Use of Decision Support Systems

Integrated Weather-Air Traffic Management (ATM) Decision Support Systems provide some interesting challenges in development and benefits assessment because there is a mixture of aviation weather product development and explicit modeling of some aspects of the aviation system itself.

Conventional Approach to Aviation Decision Making

In the “conventional” approach to aviation decision making shown in Figure 1 (below), the human users must determine the impact of the weather on the ATC system before a decision can be made. If decision support tools (Slideshow 1), such as the Cooperative Route Coordination Tool (CRCT) in the Enhanced Traffic Management System (ETMS), are used to aid in the execution of the conventional operational decision loop, the user must provide results from the impact assessment phase (e.g., Flow Constrained Areas) to the tools in order to take advantage of the decision guidance they provide. Although this approach has been successful in a number of specific applications, it is becoming clear that it is not adequate overall, since often the task of determining the weather impact manually can be extremely difficult.

Use of Decision support systems - figure 1Figure 1: Overall convective weather impact mitigation process. The TMU workload associated with convective weather management includes all 5 elements shown in the “operational decision loop.”

Integrated Weather–ATM Decision Support System Approach

The new direction (e.g., as exemplified by NextGen initiatives) is toward the development of integrated Weather–ATM decision support systems as shown in Figure 2 (below). In this approach, the weather products go directly into an explicit model of some element of the aviation system, which is embedded in a decision support system. At a minimum, this integrated Weather–ATM system then determines the air traffic control impact of the weather (e.g., as in the Route Availability Planning Tool [RAPT]) and may (depending on the degree of sophistication) generate mitigation plans, decide on a mitigation plan, and assist in the execution of the plan. Other non-integrated ATM decision support tools may assist in the decision elements that are not accomplished by the Weather–ATM system, but the majority of the complex analysis is provided by the integrated system.

User support systems figure 2Figure 2: Decision process for use of weather products with an integrated Wx–ATM system.

In a “conventional” ATM aviation weather decision making process, the developer generally worries only about the weather products and not the fidelity of the human user’s model of the aviation system. With integrated systems, there must be more attention to both the weather product (e.g., forecast) quality and the aviation system model.

Key Operational Users

This important difference between integrated and non-integrated Weather–ATM systems suggests a somewhat different approach to benefits quantification. At the outset, identification of the operational users most likely to generate the benefits and the tailoring of both the forecasts and training to facilitate improved end-user decision making is crucial. The experience with CIWS, ITWS and RAPT showed the importance of identifying as many of the key decision makers as possible. In addition, with an integrated Weather–ATM system, it will be very important to involve operational users who have an interest in NAS system design, since it will be important for them to articulate what is needed to improve the aviation system model features. Thus, particularly for the initial product development, one wants both users with a significant operational need, as well as an interest in system design.

Determining meteorological and/or operational situations that should generate significant benefits is also very important. For example, we suggested that the CTAS TMA testing consider some explicit experimentation to better determine the benefits in a highly sheared wind environment.

Weather–ATM System Training

As in the case of “conventional” aviation weather decision support systems, the quality of the training that is made available to the users is key to the effective use of the tools. First of all the training needs to contain advice to the end users on how to properly apply the output of the Weather–ATM system to the operational decisions that they make on a daily basis. Next, the training material needs to describe the basis of the combined system (that is, both the approach by which the weather products are generated and the aviation system model that is an implicit part of the Weather–ATM system). Finally, the training information needs to provide a clear explanation on how to use the display.

Given the complexities of the basis for the integrated Weather–ATM system, this training should include face-to-face training by a subject matter expert (SME) on the operational use of the integrated system so that the user can ask questions.

Avoidable Delay Issues

An additional element of the “baseline” analysis is determining how much of the existing weather delay is “avoidable” and to what extent the “avoidable” delay is being reduced by other systems. The issues here are identical to those for convectional aviation weather decision support systems, that is, how much of the potential benefits associated with reductions in delay can be attributed to a specific system. Determining how much of the total delay is “avoidable” requires a NAS model that considers the capacities of terminals, en route sectors, and the optimal allocation of flights to the available capacity. Research is underway to develop such a model.

Normalizing Performance Metrics

Another important topic is how to normalize the performance metrics (e.g., delays) to deal with issues such as

  • finding comparable convective weather events (e.g., spatial patterns and time history of significant convective weather),
  • compensating for changes in the NAS (e.g., demand, traffic mix, operational procedures, airline scheduling and automation capabilities), and
  • determining the influence of other weather decision support systems [e.g., Collaborative Convective Forecast Product (CCFP) and the decision system that makes use of the CCFP].

Even if it is believed that the operational benefits of the Weather–ATM system can be estimated by objective metrics, such as those used to assess TMA performance in the Free Flight studies, when a system is introduced we would strongly recommend that there also be real-time observations made in situ within the user facilities; this is the approach that was used in assessing the benefits for CIWS. Although user inputs on their operational needs presumably were obtained in the initial system design, our experience has been that it is difficult a priori to anticipate all of the adverse aviation weather situations that would be of concern for an integrated system.

top of page