Characterization and correction of detector nonlinearity in Fourier-transform interferograms

Abstract. Atmospheric gas products obtained from remote sensing observational networks operating groundbased FTIR interferometers such as COCCON, NDACC and TCCON are known to be sensitive to nonlinear detector response. Existing methods to correct interferograms for nonlinear detector response showed that the effect of the correction on the retrieved atmospheric gas concentrations can exceed the reported uncertainties on these data products. Several methods to correct the recorded interferograms or spectra exist but turned out to have only a limited applicability mainly due to underlying assumptions. A new nonlinearity characterization method is presented which builds upon previous methods and overcomes most assumptions on the underlying measurement setup. The method is demonstrated on five distinct nonlinearity episodes for measurements obtained from two TCCON instruments, one COCCON low-spectral-resolution instrument and one NDACC instrument and includes a study on the effect of the nonlinearity correction on the atmospheric trace gas products derived from these measurements. For the TCCON instruments the new method is compared to the standardized TCCON nonlinearity correction method. New diagnostic metrics that quantify nonlinearity strength are introduced and these are shown to have the potential to characterize the cause of the underlying detector nonlinearity response, being either saturation or optical misalignment.