Column-averaged dry-air greenhouse gas mole fractions observed at the TCCON site at Wollongong, Australia

Nicholas M. Deutscher, Voltaire A. Velazco, David W.T. Griffith, Ronald Macatangay, Clare Paton-Walsh, Nicholas B. Jones, Graham Kettlewell, Stephen R. Wilson, Debra Wunch, Steven Wofsy, (Vanessa, Rebecca, anyone else?)



Knowledge of the distribution of global sources and sinks of the greenhouse gases (GHGs) CO\(_{\mathrm{2}}\) and CH\(_{\mathrm{4}}\) is usually based on surface in situ measurements of these gases (Gurney 2002). Yang et al. (2007) and Stephens et al. (2007) have shown that inversions using only these measurements are sensitive to model parameterization of vertical mixing. On the other hand, column measurements are relatively insensitive to vertical transport (Keppel-Aleks 2011). Column measurements can be performed from ground-based (usually stationary) or satellite platforms. Satellite GHG columns from instruments such as SCIAMACHY (Burrows 1995) and GOSAT (Kuze 2009) can achieve quasi-global coverage, and with sufficient precision and lack of spatial and temporal biases can provide further information about GHG fluxes (Rayner 2001). Achieving retrievals that are free of bias is a challenge, particularly because of the influence of scattering processes on reflected sunlight, as measured by these satellite instruments (Oshchepkov 2012, Oshchepkov 2013).

The Total Carbon Column Observing Network (TCCON) is a network of high precision and accuracy ground-based Fourier Transform Infrared (FTIR) spectrometers, making direct solar absorption measurements. Direct solar absorption measurements result in a significantly higher signal than those from reflected sunlight, and therefore scattering processes make an insignificant contribution to the measurement intensity, resulting in an inherently higher measurement precision. TCCON CO\(_{\mathrm{2}}\) column measurements have been shown to have a precision better than 0.1% (Keppel-Aleks 2007, Deutscher 2010). The accuracy of TCCON is achieved via calibration against independent aircraft profiles (Deutscher 2010, Messerschmidt 2011, Wunch 2010). In addition to providing highly-precise and accurate ground-based column measurements, the TCCON serves as the primary validation for satellite retrievals and a transfer standard between satellite and surface in situ measurements.

Site description


Wollongong, Australia is situated on a narrow coastal plain bordered by the Pacific Ocean to the east and a steep sandstone precipice to the west. Wollongong, population ~300,000, is located approximately 82 kilometres south of Sydney (population ~4.6 million). The measurement site is located at the University of Wollongong (34.406\(^{\circ}\)S, 150.879\(^{\circ}\)E, 30 m above sea level). The escarpment to the west rises approximately 400 m within 2 km of the measurement site, making Wollongong a challenging site for satellite validation.


Wollongong has been a long-standing contributor to the Infrared Working Group (IRWG) of the Network for Detection of Atmospheric Composition Change (NDACC), making mid infrared (MIR) measurements from 1996 to 2008 with a Bomem DA8 spectrometer. Since 2007 this instrumentation was accompanied and then replaced by a Bruker IFS125HR of sufficient standard to perform measurements in the near infrared (NIR) for the Total Carbon Column Observing Network (TCCON). The Bruker became the sole instrument in May 2008, and from this point onwards was operated in a configuration appropriate to TCCON standards. The FTS is equipped with four detectors: a room temperature silicon diode (Si) and a room temperature Indium Gallium Arsenide (InGaAs) operating in the NIR, and cryogenically cooled Indium Antimonide (InSb) and Mercury Cadmium Telluride (MCT) detectors operating in the MIR. The measurements made in the NIR are the subject of this paper.

For the NIR measurements, the instrument follows standard TCCON configuration (Washenfelder 2006, Wunch 2011). An updated Bruker solar tracker with 80mm mirrors is used. Pressure sensors... other MET instrumentation. The automation is run by in house developed software “Oscar”, which communicates with OPUS via DDE interface under Windows.

Aerosol optical depth has been measured since 2009 using Middleton Solar SP02 Sunphotometers mounted on an active altitude - azimuth tracker as used by the Bureau of Meteorology (Mitchell 2003). The sunphotometers have been changed intermittently, and hence the wavelengths measured have altered during the measurement period. The calibration procedure involves analysis of suitable clear morning or afternoon periods (Mitchell 2003). Aerosol optical depths have been calculated using the minute observations, filtered for cloud (Mitchell 2003) and then averaged to hourly values provided 6 measurements exist in that hour (10 %).