Two Decades of in situ Halocarbon Trace Gas Measurements

Motivated by the signing of the Montreal Protocol in 1987 and interests in greenhouse gases, the NOAA Halocarbons and other Atmospheric Trace Species (HATS) group (now in the ESRL/GMD laboratory) focused on frequently measuring some of the regulated ozone depleting gases such as chlorofluorocarbons (CFC-11 and CFC-12), methyl chloroform (CH3CCl3), and carbon tetrachloride (CCl4). The original HATS in situ program, the Radiatively Important Trace Species (RITS) program, measured these four gases and nitrous oxide (N2O) using gas chromatographic (GC) techniques. The RITS GCs were deployed at the NOAA baseline observatories and a cooperative research station where they remained in operation for the next 13 years. Throughout the 1990s, the HATS in situ and flask programs documented the steady decline in global growth rates of the major chlorinated solvents and chlorofluorocarbons as a result of the Montreal Protocol. Widespread use of the replacement compounds to the now banned CFCs prompted improvements to the HATS in situ program. The RITS instruments were replaced from 1998-2000 by the four-channel Chromatograph for Atmospheric Trace Species (CATS) GCs. In addition to the gases measured by RITS, the CATS GCs added nine compounds including halon-1211, methyl chloride (CH3Cl) and CFC alternatives such as HFC-142b and HCFC-22. Since the RITS instruments have been taken offline, efforts have been focused on finalizing this important data set. A number of calibration scale changes from and improvements with the data reduction algorithms have facilitated comparing and combining the RITS and CATS data sets. In particular, the RITS calibration gas concentration tables were updated to reflect the most recent calibration scale changes. Ratios of calibration gas concentrations were fitted to ratios of RITS system responses to those gases in order to derive a detector response nonlinearity factor for each compound measured. The residuals of these fits were used to simultaneously derive a set of calibration gas concentration offsets via a singular value decomposition algorithm for an over-determined linear system. The net effect was to reduce systematic errors associated with calibration cylinder changes and provide a consistent, calibrated data set. The RITS and CATS data can now be compared and combined into a 20 year in situ record.