Risk-Hedged Approach for Re-routing Air Traffic Under Weather Uncertainty

Abstract

This paper explores a new risk-hedged approach for re-routing air traffic around forecast convective weather. In this work, flying through a more likely weather instantiation is considered to pose a higher level of risk. Current operational practice strategically plans re-routes to avoid only the most likely (highest risk) weather instantiation, and then tactically makes any necessary adjustments as the weather evolves. The risk-hedged approach strategically plans re-routes by minimizing the risk-adjusted path length, incorporating multiple possible weather instantiations with associated likelihoods (risks). The resulting model is transparent and is readily analyzed for realism and treated with well-understood shortest-path algorithms. Risk-hedged re-routes are computed for some example weather instantiations. The main result is that in some scenarios, relative to an operational-practice proxy solution, the risk-hedged solution provides the benefits of lower risk as well as shorter path length. In other scenarios, the benefits of the risk-hedged solution are ambiguous, because the solution is characterized by a tradeoff between risk and path length. The risk-hedged solution can be executed in those scenarios where it provides a clear benefit over current operational practice.

Tradeoffs for Routing Flights in View of Multiple Weather Hazards

“…Consideration of weather nowcast/forecast and associated uncertainty can yield valuable insights, as has been demonstrated in several studies on aircraft routing [28,29,46], air traffic flow management [23], and airspace congestion. Future research may also allow for penetration of larger moderate or greater areas of turbulence and icing hazards because severe levels of such hazards are usually intermittent.…”

“…Initially, constraints were treated deterministically (e.g., no-fly zones) (e.g., [21]); but soon, concepts of soft versus hard constraints (accommodating different types of aircraft limitations or airline procedures) [22] and treating constraints probabilistically [23][24][25][26] emerged. The latter opened the door for capturing forecast weather and demand uncertainty [27][28][29]. Also, finding path solutions in a maze of constraints may require minimal-sized bottlenecks to safely pass through [30][31][32], accounting for limitations on how hard an aircraft may be allowed to turn or the number of turns as part of the rerouting [33,34], accommodating airspace structures (e.g., sectors, waypoints, fixes) and considering airspace capacity (i.e., other air traffic) [35][36][37].…”

Optimality and robustness in path-planning under initial uncertainty

“…Given a forecast of probable storm front trajectories, one option is to devise a good flight path and commit to it in advance all the way to the target. This is the key idea of the approach developed in [34], but one can do much better by taking into account the future re-routing once a more precise forecast becomes available. What follows is an illustration of how this can be done in our framework under a simplifying assumption that the true position of the storm front will be revealed at a known time T .…”

Weather Design Considerations for the TASAR Traffic Aware Planner

“…While CWAM is an excellent product with which NASA has experience in the context of several research projects [17][18][19][20], it is not yet widely available to airline users (although it is a part of the FAA's impending rollout of the NextGen Weather Processor [21]). For the operational evaluation with Alaska Airlines, it was important to identify a weather data source that would be compatible with the airline's current operational approvals and practices.…”