An approach is presented for using the probabilistic forecast of stratus clearing time at San Francisco (SFO) to achieve more efficient Ground Delay Programs (GDPs) by better determining GDP end time and scope. Given a probabilistic forecast, we use a Monte-Carlo simulation approach to generate many stratus clearing times for each discrete GDP end time and scope under consideration. Various key measures are evaluated such as unnecessary ground delay if the GDP ends later than stratus clearing and the risk of airborne holding at the end of the GDP if the GDP ends earlier than stratus clearing. An objective function that includes each of the key metrics captures the cost of each scenario under consideration, and the optimal GDP parameters can then be selected. Results show reductions of 29% for unnecessary issued ground delay and reductions of 39% for unnecessarily delayed flights over the SFO GDPs during the severe weather seasons in 2006 and 2007.
Keywords: Ground Delay Program, Monte-Carlo Simulation, Probabilistic Forecast, SFO Stratus Forecast System
/ Other authors: