Assessment of Vertical Resolution's Effect in the Intercomparison of Temperature Profiles from Hyperspectral Infrared Sounders and GPS Radio Occulation

With recent advancements in space remote sensing technology, the fields of atmospheric and oceanic sciences have been increasingly utilizing satellite data to push forward the boundaries of research. Specifically for climate applications, the potential of satellite systems to provide stable, accurate and systematic observations has become a large interest. Though satellite observations such as infrared radiances or GPS radio occultation (RO) refractivities inherently have less uncertainties than derived satellite products such as temperature, derived products commonly give more physical insight into the Earth's atmospheric state than the observations. It follows that the validation of such derived satellite products should be highlighted as well. By inter-comparing the products of various satellite instruments, the error characteristics of these products can be identified.

GPS RO, due to its stability and measurement principle that is fundamentally different from that of radiometric instruments, can offer an independent dataset for comparison against infrared sounding retrievals. Unlike the radiosonde network, which has been traditionally used for sounder validation purposes, GPS RO is unbiased to have more samples over land or ocean, has a more continuous temporal and spatial range, and provides more coverage in the upper troposphere and lower stratosphere.

By spatiotemporally matching individual GPS RO profiles to sounder profiles from different instruments and platforms, various comparisons of sounder retrieval products and GPS RO products can be made. Monthly temperature profile comparisons involving NASA AIRS version 5 and 6, IDPS CrIMSS, NOAA and EUMETSAT version 6 IASI, CDAAC COSMIC, and CDAAC GRAS for a global and 5 latitude zones have been made. Advancement from previous work includes the use of sounder averaging kernels to examine smoothing errors and account for the different vertical resolution of the sounder and GPS RO profiles. For an example case study in the Antarctic winter, application of the averaging kernel to both the GPS RO and sounder profiles is performed in an investigation of error types.