i
Toward Improved Use of GOES Satellite-Derived Winds at the National Centers for Environmental Prediction ( NCEP)
-
2003
-
Details:
-
Personal Author:
-
Corporate Authors:
-
NOAA Program & Office:
-
Description:Satellite derived winds have been used in the NCEP global data assimilation system since 1979. These winds provide valuable information for numerical model initialization over regions where conventional observations are not available. For example, the European Center for Medium-Range Weather Forecasts (ECM WF) (Tomassini et al.,1999, Kelly et al., 2000), U.K. Met. Office (Butterworth et al., 2000), and Australian Bureau of Meteorology (Le Marshall et al., 2000) found a positive impact from satellite-derived winds on routine operational forecasts. Some studies have reported a positive impact on hurricane track forecasts (Soden et al., 2000, Bhatia et al., 2000, Evans, 2000, Le Marshall, 2000). However, the imperfect wind data have the potential to degrade the quality of model forecasts. For example, the ECMWF and U.K. Met. Office found satellite winds had a negative impact on forecasts when they switched from assimilating low density to high density winds (Butterworth et al., 2000, Kelly et al., 2000). Various quality indicators have been developed and are distributed along with the observations as the guides to select better quality data in the data assimilation system. The Recursive Filter Flag ( RFF), developed by Cooperative Institute for Meteorological Studies(CIMSS) (Hayden and Purser, 1995), is currently used as a quality mark in the satellite winds produced at the National Environmental Satellite, Data & Information Services (NESDIS). The Quality Indicator (QI), developed by European Organization for the Exploitation of Meteorological Satellites (EUM ETSAT ) (Holmlund, 1998), is currently used as a quality mark in the satellite wind data produced at EU METSAT. These quality marks can be used to select which data are used in the analysis system. For instance, ECMWF and U .K. Met Office use d QI to filter EUMETSATE satellite wind data (Kelly a nd Rohn, 2000; Butterworth and Ingle by, 2000). Holmlund et al. (2000) use a combination of the RFF and QI to filter GOES satellite winds and have shown some improvements in the forecast skill. With satellite winds, it is not only necessary to account for random error, but also spatially correlated errors. When clusters of winds are derived from similar cloud systems by the same tracking technique, spatially correlated errors are likely. These problems have stimulated many efforts toward making 3 better use of the data and reducing correlated errors, especially for the high density winds. Both the ECM WF and the U.K. Met. Office try to reduce correlated errors in the winds by thinning the data (Kelly and Rohn, 2000; Butterworth and Ingleby, 2000). The results fr om Kelly and Rohn ( 2000) showed that thinning high density EUMETSAT satellite wind data(based on higher QI values) improved model forecast skill. However, the results from Butterworth and Ingleby (2000) showed that high density EUMETSAT satellite winds have a negative impact on the forecast regardless of thinning. In this study, we use two NCEP low resolution (T62L28) and one higher resolution (T170L42) assimilation systems in a study of thinning algorithms for high density GOES cloud-drift infrared and cloud-top water vapor winds. In the following sections, we will describe the satellite wind data quality, the thinning algorithms, and the experimental results"--Introduction.
-
Content Notes:Xiujuan Su, John Derber, Steve Lord, Christopher S. Velden, Jaime Daniels.
"May 2003."
System requirements: Adobe Acrobat Reader.
Includes bibliographical references (pages 18-20).
-
Keywords:
-
Series:
-
Document Type:
-
Place as Subject:
-
Rights Information:Public Domain
-
Compliance:Library
-
Main Document Checksum:
-
Download URL:
-
File Type: