A "Demand Pull" Approach to Short Term Forecast Development and Testing
For short term aviation weather forecasts, in an era of significant federal government and airline budget austerity for civil aviation investments, it is becoming increasingly important to quantitatively demonstrate the benefits of the improved aviation weather forecasts. If such a benefits demonstration is in fact going to be an integral element of the overall short term aviation forecast development process as opposed to something that suddenly arises as an important issue after the forecast has been developed and deployed, then one needs to give some thought as to how the benefits assessment will proceed and whether it is likely to be successful.
In this paper, we will consider some key elements that need to be considered if this is to be an “demonstrated user benefits driven” overall short term aviation forecast process:
1. What is the overall decision process for the effective use of the short term forecast if it is to have the desired quantifiable user benefit? For example, in the case of products intended to reduce aviation weather delays, one need to understand carefully the overall decision process that is involved in the user taking actions that will result in a reduction of delays. We discuss two specific short term aviation weather forecasts – convection and ceiling – to illustrate the issues that arise in thinking about the overall decision support system needed to generate benefits. 2. What is the preexisting “baseline” of short term aviation forecasts/decision processes that exists already to address the user needs? In most cases, there are already various weather information sources that can be viewed as providing a short term forecast (e.g., a CWSU meteorologist, persistence or, animation loops of the past weather). How well do we understand how the “baseline” forecast and the associated user decision support system operates? How will the new forecast and its decision support compare? What about training (especially if the new forecast is rather different than the “baseline”)? 3. How will we measure the change in system performance? For example, if the new forecast claims to help reduce delays and/or accidents, how will one address differences in the weather between the “before” and “after” time periods? How will one determine whether the new forecast is in fact the key factor if there was a change? 4. The paper concludes with some suggestions for development and testing of new 0-6 hour aviation forecasts to improve safety and reduce delays.