Intercalibration between HIRS/2 and HIRS/3 channel 12 based on physical considerations

Abstract

HIRS brightness temperatures at channel 12 (T12) can be used to assess the water vapour content of the upper troposphere. The transition from HIRS/2 to HIRS/3 in 1999 involved a shift in the central wavelength of channel 12 from 6.7 μm to 6.5 μm causing a discontinuity in the time series of T12. To understand the impact of this change in the measured brightness temperatures, we have performed radiative transfer calculations for channel 12 of HIRS/2 and HIRS/3 instruments, using a large set of radiosonde profiles of temperature and relative humidity from three different sites. Other possible changes within the instrument, apart from the changed spectral response function, have been assumed to be of minor importance and in fact, it was necessary to assume as a working hypothesis that the spectral and radiometric calibration of the two instruments did not change during the relatively short period of their common operation. For each radiosonde profile we performed two radiative transfer calculations, one using the HIRS/2 channel response function of NOAA14 and one using the HIRS/3 channel response function of NOAA15, resulting in negative differences of T12 (denoted as Delta_T12 := T12/15 − T12/14) ranging between −12K and −2 K. Inspection of individual profiles for large, medium and small values of Delta_T12 pointed to the role of the middle tropospheric humidity. This guided us to investigate the relation between Delta_T12 and the channel 11 brightness temperatures which are typically used to detect signals from the middle troposphere. This allowed us to construct a correction for the HIRS/3 T12, a correction that leads to a pseudochannel 12 brightness temperature such as if a HIRS/2 instrument had measured it. By applying this correction we find an excellent agreement between the original HIRS/2 T12 and the HIRS/3 data inferred from the correction method with R = 0.986. Upper tropospheric humidity (UTH) derived from the pseudoHIRS/2 T12 data compared well with that calculated from intersatellite-calibrated data, providing independent justification for using the two intercalibrated time series (HIRS/2 and HIRS/3) as a continuous HIRS time series for long term UTH analyses.