Two recent studies [1, 2] present models that estimate the probability of pilot deviation around convective weather in en route airspace as a function of location and flight altitude. Weather Avoidance Fields (WAFs) calculated using these models may be used to determine which regions of airspace are likely to be passable. However, models to predict pilot behavior are very difficult to validate, since pilot intent must be inferred using only flight plans and actual trajectories. A recent study  illustrates the difficulties and limitations of attempting to infer pilot intent. Furthermore, observed flight trajectories may not correctly represent pilot preference. In some instances, pilots may have penetrated airspace that they would rather have avoided or they may have avoided airspace that was easily passable.
It is desirable to replace the inference of pilot intent with an objective measure (e.g., turbulence, lightning, hail, etc.) of airspace penetrability in the validation of WAFs. In this study, WAFs are validated by comparing the deviation probabilities encountered along flight trajectories to in situ turbulence measurements. Turbulence measurements were obtained from two sources: two flight missions flown by instrumented aircraft in and around thunderstorms in the upper Midwest, and in situ turbulence measurements from instrumented commercial aircraft . On-board weather radar images, video, photographs and pilot narrative from the cockpit were also obtained from the flight missions. These data are also compared to the WAF probabilities and turbulence encountered along the flight trajectory.