Comparison of AIRS and IASI co-located radiances for cold scenes

摘要

Calibration of infrared radiometers at cold scene temperatures is very difficult. But high accuracy even at cold temperatures is critical for establishing a climate-quality data record. This paper describes the comparison of radiances from two sensors—the Infrared Atmospheric Sounding Interferometer (IASI) and the Atmospheric Infrared Sounder (AIRS) for cold scenes. We compare thirty-two months of IASI and AIRS data for Dome Concordia, which is on a high plateau in Antarctica and thus provides a source of nearly uniform dry scenes with a temperature range from about 190 K to about 250 K. Located on this plateau is a research station, an operational automated weather station that provides ground truth. The AIRS LIB and IASI L1C radiometric calibrations agree for large spatial and temporal averages of data taken over 32 months at Dome Concordia at the 200 mK level, in spite of large differences in the instrument implementations. However, both AIRS LIB and IASI L1C data show scene-temperature-dependent differences as large as IK, which appear to be calibration artifacts that are only partially understood. In the case of AIRS L1B spectra, some of the effects will be corrected in the forthcoming release of the L1C data. In addition, the IASI quality flag identifies a disproportionate number of spectra in the 240-250 K brightness temperature range as "low quality". Uncorrected calibration artifacts and quality flag related issues are likely to be of significance for climate applications, where 100 mK absolute accuracy is required. Both effects create sampling biases, which cannot be decreased by massive data averaging. The effects are small compared to the absolute radiometric calibration accuracy requirements of AIRS or IASI, but both will need to be accounted for in the radiometric accuracy analysis of future instruments specifically designed for climate research.