Impact of Atmosphere Circulation on Time-Optimal Flight Routes Using Two Different Flight Trajectory Modeling Methods

Upper-level winds embedded on large-scale flow play a crucial role in the efficiency and safety of commercial flights, especially on long-haul journeys. This importance has been demonstrated in the previous studies through analyzing historical flight records. Although various factors can influence on the actual flight planning and operation, to isolate the atmospheric effects many studies have used a concept of Wind-Optimal Route (WOR), which is a flight trajectory that minimizes flight time by utilizing wind variation along the route. The adoption of the WOR offers operational benefits by reducing flight time, consequently leading to a reduction in carbon emissions in atmosphere. For WOR adoption, significant research has been conducted using the shooting method based on the Pontryagin's maximum principle to compute optimal routes. However, this method requires an additional necessary condition of initial heading angle that leads to a genuine optimal solution, which sometimes results in the failure of finding the WOR in several cases. Another alternative approach involves the constructing of node networks and utilizing pathfinding algorithms (here, A* algorithm). This method guarantees optimal solutions and offers greater flexibility compared to the shooting method approach. However, fundamental differences between these methodologies have yet to be extensively explored. In this study, both of the two methods were employed to calculate WORs between the same city pairs, and differences in the simulated flight times and spatial distributions were examined. When computed from the daily averaged wind fields from reanalysis data, the two methods exhibited slightly different simulated flight times with a strong correlation (above 0.99). In general, flight times from the shooting method slightly faster than those from the A* algorithm, although there are some minor discrepancies in accuracy and computing times depending on the number of regridding and neighbouring schemes used. There are no significant discrepancies in the spatial distributions between two WORs, which confirms that the A* algorithm with no failure of WOR solution in all cases is more useful for calculating the impact of large-scale atmospheric circulation on long-term WORs and subsequent emissions. Using this method, we additionally examine the potential encounters of clear-air turbulence (CAT) along the A*-based daily WORs for a climatological time scale (40-yrs). And, we also calculate the flight paths avoiding potential areas of CAT encounters exceeding certain intensities, estimating the costs incurred by detours that are expected to be a positive long-term trend in accordance with many previous studies investigating the changes in CAT at flight routes around the world due to climate change. The detailed results will be shown in the presentation in the upcoming conference.

Key Words: Optimization, Flight planning, A* algorithm, Turbulence, Climatology

Acknowledgment: This work was funded by the Korea Meteorological Administration Research and Development Program under Grant KMI2022-00310.