Home | About Us | Contacts | Help | Feedback | SEARCH Global | Australia | NSW | Vic. | Qld | WA | SA | Tas. | ACT | NT | Ant. | Weather & Warnings | Hydrology | Climate | Numerical Prediction | About Services | Learn About Meteorology | Registered User Services |

Global Guide
Main Menu

Global Guide to Tropical Cyclone Forecasting:
CHAPTER 6: OPERATIONAL STRATEGY

6.4 WORKSTATIONS

A computer based workstation in every TCWC was a dream even recently during the first IWTC, but this is fast becoming realisable as personal computers become increasingly more sophisticated and less expensive. Workstations are becoming affordable even to modestly budgeted weather services, and more than that, they are becoming absolutely essential to the efficient running of a forecasting office.

Around 50% of countries surveyed at IWTC-II did not have access to personal computers within their forecasting office. It is hoped that this number will shrink rapidly within the next few years.

Initiatives taken primarily by the United States and Australia should mean that individual countries will not need to expend precious resources in designing and developing a workstation with applications for tropical cyclone forecasting.

The U.S. Navy and Air Force Joint Typhoon Warning Center (JTWC) requested the Naval Environmental Prediction Research Facility (NEPRF) to develop such a system. The Naval Automated Tropical Cyclone Forecasting System (ATCF) has been used operationally at JTWC since 1988 (Miller et al., 1990). Features of the system, which will run on IBM-XT or higher personal computers, include:

1. database management system,
2. overlaying of graphical fields,
3. looping of graphical fields,
4. CLIPER and climatology models for track prediction guidance,
5. field analysis capabilities,
6. friendly user interface.

The Australian Tropical Cyclone Workstation (ATCW) is currently being developed with flexibility in mind, both in hardware requirements and in the level of sophistication of operation. Features planned to be incorporated into the ATCW include:

1. high resolution map library for tropical cyclone basins,
2. integrated map-cyclone track display, with colour icon track and fix-type identifiers and capability to display distance and bearing of cyclone from map features,
3. automated verification scheme,
4. warning preparation capability,
5. numerical models eg storm surge model, barotropic model,
6. other integrated applications, including strike probability and CLIPER forecasts, and wind-radii estimation,
7. incorporation of expert systems, including several of the decision trees in this Guide.

The development of artificial intelligence (AI) and expert systems has important consequences for meteorology including tropical cyclone forecasting. An overview of the applications of expert systems to weather forecasting is given in Conway (1989), while an introduction to expert systems is found in Jackson (1986) and Hayes-Roth et al. (1983). More fundamental reading on AI is given in Winston (1984).

Expert systems are a method of using computers to perform tasks normally requiring human judgement, and for this reason offer a new dimension to the task of tropical cyclone forecast decision making. Certain tropical cyclone events may occur relatively infrequently and can therefore transcend the experience of forecasters. Consulting a suitable expert system may provide the necessary guidance to compensate for any lack of experience in a TCWC. Quite apart from considerations of experience, forecasting rules and aids can be hard to find or even neglected in the stress of operational forecasting. A properly designed expert system resident on a workstation can be (either actively or passively) consulted, particularly in times of indecision. Information and "rules" can be input and adjusted locally so as to be region specific.

The introduction to a cyclone forecast office of a personal-computer workstation designed specifically for tropical cyclone applications, and incorporating a user-friendly interface appears to be an effective way of bringing about a significant advance in operational forecasting efficiency, globally, in a relatively short time. Such a system will give forecasters access to sophisticated technology, state of the art forecasting techniques, and increased efficiency in message preparation.

Another example of workstation technology already being used successfully in forecast offices and in the first instance by the NHC in Miami in 1983 (Sheets, 1990) is the Man-computer Interactive Data Access System (McIDAS), which was developed by the University of Wisconsin (Suomi et al., 1983). In Australia, the McIDAS system has been modified to run operationally on a personal computer (Le Marshall et al., 1987) and has been a prominent forecasting tool in its three TCWCs since 1988, particularly with respect to the analysis of satellite information. Capabilities of this system include the processing, display, enhancement and animation of satellite data, and the overlaying of data as well of fields from numerical models. Significant computing power is required to operate this system, however.

Contents Chapter 6.5