TCCON > TCCON Data and Network Policies > TCCON Requirements

TCCON Requirements

The Total Carbon Column Observing Network (TCCON) is a network of ground-based Fourier Transform Spectrometers that record direct solar spectra in the near-infrared. From these spectra, accurate and precise column-averaged abundances of atmospheric constituents including CO2, CH4, N2O, HF, CO, H2O, O2 and HDO, are retrieved.

These column observations, in combination with the existing surface measurements, will improve estimates of surface fluxes of greenhouse gases allowing improved predictions of their future concentration, and ultimately climate. In addition to their direct use for carbon flux studies, TCCON measurements are used to validate satellite column measurements.

This page is a description of our current network organization and best practice for TCCON measurements. Both of these items are understood to be subject to change as the network grows and we learn more about our measurements and the scientific requirements.

This page contains the following sections:


TCCON Database

A defined dataset of X(gases) and supporting data sufficient to allow the calculation of the tropospheric average mixing ratio (e.g. averaging kernels, a priori profiles, XHF) will be archived in a central TCCON database. All co-Is will have access to all data in the TCCON archive. Use of these data is contingent upon reciprocation; all PI's agree to make their own data available to the TCCON community.

Archived data will include flags to identify which data are deemed of sufficient quality for specific purposes. Data which pass all defined quality control (QC) tests (i.e. data which remain unflagged (flag =0) at the end of the QC procedure) are deemed suitable for public release. Quality control criteria may evolve in time as our characterisation of measurement errors improves. The QC criteria applied will be documented for each public release. PIs are encouraged to release their data to the public as soon as possible, but no later than one year after acquisition.

Each TCCON site PI is responsible for archiving sufficient raw data and ancillary information at their own site to allow the data to be reprocessed when necessary.

TCCON Data Processing

Individual TCCON sites are responsible for processing their collected interferograms in a standardized procedure to ensure consistency across the network.  Standard processing should use the current release versions of the relevant software packages and requires:

  1. transformation of interferograms to spectra using I2S, which reads OPUS- and slice-format interferograms, computes spectra using standardized phase correction, Fourier transformation and DC correction algorithms, and writes OPUS-format spectra and
  2. fitting of spectra using GGG to retrieve total column amounts and derived output data.  TCCON standards (spectral windows, levels, correction factors, vmrs, etc.) are documented on the wiki. The results are delivered in a *.eof.csv file is which specifies the site, date, and current version of the software is indicated in a standardized manner.

Following release of new software versions, each site shall reprocess and deliver their data to the TCCON archive within 4 months. 

TCCON Calibration

Calibration of the TCCON facilities follows the description identified in Wunch et al. (AMT, 2010).  Members are encouraged to obtain in situ profiles as often as possible and to the highest possible altitudes to assure network accuracy and lack of bias.


FTS Requirements

  • a minimum spectral range of 4000-9000 cm-1.
  • Maximum OPD >= 45 cm with phase resolution of at least 1 cm-1.
  • FTS sun tracker pointing with an accuracy of 1 mrad (~0.05º, or 3 arcminutes). This tracker requirement corresponds to an airmass error of 0.2% at 63º SZA, or 0.1% at 45º SZA. This is less strict than the other requirements, because of the practical limitations of available trackers, and may limit overall accuracy. 0.2 mrad, or 0.01º, corresponds to 0.1% airmass error at 80º SZA, and would be ideal. Tracker accuracy should be routinely monitored in one dimension using the solar-telluric shift (S-G shift).
  • surface pressure measurements accurate better than 0.3 mbar.
  • surface temperature measurements accurate better than 1 K.
  • accurate knowledge and reporting of the interferogram ZPD crossing time (within 1 sec).
  • laser sampling errors less than 0.00024 of the sample step. This corresponds to a ghost-parent intensity ratio of 0.0001 for parent frequency 4150 cm-1 (NDACC HF filter) and a ghost-parent ratio of 0.00014 for parent frequency of 5970 cm-1 (TCCON ghost filter). The corresponding bias in XCO2 is estimated to be 0.02% or less.
  • routine procedures for monitoring instrument line shape (ILS):
    • a line shape monitoring device in the solar beam during all or a continually repeated subset of observations capable of characterizing the modulation efficiency to at least 10% (which implies 0.1% O2 column).
    • monthly ILS retrieval from HCl cell spectra acquired with the tungsten lamp source and analysed with Frank Hase's LINEFIT code using the TCCON ILS parameterisation. A SNR of ~2500 is recommended for ILS retrieval from cell spectra, and required for manned sites. A relaxed SNR criteria of ~1300 (typically achieved with ~20 scans) is acceptable for container deployments with limited duration lamp operation.
  • criteria for optical alignment, expressed in terms of ME and phase error, to be determined
    • Provisional: The ME shall vary by less than 5% over the 0 to 45 cm OPD, and shall be known to better than 2%.
    • TBD: What is an acceptable limit for the phase error?

Instruments demonstrating the necessary precision to date have been Bruker 120/125 HRs, with CaF2 beamsplitters and room-temperature DC-enabled InGaAs detectors, and with DC-enabled Si detectors for the O2 A-band if equipped for dual-beam acquisition.

Examples of instrument hardware, which are already part of the network, can be found on the page Bruker Settings.

TCCON Data Delivery

For all sites, retrieved vertical column abundances of CO2, O2, CO, N2O, CH4, H2O, HDO and HF using validated software must be provided to the TCCON archive in the agreed format. Additional diagnostics as required by the Steering Committee will also be delivered.

Optional requirements

Site characteristics

For satellite validation, a relatively uniform surroundings (in terms of altitude, surface albedo), remote from local pollution is preferable. Additionally, co-location with tall tower measurements, regular low-altitude profile measurements, ancillary meteorological and aerosol measurements are desirable. To maintain the high quality TCCON standard an on-site technical support and active scientific support is strongly recommended.


All PIs are encouraged to calibrate their measurements with aircraft flights whenever/wherever possible. To improve the data retrieval, all working groups are encouraged to compare results from experimental software versions and different retrieval approaches with the TCCON standard product.


Particular settings are recommended to achieve the required precision of the data and those for the TCCON sites can be found on the Bruker Settings page.  The abbreviations follow the names of the parameters used in OPUS, the software delivered by Bruker.

Tag page