[Contents]
[Next]
[Previous]
Enhancing COMET Case Studies For Use In A Diverse Community
4. Methodology
4.1 Philosophy for Selecting COMET Cases
A case is chosen by COMET staff with input from the COMET
Residence Program Advisory Panel and former/current COMET
instructors. COMET employs three main considerations when
choosing to publish a case: the ease of which the case can be
"built"; the meteorological significance of the case; and its
contribution to a well-rounded library. Members of the
operational community have expressed a growing interest in
obtaining case studies that are from recent events, because
they contain the most up-to-date weather observing and
forecast systems, and SOOs and forecasters can conduct
original research on the data.
4.2 Building a Case Study at COMET
Once the case is selected, the case study meteorologist
gathers the data from a variety of sources. These sources
include data archive centers, real-time Family of Services
data accessed from the NWS Telecommunications Gateway and
commercial data providers. The data are then organized by
type.
The data from the WSR-88D radars, including Level III NIDS
products and NOWRAD composite radar images, and from the
multi-spectral geostationary (GOES 8 and GOES 9) satellites
are converted into McIDAS "AREA" format files. Storing the
files in this format allows users to view the data with a
variety of software tools (i.e., McIDAS, GEMPAK, and
NAWIPS). The gridded model data from NCEP are stored as GEMPAK
files, pre-generated "metafiles" for use with NAWIPS software
are included. Data from the following models are normally
included: Medium Range Forecast (MRF), the Eta, NCEP's Nested
Gridded, the Rapid Update Cycle (RUC), the RUC Surface
Analysis, and the Aviation model. Surface and upper air
observations as well as atmospheric wind profiler data are
also stored as GEMPAK files. Textual weather data products are
kept in raw ASCII format. When available, data from mesoscale
models and non-traditional data sources (e.g. ACARS, local
mesonets) are included in the case study.
Other components of building the case study include: data
review, documentation, quality control, case release, and
finally, the case distribution.
4.3 Current COMET Case Studies
As of this writing (September, 1996), four cases have been
released and are available at no cost to NWS and university
users. These cases are: 11-14 March 1993, the East Coast
"Storm of the Century"; 06 December 1995, a Mid-West Snow
Storm; 03 August 1995, Hurricane Erin; and 05 May 1996, Bow
Echo over Paducah, KY.
Figure 1 is an infrared
satellite image from 11 March 1993 and is an example of
imagery available in the case study data. A description of
these cases and ordering information is available at
<http://www.joss.ucar.edu/codiac/>. Go to the research
project page and click on the COMET Case Study of
interest. The descriptions include references (if applicable),
and users are encouraged to contribute to the case study
information.
4.4 Loading data into CODIAC
The CODIAC system consists of three main sections
(see Figure 2): the
data management system; the WWW based user interface;
and the CODIAC server. The data management system contains
metadata for each data set that describes the data within that
data set and the location of the data. The actual data files may
be available from disk on-line or may be stored off-line on
archive tapes. The metadata is stored in several relational
databases that are accessed through the Empress Relational
Database Management System. For each case study and for each
data set in the case studies, the metadata consists of textual
descriptions, time and latitude/longitude ranges of the data, a
description of the sensor or platform, the parameters contained
in each data set, and various other information relevant to the
data set.
The WWW user interface, which consists of object-oriented CGI
programs, allows users to search for desired data sets, view
the metadata that describes each data set, browse data with
graphical plots and images, and retrieve desired subsets of
the data sets either over the Internet by FTP or by removable
media. The CODIAC server receives requests from the WWW user
interface CGI programs and passes them on to the data
management system. When the data management system responds to
the request, the answer returns to the WWW user interface
through the CODIAC server. For information on the evolution of
and workings of the CODIAC system, see McGuirk
(1991) and
Brazille (1992,
1993,
1996).
Because CODIAC deals mainly with metadata, it offers both
flexibility and extensibility. New data sets and new browse
capabilities can easily be added. There are several steps
involved in adding COMET case study data to the CODIAC
system. First, each case study must be split into data sets
that are groupings of similar data. Second, metadata to
describe each data set must be gathered. The metadata must
include a title and description of the data set, the time
range that the data covers and the frequency of the
observations, the latitude/longitude range of the data, the
parameters the data contains, and information on how to access
and browse the data. The final step is to load the metadata
into CODIAC. Once that has been accomplished, the data set is
available through the CODIAC WWW user interface for searching,
browsing, and retrieving.
This page was Webified by
Ethan Davis
Questions or comments can be sent to
<edavis@unidata.ucar.edu>.
This page was last updated on
.
