Remote sensing of cirrus cloud vertical size profile using MODIS data

• Published in: Journal of Geophysical Research. D. Atmospheres Vol. 114; no. D9
• Main Authors: Wang, Xingjuan, Liou, K. N., Ou, Steve S. C., Mace, G. G., Deng, M.
• Format: Journal Article
• Language: English
• Published: Washington, DC Blackwell Publishing Ltd 16.05.2009; American Geophysical Union
• Subjects: cirrus cloud; Earth sciences; Earth, ocean, space; Exact sciences and technology; remote sensing; vertical size profile; Great Plains; Far East; Asia; Southern Great Plains; China; China, People's Rep; Rayleigh scattering; Particle size; Spectroradiometry; radiative transfer; accuracy; North America; thermal emission; Radar remote sensing; instruments; noise; Optical thickness; radiance; absorption; libraries; Satellite observation; Perturbation; Ice cloud; Cloud base; corrections; Cirrus; water vapor; programs; Cloud top; Retrieval algorithm; Reflectance
• ISSN: 0148-0227; 2156-2202
• DOI: 10.1029/2008JD011327

This paper describes an algorithm for inferring cirrus cloud top and cloud base effective particle sizes and cloud optical thickness from the Moderate Resolution Imaging Spectroradiometer (MODIS) 0.645, 1.64 and 2.13, and 3.75 μm band reflectances/radiances. This approach uses a successive minimization method based on a look‐up library of precomputed reflectances/radiances from an adding‐doubling radiative transfer program, subject to corrections for Rayleigh scattering at the 0.645 μm band, above‐cloud water vapor absorption, and 3.75 μm thermal emission. The algorithmic accuracy and limitation of the retrieval method were investigated by synthetic retrievals subject to the instrument noise and the perturbation of input parameters. The retrieval algorithm was applied to three MODIS cirrus scenes over the Atmospheric Radiation Measurement Program's southern Great Plain site, north central China, and northeast Asia. The reliability of retrieved cloud optical thicknesses and mean effective particle sizes was evaluated by comparison with MODIS cloud products and qualitatively good correlations were obtained for all three cases, indicating that the performance of the vertical sizing algorithm is comparable with the MODIS retrieval program. Retrieved cloud top and cloud base ice crystal effective sizes were also compared with those derived from the collocated ground‐based millimeter wavelength cloud radar for the first case and from the Cloud Profiling Radar onboard CloudSat for the other two cases. Differences between retrieved and radar‐derived cloud properties are discussed in light of assumptions made in the collocation process and limitations in radar remote sensing characteristics.