The optimization of traffic flow in congested airspace is a highly complex problem. Although air traffic managers have several tools such as the Corridor Integrated Weather System (CIWS) and the new Consolidated Storm for Prediction Aviation (CoSPA) system to provide weather forecasts, they also need automated decision support tools to integrate flight information, trajectory models, and convective weather products to assist in rerouting flights most efficiently around convective weather cells. In particular, it is important to know when pilots will choose to deviate from their planned flight route rather than fly through convective weather.

Two Convective Weather Avoidance Models (CWAM1 and CWAM2) have been developed to predict automatically when pilots in en route airspace will choose to deviate around convective weather. The model used the CIWS high-resolution Vertically Integrated Liquid (VIL) precipitation measure and the echo tops mosaic to characterize the weather. Over 30 weather predictors such as mean and 90th percentile values were calculated from the VIL and echo tops.

CWAM2 confirmed the findings of CWAM1 that the difference between flight altitude and echo top height is the single best predictor of pilot deviation. However, it is often observed that pilots tend to avoid the region downwind of a convective cell, and will readily fly through stratiform rain that trails upwind of a convective cell. Many of the misclassified encounters from CWAM2 were stratiform rain encounters that were incorrectly classified as deviations.

In this paper, we present a modified Weather Avoidance Model incorporating predictors based on the standard deviation of VIL and echo tops. The standard deviation measure is used by the Corridor Integrated Weather System (CIWS) to distinguish between regions of vigorous convection and stratiform rain. Operational experience with CIWS and observations of flight traffic suggest that pilots are more willing to fly through stratiform rain and a predictor that reliably discriminates between convective and stratiform weather may have skill as a deviation predictor. CWAM2 used the difference between 90th and 10th percentile values of VIL and echo tops to estimate standard deivation; in this analysis, the standard deviation is calculated directly. The analysis is based on data from approximately 1,900 flight trajectories through three Air Traffic Control en route super-sectors over six days from summer 2006 used in the CWAM2 study.

This work was sponsored by the National Aeronautics and Space Administration (NASA) under Air Force Contract FA8721-05-C-0002. Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the United States Government.