It might seem obvious, but at any given moment, an airport can only accommodate so many aircraft.
The accurate estimation of an airport capacity is critical for the prediction and simulation of delays in the airspace network—and therefore, also for creating Lumo’s prediction delay scores.
When demand exceeds the current capacity at an airport, queues develop and persist on the airport surface, often resulting in large departure and arrival delays at the airport in question that, in turn, propagate and magnify throughout the airspace network.
At Lumo, we analyze years of historical airport delay data to determine the capacity of airports under varying weather and operating conditions. In so doing, we identify periods during which airports are expected to be constrained and account for such constraints in our predictions of delay.
What the figure shows: Newark International Airport (EWR) is constrained when there are 26 or more aircraft on the airport surface (this happens historically 10.6% of the time). During highly constrained periods (7.4% of the time) the typical achievable throughput is only 19 operations per 15 minutes.