Comparisons of IASI-A and AATSR measurements of top-of-atmosphere radiance over an extended period

Abstract

Abstract. This study examines the trustworthiness of the Advanced Along-Track Scanning Radiometer (AATSR) and the Infrared Atmospheric Sounding Interferometer (IASI-A), as on-orbit reference instruments that are useful in re-calibrating the Advanced Very High Resolution Radiometer (AVHRR) series (Mittaz and Harris, 2011). To do this, a 39-month period (1 January 2008 to 31 March 2011) of AATSR and IASI-A inter-comparisons of top-of-atmosphere (TOA) radiance measurements is examined. Our inter-comparison reveals features of the AATSR and IASI-A bias with respect to scan angle, scene temperature, time and orbital maneuvers, and gives insight into their trustworthiness as an in-orbit reference instruments. The first feature that our study reveals is that the AATSR (nadir view) and IASI-A are both stable (have no perceptible trends in the period of study). The second feature is that IASI-A is perhaps more accurate ( ∼ 0.05 K) than its stated accuracy (0.5 K). In fact the AATSR and IASI-A bias is close to the AATSR pre-launch bias (plus a small offset of +0.07 K) implying that IASI-A can get close to pre-launch levels of accuracy. Third, a very small scan angular dependence of AATSR and IASI-A bias indicates that the IASI-A response vs. scan angle algorithm is robust, while the instrument is in orbit. Inter-comparisons of AATSR with IASI-A further reveal the impact of orbital maneuvers of the ENVISAT, the platform carrying AATSR, done in October 2011 and not anticipated previously. Our study reveals that this maneuver introduced a temperature-dependent bias in the AATSR measurements for low temperatures (< 240 K) in the period following this maneuver (Cocevar et al., 2011). Our study also shows that the known AATSR 12 µm channel offset is in fact temperature dependent, grows up to 0.4 K, varies seasonally and is correlated with instrument temperature and cannot be corrected by shifting the spectral response function (SRF) of AATSR. We also present a set of recommendations to help identify the parameters under which these instruments can provide the most trustworthy observations for the AVHRR re-calibration.