Showing papers by "Michael R. Gunson published in 2003"
TL;DR: Constraint of the AIRS product uncertainties to within the claimed specification of 1 K/1 km over well-instrumented regions is feasible within 12 months of launch, but global validation of all AIRS/AMSU/HSB products may require considerably more time due to the novelty and complexity of this dataset.
Abstract: The Atmospheric Infrared Sounder/Advanced Microwave Sounding Unit/Humidity Sounder for Brazil (AIRS/AMSU/HSB) instrument suite onboard Aqua observes infrared and microwave radiances twice daily over most of the planet. AIRS offers unprecedented radiometric accuracy and signal to noise throughout the thermal infrared. Observations from the combined suite of AIRS, AMSU, and HSB are processed into retrievals of atmospheric parameters such as temperature, water vapor, and trace gases under all but the cloudiest conditions. A more limited retrieval set based on the microwave radiances is obtained under heavy cloud cover. Before measurements and retrievals from AIRS/AMSU/HSB instruments can be fully utilized they must be compared with the best possible in situ and other ancillary "truth" observations. Validation is the process of estimating the measurement and retrieval uncertainties through comparison with a set of correlative data of known uncertainties. The ultimate goal of the validation effort is retrieved product uncertainties constrained to those of radiosondes: tropospheric rms uncertainties of 1.0 degC over a 1-km layer for temperature, and 10% over 2-km layers for water vapor. This paper describes the data sources and approaches to be used for validation of the AIRS/AMSU/HSB instrument suite, including validation of the forward models necessary for calculating observed radiances, validation of the observed radiances themselves, and validation of products retrieved from the observed radiances. Constraint of the AIRS product uncertainties to within the claimed specification of 1 K/1 km over well-instrumented regions is feasible within 12 months of launch, but global validation of all AIRS/AMSU/HSB products may require considerably more time due to the novelty and complexity of this dataset and the sparsity of some types of correlative observations.
120 citations
TL;DR: In this paper, a case study of an orographic cirrus cloud observed on October 2, 2002, over the central Andes of South America is presented with spectral BT differences up to 63K between 998 and 811 cm−1.
Abstract: [1] The high-resolution spectra of the Atmospheric Infrared Sounder (AIRS) provide a global view of small- particle-dominated cirrus clouds, and they exist over much larger spatial extents than seen in previous aircraft campaigns. As shown by simulations using a plane-parallel scattering radiative transfer (RT) model and realistic ice particle shapes, the shape of the radiance spectra in the atmospheric windows is uniquely influenced by small ice crystals. Minima in the brightness temperature (BT) spectra between 800 to 850 cm−1 are seen for ice particles smaller than 3 μm in the RT simulations and AIRS spectra. A case study of an orographic cirrus cloud observed on October 2, 2002, over the central Andes of South America is presented with spectral BT differences up to 63K between 998 and 811 cm−1.
38 citations
TL;DR: Simulated data for synthesizing radiance measurements by the Atmospheric Infrared Sounder have been a valuable tool for developing retrieval algorithms and studying error characteristics and will continue to be so after launch.
Abstract: Models for synthesizing radiance measurements by the Atmospheric Infrared Sounder (AIRS) are described. Synthetic radiances have been generated for developing and testing data processing algorithms. The radiances are calculated from geophysical states derived from weather forecasts and climatology using the AIRS rapid transmission algorithm. The data contain horizontal variability at the spatial resolution of AIRS from the surface and cloud fields. This is needed to test retrieval algorithms under partially cloudy conditions. The surface variability is added using vegetation and International Geosphere Biosphere Programme surface type maps, while cloud variability is added randomly. The radiances are spectrally averaged to create High Resolution Infrared Sounder (HIRS) data, and this is compared with actual HIRS2 data on the NOAA 14 satellite. The simulated data under-represent high-altitude equatorial cirrus clouds and have too much local variability. They agree in the mean to within 1-4 K, and global standard deviation agrees to better than 2 K. Simulated data have been a valuable tool for developing retrieval algorithms and studying error characteristics and will continue to be so after launch.
34 citations