altogether, but emphasising that the emissions increase annually, e.g. "... emissions have generally increased each year, but with occasional declines...";
Line 109: "... in the Southern Hemisphere where CO2 growth is largely influenced by transport from the Northern Hemisphere" -- please can you provide a reference supporting this statement?
Figure 3: this figure is a little hard to interpret -- for each year, I suggest plotting the "CAMS ds" bars next to the "TCCON" bars, since I think the key point is the comparison between these two. I would then have the "CAMS non-ds" data on the right of the three for each year, so that the reader can see the effect of the down-sampling more easily (I would also explicitly label these "CAMS non-ds" instead of just "CAMS"). So in summary, for each year, plot from left to right: TCCON, CAMS ds, CAMS non-ds;
Figure A3: it would be useful for the reader to have this referred to somewhere in the main text, I think Section 3.1 (Data Preparation) would be most appropriate.