Predicting Peak Sector Occupancy with Two-Hour Convective Weather Forecasts

An important function of traffic flow management ensuring the number of aircraft entering a sector does not exceed the amount that can be safely controlled by the sector controller. One factor that makes this task difficult is the uncertainty of the impact of convective weather, as both the weather forecast and the impact given specific weather is uncertain. In this investigation, we study this effect indirectly by exploring the relationship between convective weather forecasts and observed peak sector occupancy. Specifically, we measure how well the peak sector occupacy can be predicted using area-based and directional-based weather translation models. We also present a methodology for comparing weather translation models using a machine learning approach. Our results over the entire range of weather forecasts show that weather translation models produce better predictions than models without translation, but little significant difference is observed among the weather translation models. When restricted to heavy forecasted weather, however, some statistically significant differences were observed among the weather translation models.