Automatic classification and compression of GMS cloud imagery in heavy rainfall monitoring application

Li, Jun; Fengxian, Zhou; Wang, Luyi

doi:10.1007/bf02677078

Cited by 2 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If auxiliary data is not available or in certain situations (such as in the presence of snow) the cloud mask may not be of good quality, a simple visible-infrared box classification can be used for initial classification (Li and Zhou 1990). Also, if the previous near time classification center values were stored as training or reference data, these center values could also be used for initial classification based on the Bayesian decision method (Li et al 1992). Future work includes more case studies, especially in polar regions and African deserts where the surfaces may have a very unique appearance in the MODIS imagery.…”

Section: Discussionmentioning

confidence: 99%

“…MODIS atmospheric and surface parameter retrievals require cloud free measurements (Li et al 2001a), while cloud type information such as single/multi-layer or high/medium/low cloud information will greatly benefit cloud parameter retrievals (Frey et al 1999;Li et al 2001b) and the derivation of cloud motion vectors (Velden et al 1997). Cloud classification can also improve the monitoring of deep convective clouds and rainfall estimation from IR cloud imagery data (Li et al 1992(Li et al , 1993. MODIS cloud information can further the International Satellite Cloud Climatology Program (ISCCP) that was stimulated by research on several methods of cloud classification that have been tested in a systematic algorithm intercomparison (Rossow et al 1985).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

High-Spatial-Resolution Surface and Cloud-Type Classification from MODIS Multispectral Band Measurements

Menzel

Yang

et al. 2003

J. Appl. Meteor.

Self Cite

View full text Add to dashboard Cite

A method for automated classification of surface and cloud types using Moderate-Resolution Imaging Spectroradiometer (MODIS) radiance measurements has been developed. The MODIS cloud mask is used to define the training sets. Surface and cloud type classification is based on the maximum likelihood (ML) classification method. Initial classification results define training sets for subsequent iterations. Iterations end when the number of pixels switching classes becomes smaller than a predetermined number or when other criteria are met. The mean vector in the spectral and spatial domain within a class is used for class identification and a final 1km-resolution classification mask is generated for such field of view in a MODIS granule. This automated classification refines the output of the cloud mask algorithm, and enables further applications such as clear atmospheric profile or cloud parameter retrievals from MODIS and Atmospheric Infrared Sounder (AIRS) radiance measurements. The advantages of this method are that the automated surface and cloud type classifications are independent of radiance or brightness temperature threshold criteria, and that the interpretation of each class is based on the radiative spectral characteristics of different classes. This paper describes the ML classification algorithm and presents daytime MODIS classification results. The classification results are compared with the MODIS cloud mask, visible images, infrared window images, and other observations for an initial validation.

show abstract

Section: Discussionmentioning

confidence: 99%