ECMWF COPERNICUS REPORT
Copernicus Atmosphere Monitoring Service
Validation report for the CO2 fluxes estimated by atmospheric inversion, FT18r1 Version 1.0
Issued by: CEA / Frédéric Chevallier
Date: 11/07/2019
REF.: CAMS73_2018SC1_D73.4.6.3-‐2019-‐v1_201907_Validation FT inversion CO2
fluxes_v1
Copernicus Atmosphere Monitoring Service
Author 2 of 14
This document has been produced in the context of the Copernicus Atmosphere Monitoring Service (CAMS). The activities leading to these results have been contracted by the European Centre for Medium-‐Range Weather Forecasts, operator of CAMS on behalf of the European Union (Delegation Agreement signed on 11/11/2014). All information in this document is provided "as is" and no guarantee or warranty is given that the information is fit for any particular purpose. The user thereof uses the information at its sole risk and liability. For the avoidance of all doubts, the European Commission and the European Centre for Medium-‐Range Weather Forecasts has no liability in respect of this document, which is merely representing the authors view.
Copernicus Atmosphere Monitoring Service
Author 3 of 14
Contributors
CEA Frédéric Chevallier
Copernicus Atmosphere Monitoring Service
Author 4 of 14
Table of Contents
1. Introduction 5
2. Inversion configuration 5
3. Assimilated data 6
4. Evaluation 7
4.1 Fit to the unassimilated surface measurements 7 4.2 Fit to unassimilated aircraft measurements 8 4.3 Fit to TCCON GGG2014 9
Acknowledgements 11
References 11
Copernicus Atmosphere Monitoring Service
Author 5 of 14
1. Introduction The inversion system that generates the CAMS global CO2 atmospheric inversion product is called PyVAR. It has been initiated, developed and maintained at CEA/LSCE within CAMS and its precursor projects GEMS/MACC/MACC-‐II/MACC-‐III (Chevallier 2019a, and references therein). Here, we synthesize the evaluation of version FT18r1 that was released in July 2019. Version FT18r1 covers the period between September 2014 and December 20181 and is constrained by satellite retrievals from the second Orbiting Carbon Observatory (OCO-‐2). ‘FT’ stands for Fast Track, since the satellite observations are available faster than most surface measurements: it is planned to update and extend this inversion faster and more often than the surface-‐driven inversion. The presentation of the evaluation procedure is primarily based on the fit of the inversion posterior simulation to large databases of atmospheric observations: ObsPack Globalview+ v4.2.1 of Cooperative Global Atmospheric Data Integration Project (2019), ObsPack NRT v4.4.1 of NOAA Carbon Cycle Group ObsPack Team (2019), ObsPack INPE_RESTRICTED v2.0 of NOAA Carbon Cycle Group ObsPack Team (2018) and the Total Carbon Column Observing Network (TCCON) GGG2014 archive (Wunch et al. 2011). In all cases we compare the results with those obtained by the latest CAMS surface-‐driven inversion, v18r2 (Chevallier, 2019b). More scientific detail can be found in the study that prepared FT18r1 (Chevallier et al., 2019). Section 2 describes the PyVAR-‐CO2 configuration that was used to assimilate the retrievals described in Section 3. Section 4 presents the evaluation synthesis.
2. Inversion configuration The transport model in PyVAR-‐CO2 is the global general circulation model LMDz in its version LMDz6A (Remaud et al. 2018), that uses the deep convection model of Emanuel (1991). This version has a regular horizontal resolution of 3.75o in longitude and 1.875o in latitude, with 39 hybrid layers in the vertical. The inferred fluxes are estimated in each horizontal grid point of the transport model with a temporal resolution of 8 days, separately for day-‐time and night-‐time. The state vector of the inversion system is therefore made of a succession of global maps with 9,200 grid points. Per month it gathers 73,700 variables (four day-‐time maps and four night-‐time maps). It also includes a map of the total CO2 columns at the initial time step of the inversion window in order to account for the uncertainty in the initial state of CO2. The prior values of the fluxes combine estimates of (i) gridded annual anthropogenic emissions (EC-‐JRC/PBL EDGAR version 4.2, CDIAC and GCP), (ii) monthly ocean fluxes (Landschützer et al. 20182), 3-‐ 1 Observations after December 2018 are used to constrain the year 2018 better, but fluxes for those months are not publicly distributed. 2 This database covers the period 1982-‐2017. We use the monthly values for the year 2017 after it.
Copernicus Atmosphere Monitoring Service
Author 6 of 14
hourly (when available) or monthly biomass burning emissions (GFED 4.1s until 2016 and GFAS afterwards) and climatological 3-‐hourly biosphere-‐atmosphere fluxes taken as the 1989-‐2010 mean of a simulation of the ORganizing Carbon and Hydrology In Dynamic EcosystEms model (ORCHIDEE, Krinner et al. 2005), version 1.9.5.2. The mass of carbon emitted annually during specific fire events is compensated here by the same annual flux of opposite sign representing the re-‐growth of burnt vegetation, which is distributed regularly throughout the year. The gridded prior fluxes exhibit 3-‐hourly variations but their inter-‐annual variations over land are only caused by anthropogenic emissions. This feature was explicitly demanded by some users who wanted the interannual signals in the inverted natural fluxes to be strictly driven by the atmospheric measurements. Over land, the errors of the prior biosphere-‐atmosphere fluxes are assumed to dominate the error budget and the covariances are constrained by an analysis of mismatches with in situ flux measurements (Chevallier et al. 2006, 2012): temporal correlations on daily mean Net Carbon Exchange (NEE) errors decay exponentially with a length of one month but night-‐time errors are assumed to be uncorrelated with daytime errors; spatial correlations decay exponentially with a length of 500 km; standard deviations are set to 0.8 times the climatological daily-‐varying heterotrophic respiration flux simulated by ORCHIDEE with a ceiling of 4 gC·∙m-‐2 per day. Over a full year, the total 1-‐sigma uncertainty for the prior land fluxes amounts to about 3.0 GtC·∙yr-‐1. The error statistics for the open ocean correspond to a global air-‐sea flux uncertainty about 0.5 GtC·∙yr-‐1 and are defined as follows: temporal correlations decay exponentially with a length of one month; unlike land, daytime and night-‐time flux errors are fully correlated; spatial correlations follow an e-‐folding length of 1000 km; standard deviations are set to 0.1 gC·∙m-‐2 per day. Land and ocean flux errors are not correlated.
3. Assimilated data OCO-‐2 is a NASA satellite that was launched in July 2014 (Eldering et al. 2017). It orbits around the Earth from pole to pole with a local crossing time at the Equator in the early local afternoon. It carries a spectrometer that measures the sunlight reflected by the Earth and its atmosphere in the near-‐infrared/ shortwave infrared spectral regions, with high spectral resolution (>~ 20,000) such that individual gas absorption lines are resolved. OCO-‐2 provides spatially dense data with a narrow swath and with footprints of a few km2. We use NASA’s Atmospheric CO2 Observations from Space (ACOS) bias-‐corrected retrievals of the column-‐average CO2 dry air-‐mole fraction (XCO2), version 9 (O’Dell et al., 2018; Kiel et al., 2019) from September 2014 until April 2019. We add a trend of +0.1 ppm/a to the retrievals further to some advice from the ACOS team (O’Dell, personal communication, 2019) and we remove a +0.2 ppm offset based on our own assessment of the retrievals. Our mean correction is therefore about zero on average over the first two years of scientific measurement and slightly positive afterwards. To reduce data volume without loss of information at the scale of a global model, glint and nadir OCO-‐2 retrievals have been averaged in 10-‐s bins following the approach defined by the Model Intercomparison Project (MIP) of OCO-‐2 (Crowell et al., 2019). The retrieval averaging kernels, prior
Copernicus Atmosphere Monitoring Service
Author 7 of 14
profiles and Bayesian uncertainty are accounted for in the assimilation. The interpolation procedure between the model vertical grid and the retrieval grid is described in Section 2.2 of Chevallier, 2015). We also introduce a transport uncertainty term, based on the variability across several models at the OCO-‐2 sounding locations (Crowell et al., 2019). Last, in order to account for likely correlations between the transport model errors at the sub-‐grid scale, we de-‐weigh the binned retrievals that fall within a same grid box for a same orbit by inflating the assigned error variance (σ2) by the number of retrievals in the box. Maps of the coverage of the OCO-‐2 retrievals are shown in O’Dell et al. (2018). We only consider “good” retrievals as identified by variable xco2_quality_flag. Though the ocean biases in OCO-‐2 have been substantially reduced since the initial version 7 (O’Dell et al. 2018), initial inversion tests using OCO-‐2 ocean observations still produced highly unrealistic results (annual global ocean sinks about 5 GtC·∙a-‐1, to be compared with the much smaller state-‐of-‐the-‐art estimates in Le Quéré et al., 2018) and are hence left out of this work (as are retrievals over inland water or over mixed land-‐water surfaces).
4. Evaluation We have run the LMDz global transport model using the surface fluxes and the initial CO2 state inferred by the inversion as boundary conditions and now compare it with various independent observations.
4.1 Fit to the unassimilated surface measurements We first consider the surface measurements assimilated in the CAMS surface-‐driven inversion, v18r2 (Chevallier, 2019b). Figure 1 shows the posterior root mean square (RMS) and bias of the model-‐minus-‐measurement difference (for FT18r1) as a function of the corresponding error statistics that we have assigned at each assimilated data in v18r2. Measurement error is negligible here and the assigned error statistics refer to transport model errors and to representation errors. As expected, the inversion fits the data less well than the inversion that assimilated them (Figure 3 of Chevallier, 2019b), but the RMS is still overall better than the assigned observation uncertainty. Biases are usually less than 2 ppm in absolute value. Sites CIB (Centro de Investigacion de la Baja Atmosfera, ES), ELL (Estany Llong, ES), TIK (Hydrometeorological Observatory of Tiksi, RU), UTA (Wendover, Utah, US) and TAP (Tae-‐ahn Peninsula, KR) show larger RMS differences than the other sites. The mean bias at station Mauna Loa, Hawaii (MLO) is -‐0.5 ppm. Over the four years 2015-‐2018, FT18r1 fits the annual trend of globally-‐averaged marine measurements (http://www.esrl.noaa.gov/gmd/ccgg/trends/, access 5 June 2019) with a bias of -‐0.02 ppm/a and a standard deviation3 of 0.11 ppm/a 4.
3 With 4 samples only, the standard deviation estimate has hardly any statistical meaning. 4 We assume a conversion factor of 2.086 GtC·∙ppm-‐1, from Prather (2012).
Copernicus Atmosphere Monitoring Service
Author 8 of 14
Figure 1 -‐ Statistics of the differences between the posterior inversion simulation and individual surface measurements as a function of the assigned observation error standard deviation for each measurement site. The statistics cover the full assimilation period, starting in September 2014 and including the first months of 2019.
4.2 Fit to unassimilated aircraft measurements Following the approach defined in Chevallier et al. (2019), we now focus on the continuous or flask dry air mole fraction measurements made by aircraft in the free troposphere. The free troposphere is simply defined here as the atmospheric layer between 2 and 7 km above sea level. The measurements are all from ObsPack Globalview+ v4.2.1, NRT v4.4.1 and INPE_RESTRICTED v2.0 for the period 2014-‐2018. All model equivalents to individual Globalview+ v4.2.1 data are publicly available from http://dods.lsce.ipsl.fr/invsat/CAMS/FT18r1_GV+4.2.1.txt or on request to copernicus-‐[email protected]. Apart from the Amazonian campaigns, biases do not exceed 0.6 ppm for FT18r1 (Figure 2). FT18r1 and v18r2 have usually similar biases. Larger differences between the two inversions are seen for the campaigns at Rarotonga, Cook Islands (RTA), at Pantanal, Brazil (PAN), for the Comprehensive Observation Network for TRace gases by AIrLiner (CON) and at Alaska Coast Guard (ACG), where FT18r1 performs less well by a few tenths of ppm 5. FT18r1 clearly outperforms v18r2 at Park Falls, Wisconsin (LEF). There is no obvious latitudinal trend (even when reporting the sign of the biases), and therefore no obvious flaw of the model vertical mixing (Stephens et al., 2007). Standard deviations are very similar between the two inversions. When taking all free tropospheric aircraft campaigns together, irrespective of their data number, the posterior simulation deviates from the
5 We do not count MRC here because this campaign in Pennsylvania includes 7 measurements only at our altitudes.
Copernicus Atmosphere Monitoring Service
Author 9 of 14
measurements by 0.0±1.5 ppm (bias ± standard deviation), which is within the specification (key performance indicator) of the CAMS CO2 inversion. Figure 2 -‐ Model-‐minus-‐observation absolute differences and standard deviations per measurement program for FT18r1 and v18r2. The number of measurement per site, campaign or program varies between 7 (MRC) and 290,361 (ACT). The programs are ranked by increasing mean latitude (North is on the right), irrespective of their latitudinal coverage (which is large of several tens of degrees for ORC, TOM and CON). These mean latitudes are shown in the middle of the panel. The statistics cover the period December 2014-‐December 2018.
4.3 Fit to TCCON GGG2014 Figure 3 shows the misfit statistics for the column retrievals at each TCCON station. Results for versions FT18r1 (OCO-‐2-‐driven) and v18r2 (surface-‐driven) are both displayed. For the comparison, the model has been convolved with the retrieval averaging kernels. All available TCCON station records are shown for the sake of completeness, but sites Pasadena, JPL and Paris are located in urban areas that are not well represented at the horizontal resolution of the transport model (3.75o in longitude and 1.875o in latitude): in this case the statistics logically show large negative model biases about -‐1 ppm. Apart from these urban stations, we also note large absolute biases (about 1 ppm) for the OCO-‐2-‐driven inversions at the European and Canadian stations, while the surface-‐driven inversion performs slightly better there. For the other sites, no inversion product clearly outperforms the other one. In non-‐urban sites, the standard deviation is usually about 1 ppm, but it reaches 2 ppm at the Zugspitze mountain site for both inversions. We note that the model usually fits TCCON retrievals better than the satellite retrievals presented by Wunch et al. (2017).
Copernicus Atmosphere Monitoring Service
Author 10 of 14
Figure 3 -‐ Statistics of the difference between the posterior model and individual TCCON measurements, ordered by increasing latitude indices in the LMDz model. A site may appear several times if several instruments have been used over time there. The statistics cover the period September 2014 – January 2019.
Copernicus Atmosphere Monitoring Service
Author 11 of 14
Acknowledgements The author is very grateful to the many people involved in the surface and aircraft CO2 measurements and in the archiving of these data that were kindly made available to him by various means. The OCO-‐2 data have been obtained from http://co2.jpl.nasa.gov. They were produced by the OCO-‐2 project at the Jet Propulsion Laboratory, California Institute of Technology. This work benefited from the kind help and advice from the OCO-‐2 Science Team. TCCON data were obtained from the TCCON Data Archive, operated by the California Institute of Technology from the website at http://tccon.ornl.gov/. Obspack data were obtained from https://www.esrl.noaa.gov/gmd/ccgg/obspack/. Mass fluxes for the LMDz transport model have been provided by M. Remaud. Some of this work was performed using HPC resources of DSM-‐CCRT and of CCRT under allocation A0050102201 made by GENCI (Grand Équipement National de Calcul Intensif).
References Chevallier, F., N. Viovy, M. Reichstein, and P. Ciais: On the assignment of prior errors in Bayesian inversions of CO2 surface fluxes. Geophys. Res. Lett., 33, L13802, doi:10.1029/2006GL026496, 2006. Chevallier, F., T. Wang, P. Ciais, F. Maignan, M. Bocquet, A. Arain, A. Cescatti, J.-‐Q. Chen, H. Dolman, B. E. Law, H. A. Margolis, L. Montagni, and E. J. Moors: What eddy-‐covariance flux measurements tell us about prior errors in CO2-‐flux inversion schemes. Global Biogeochem. Cy., 26, GB1021, doi:10.1029/2010GB003974, 2012. Chevallier, F.: On the statistical optimality of CO2 atmospheric inversions assimilating CO2 column retrievals, Atmos. Chem. Phys., 15, 11133-‐11145, https://doi.org/10.5194/acp-‐15-‐11133-‐2015, 2015. Chevallier, F., Description of the CO2 inversion production chain. CAMS deliverable CAMS73_2018SC1_D73.5.2.1-‐2019_201904_CO2 inversion production chain_v1. http://atmosphere.copernicus.eu/, 2019a. Chevallier, F., Validation report for the CO2 fluxes estimated by atmospheric inversion, v18r2. CAMS deliverable CAMS73_2018SC1_D73.1.4.1-‐2018-‐v1_201907. http://atmosphere.copernicus.eu/, 2019b. Chevallier, F., Remaud, M., O'Dell, C. W., Baker, D., Peylin, P., and Cozic, A.: Objective evaluation of surface-‐ and satellite-‐driven CO2 atmospheric inversions, Atmos. Chem. Phys. Discuss., https://doi.org/10.5194/acp-‐2019-‐213, in review, 2019. Cooperative Global Atmospheric Data Integration Project. (2019). Multi-‐laboratory compilation of atmospheric carbon dioxide data for the period 1957-‐2017; obspack_co2_1_GLOBALVIEWplus_v4.2.1_2019-‐05-‐20 [Data set]. NOAA Earth System Research Laboratory, Global Monitoring Division. https://doi.org/10.25925/20190520
Copernicus Atmosphere Monitoring Service
Author 12 of 14
Crowell, S., Baker, D., Schuh, A., Basu, S., Jacobson, A. R., Chevallier, F., Liu, J., Deng, F., Feng, L., Chatterjee, A., Crisp, D., Eldering, A., Jones, D. B., McKain, K., Miller, J., Nassar, R., Oda, T., O'Dell, C., Palmer, P. I., Schimel, D., Stephens, B., and Sweeney, C.: The 2015–2016 Carbon Cycle As Seen from OCO-‐2 and the Global In Situ Network, Atmos. Chem. Phys. Discuss., https://doi.org/10.5194/acp-‐2019-‐87, in review, 2019. Eldering, A., Wennberg, P., Crisp, D., Schimel, D., Gunson, M., Chatterjee, A., Liu, J., Schwandner, F., Sun, Y., and O’Dell, C.: The Orbiting Carbon Observatory-‐2 early science investigations of regional carbon dioxide fluxes, Science, 358, eaam5745, 2017. Emanuel, K.: A Scheme for Representing Cumulus Convection in Large-‐Scale Models, J. Atmos. Sci., 48, 2313–2329, doi:10.1175/1520-‐0469(1991)048<2313:ASFRCC2.0.CO;2, 1991. Kiel, M., O'Dell, C. W., Fisher, B., Eldering, A., Nassar, R., MacDonald, C. G., and Wennberg, P. O.: How bias correction goes wrong: measurement of XCO2 affected by erroneous surface pressure estimates, Atmos. Meas. Tech., 12, 2241-‐2259, https://doi.org/10.5194/amt-‐12-‐2241-‐2019, 2019. Krinner, G., Viovy, N., de Noblet-‐Ducoudré, N., Ogée, J., Polcher, J., Friedlingstein, P., Ciais, P., Sitch, S., and Prentice, I. C.: A dynamic global vegetation model for studies of the coupled atmosphere-‐biosphere system, Global Biogeochem. Cy., 19, GB1015, doi:10.1029/2003GB002199, 2005. Landschützer, P., N. Gruber and D.C.E. Bakker. 2015. An observation-‐based global monthly gridded sea surface pCO2 product from 1982 onward and its monthly climatology (NCEI Accession 0160558). Version 4.4. NOAA National Centers for Environmental Information. Dataset. https://doi.org/10.7289/V5Z899N6. [2019-‐03-‐27] Le Quéré, C., Andrew, R. M., Friedlingstein, P., Sitch, S., Hauck, J., Pongratz, J., Pickers, P. A., Korsbakken, J. I., Peters, G. P., Canadell, J. G., Arneth, A., Arora, V. K., Barbero, L., Bastos, A., Bopp, L., Chevallier, F., Chini, L. P., Ciais, P., Doney, S. C., Gkritzalis, T., Goll, D. S., Harris, I., Haverd, V., Hoffman, F. M., Hoppema, M., Houghton, R. A., Hurtt, G., Ilyina, T., Jain, A. K., Johannessen, T., Jones, C. D., Kato, E., Keeling, R. F., Goldewijk, K. K., Landschützer, P., Lefèvre, N., Lienert, S., Liu, Z., Lombardozzi, D., Metzl, N., Munro, D. R., Nabel, J. E. M. S., Nakaoka, S.-‐I., Neill, C., Olsen, A., Ono, T., Patra, P., Peregon, A., Peters, W., Peylin, P., Pfeil, B., Pierrot, D., Poulter, B., Rehder, G., Resplandy, L., Robertson, E., Rocher, M., Rödenbeck, C., Schuster, U., Schwinger, J., Séférian, R., Skjelvan, I., Steinhoff, T., Sutton, A., Tans, P. P., Tian, H., Tilbrook, B., Tubiello, F. N., van der Laan-‐Luijkx, I. T., van der Werf, G. R., Viovy, N., Walker, A. P., Wiltshire, A. J., Wright, R., Zaehle, S., and Zheng, B.: Global Carbon Budget 2018, Earth Syst. Sci. Data, 10, 2141-‐2194, https://doi.org/10.5194/essd-‐10-‐2141-‐2018, 2018. NOAA Carbon Cycle Group ObsPack Team (2018): INPE atmospheric carbon dioxide data for the period 2015-‐2017; obspack_co2_1_INPE_RESTRICTED_v2.0_2018-‐11-‐13; NOAA Earth System Research Laboratory, Global Monitoring Division. http://dx.doi.org/10.25925/20181030. NOAA Carbon Cycle Group ObsPack Team. (2019). Multi-‐laboratory compilation of atmospheric carbon dioxide data for the years 2018-‐2019; obspack_co2_1_NRT_v4.4.1_2019-‐05-‐31 [Data set].
Copernicus Atmosphere Monitoring Service
Author 13 of 14
NOAA Earth System Research Laboratory, Global Monitoring Division. https://doi.org/10.25925/20190531 O'Dell, C. W., Eldering, A., Wennberg, P. O., Crisp, D., Gunson, M. R., Fisher, B., Frankenberg, C., Kiel, M., Lindqvist, H., Mandrake, L., Merrelli, A., Natraj, V., Nelson, R. R., Osterman, G. B., Payne, V. H., Taylor, T. E., Wunch, D., Drouin, B. J., Oyafuso, F., Chang, A., McDuffie, J., Smyth, M., Baker, D. F., Basu, S., Chevallier, F., Crowell, S. M. R., Feng, L., Palmer, P. I., Dubey, M., García, O. E., Griffith, D. W. T., Hase, F., Iraci, L. T., Kivi, R., Morino, I., Notholt, J., Ohyama, H., Petri, C., Roehl, C. M., Sha, M. K., Strong, K., Sussmann, R., Te, Y., Uchino, O., and Velazco, V. A.: Improved retrievals of carbon dioxide from Orbiting Carbon Observatory-‐2 with the version 8 ACOS algorithm, Atmos. Meas. Tech., 11, 6539-‐6576, https://doi.org/10.5194/amt-‐11-‐6539-‐2018, 2018. Remaud, M., Chevallier, F., Cozic, A., Lin, X., and Bousquet, P.: On the impact of recent developments of the LMDz atmospheric general circulation model on the simulation of CO2 transport, Geosci. Model Dev., 11, 4489-‐4513, https://doi.org/10.5194/gmd-‐11-‐4489-‐2018, 2018. Stephens, B. B., et al.: Weak northern and strong tropical land carbon uptake from vertical profiles of atmospheric CO2, Science, 316, 1732–1735, 2007. Wunch, D., Toon, G. C., Blavier, J.-‐F. L., Washenfelder, R. A., Notholt, J., Connor, B. J., Griffith, D. W. T., Sherlock, V., and Wennberg, P. O., The Total Carbon Column Observing Network, Phil. Trans. R. Soc. A:2011369 2087-‐2112, doi10.1098/rsta.2010.0240, 2011. Wunch, D., Wennberg, P. O., Osterman, G., Fisher, B., Naylor, B., Roehl, C. M., O'Dell, C., Mandrake, L., Viatte, C., Kiel, M., Griffith, D. W. T., Deutscher, N. M., Velazco, V. A., Notholt, J., Warneke, T., Petri, C., De Maziere, M., Sha, M. K., Sussmann, R., Rettinger, M., Pollard, D., Robinson, J., Morino, I., Uchino, O., Hase, F., Blumenstock, T., Feist, D. G., Arnold, S. G., Strong, K., Mendonca, J., Kivi, R., Heikkinen, P., Iraci, L., Podolske, J., Hillyard, P. W., Kawakami, S., Dubey, M. K., Parker, H. A., Sepulveda, E., García, O. E., Te, Y., Jeseck, P., Gunson, M. R., Crisp, D., and Eldering, A.: Comparisons of the Orbiting Carbon Observatory-‐2 (OCO-‐2) XCO2 measurements with TCCON, Atmos. Meas. Tech., 10, 2209-‐2238, https://doi.org/10.5194/amt-‐10-‐2209-‐2017, 2017.
Copernicus Atmosphere Monitoring Service
atmosphere.copernicus.eu copernicus.eu ecmwf.int
ECMWF - Shinfield Park, Reading RG2 9AX, UK Contact: [email protected]