Retrieval of cirrus ice crystal sizes from 8.3 and 11.1 mu m emissivities determined by the improved initialization inversion of TIROS-N Operational Vertical Sounder observations

Citation
Cj. Stubenrauch et al., Retrieval of cirrus ice crystal sizes from 8.3 and 11.1 mu m emissivities determined by the improved initialization inversion of TIROS-N Operational Vertical Sounder observations, J GEO RES-A, 104(D24), 1999, pp. 31793-31808
Citations number
36
Categorie Soggetti
Earth Sciences
Volume
104
Issue
D24
Year of publication
1999
Pages
31793 - 31808
Database
ISI
SICI code
Abstract
The improved initialization inversion (31) algorithms convert TIROS-N Opera tional Vertical Sounder (TOVS) observations from the NOAA polar orbiting en vironmental satellites into atmospheric temperature and water vapor profile s as well as cloud and surface properties. Because of their relatively high spectral resolution, infrared vertical sounders are especially useful for the identification of cirrus clouds. Differences in cirrus emissivity betwe en the wavelengths 8.3 and 11.1 mu m are used to retrieve ice crystal size; the radiative transfer model is based on the anomalous diffraction approxi mation applied to different crystal morphologies. We present sensitivity st udies of ice crystal sizes estimated on a global scale to uncertainty facto rs in the retrieval as well as to assumptions in the model. On average, cir rus ice crystal mean maximum dimensions lie between 80 and 150 mu m or effe ctive ice crystal sizes between 35 and 45 mu m. Correlations between estima ted cirrus ice crystal mean maximum dimensions and cloud-top temperature se em to be positive in the tropics and midlatitude winter but depend on assum ed temperature-dependent ice crystal morphology and size distribution funct ion. These estimates can be helpful for the evaluation of general circulati on models. With satellite measurements, one estimates mean ice crystal size s mostly on the top of the cirrus clouds. However, when the clouds are thin ner, the IR sounder can reach deeper into the cloud, Yet a quantitative rel ation between cloud thickness (effective cloud emissivity) and retrieval he ight inside the cloud has still to be investigated.