ISCCP Cloud Algorithm Intercomparison

W. B. Rossow NASA Goddard Space Flight Center, Institute for Space Studies, New York, NY 10025

Search for other papers by W. B. Rossow in
Current site
Google Scholar
PubMed
Close
,
F. Mosher National Severe Storm Forecast Center, NOAA, Kansas City, MO 64106

Search for other papers by F. Mosher in
Current site
Google Scholar
PubMed
Close
,
E. Kinsella MA Com Sigma Data Inc., NASA/GISS, New York, NY 10025

Search for other papers by E. Kinsella in
Current site
Google Scholar
PubMed
Close
,
A. Arking NASA Goddard Space Flight Center, Greenbelt, MD 20771

Search for other papers by A. Arking in
Current site
Google Scholar
PubMed
Close
,
M. Desbois Laboratoire de Meteorologie Dynamique du CNRS, 91128 Palaisseau Cedex 05, France

Search for other papers by M. Desbois in
Current site
Google Scholar
PubMed
Close
,
E. Harrison NASA Langley Research Center, Hampton, VA 23665

Search for other papers by E. Harrison in
Current site
Google Scholar
PubMed
Close
,
P. Minnis NASA Langley Research Center, Hampton, VA 23665

Search for other papers by P. Minnis in
Current site
Google Scholar
PubMed
Close
,
E. Ruprecht NASA Goddard Space Flight Center, Institute for Space Studies, New York, NY 10025
Institute fur Geophysik und Meteorologie. Universitat zu Koln, 5 Koln 41, Federal Republic of Germany

Search for other papers by E. Ruprecht in
Current site
Google Scholar
PubMed
Close
,
G. Seze Laboratoire de Meteorologie Dynamique du CNRS, 91128 Palaisseau Cedex 05, France

Search for other papers by G. Seze in
Current site
Google Scholar
PubMed
Close
,
C. Simmer Los Alamos National Laboratory, Los Alamos, NM 87545

Search for other papers by C. Simmer in
Current site
Google Scholar
PubMed
Close
, and
E. Smith Department of Meteorology, Florida State University, Tallahassee, FL 32036

Search for other papers by E. Smith in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

The International Satellite Cloud Climatology Project (ISCCP) will provide a uniform global climatology of satellite-measured radiances and derive an experimental climatology of cloud radiative properties from these radiances. A pilot study to intercompare cloud analysis algorithms was initiated in 1981 to define a state-of-the-art algorithm for ISCCP. This study compared the results of applying six different algorithms to the same satellite radiance data. The results show that the performance of all current algorithms depends on how accurately the clear sky radiances are specified; much improvement in results is possible with better methods for obtaining these clear-sky radiances. A major difference between the algorithms is caused by their sensitivity to changes in the cloud size distribution and optical properties: all methods, which work well for some cloud types or climate regions, do poorly for other situations. Therefore, the ISCCP algorithm is composed of a series of steps, each of which is designed to detect some of the clouds present in the scene. This progressive analysis is used to retrieve an estimate of the clear sky radiances corresponding to each satellite image. Application of a bispectral threshold is then used as the last step to determine the cloud fraction. Cloudy radiances are interpreted in terms of a simplified model of cloud radiative effects to provide some measure of cloud radiative properties. Application of this experimental algorithm to produce a cloud climatology and field observation programs to validate the results will stimulate further research on cloud analysis techniques as part of ISCCP.

Abstract

The International Satellite Cloud Climatology Project (ISCCP) will provide a uniform global climatology of satellite-measured radiances and derive an experimental climatology of cloud radiative properties from these radiances. A pilot study to intercompare cloud analysis algorithms was initiated in 1981 to define a state-of-the-art algorithm for ISCCP. This study compared the results of applying six different algorithms to the same satellite radiance data. The results show that the performance of all current algorithms depends on how accurately the clear sky radiances are specified; much improvement in results is possible with better methods for obtaining these clear-sky radiances. A major difference between the algorithms is caused by their sensitivity to changes in the cloud size distribution and optical properties: all methods, which work well for some cloud types or climate regions, do poorly for other situations. Therefore, the ISCCP algorithm is composed of a series of steps, each of which is designed to detect some of the clouds present in the scene. This progressive analysis is used to retrieve an estimate of the clear sky radiances corresponding to each satellite image. Application of a bispectral threshold is then used as the last step to determine the cloud fraction. Cloudy radiances are interpreted in terms of a simplified model of cloud radiative effects to provide some measure of cloud radiative properties. Application of this experimental algorithm to produce a cloud climatology and field observation programs to validate the results will stimulate further research on cloud analysis techniques as part of ISCCP.

Save