Issue 044 - Aviation - Transportation throughput
What does a 31% drop in traffic at the world's busiest international airport actually represent?
Reuters reported that Dubai International Airport handled 31.5 million passengers in the first half of 2026, down from about 46 million during the same period in 2025, as the Iran war disrupted aviation throughout the Gulf.
The problem
Estimate how many passenger flights would have to disappear from Dubai's schedule to account for roughly 14.5 million fewer passenger journeys in six months.
Then put the result in context: how many days or weeks of Dubai International Airport's normal flight activity does that represent?
Optional extension: estimate the total aircraft flight-hours or passenger-hours of travel represented by the missing traffic.
Because Fermi problems target an order of magnitude, I normally use no more than two significant digits and write most calculations in scientific notation; the Fermi reference explains both conventions.
Before checking sources
Matt's first pass
I assumed there are about 150 passengers on the average flight into or out of Dubai.
missing passengers ~= 1.5 x 10^7 passengers
passengers per flight ~= 1.5 x 10^2
missing flights ~= 1.5 x 10^7 / 1.5 x 10^2
~= 1.0 x 10^5 flightsTo figure out how many days of regular air traffic that represents, I looked at the number of passengers handled in 2025, about 46 million, and divided by 150 passengers per flight:
normal first-half flights
~= 4.6 x 10^7 passengers / 1.5 x 10^2 passengers/flight
~= 3.0 x 10^5 flightsThen I figured out how many days had passed so far in 2026, rounded to about 240, and divided that into 300,000 flights:
daily flights ~= 3.0 x 10^5 / 240 days
~= 1.3 x 10^3 flights/dayTo make up for 100,000 dropped flights, at 1,300 flights handled daily, it would take:
normal-traffic equivalent
~= 1.0 x 10^5 flights / 1.3 x 10^3 flights/day
~= 77 days
~= 2.5 monthsCalibration Score
Matt's Calibration Score: 75 / 100
Higher is better: earn points for accurate pegs, sound models, correct math, and a result close to the sourced answer. The image shows percent full of it: 100 minus the Calibration Score.
Pegs: 20/30. The 150-passenger estimate was low compared with DXB's reported 214 passengers per movement, but it was within 50%.
Model: 30/30. Passenger loss divided by passengers per aircraft movement is the right core model.
Math: 5/10. The arithmetic mostly held, but using year-to-date days instead of first-half days distorted the time comparison.
Result: 20/30. The 100,000-flight estimate was high, but still comfortably within an order of magnitude of the sourced answer.
Grounding facts
DXB is unusually large by international-airport standards. Dubai Airports reported 95.2 million passengers and 454,800 flight movements in 2025, or about 1,250 movements per day.
The missing 70,000 movements are not simply 70,000 individual airplane trips erased from the planet. Some passengers rerouted, some flights may have operated emptier, some airlines suspended routes, and transfer traffic can change in complicated ways. But the scale is still enormous: about two months of normal DXB takeoffs and landings.
After checking sources
Check and recalibrate
Dubai Airports reported that 2025 average passengers per aircraft movement remained strong at 214. That is the cleanest peg for this problem because it already reflects DXB's mix of large long-haul aircraft, regional flights, and load factors.
passenger decline
~= 4.6 x 10^7 - 3.15 x 10^7
~= 1.45 x 10^7 passengers
missing aircraft movements
~= 1.45 x 10^7 passengers / 2.14 x 10^2 passengers/movement
~= 6.8 x 10^4 movementsSo the missing traffic is on the order of 70,000 aircraft movements. AP's reported aircraft-movement figure gives a nice independent check: 150,600 movements in the first half of 2026 was down 32.1% from the prior year.
estimated 2025 first-half movements
~= 1.506 x 10^5 / (1 - 0.321)
~= 2.2 x 10^5 movements
movement decline
~= 2.2 x 10^5 - 1.5 x 10^5
~= 7.0 x 10^4 movementsNow compare that with normal flight activity. Dubai Airports reported 454,800 total flight movements in 2025:
normal daily movements
~= 4.55 x 10^5 movements/year / 365 days/year
~= 1.2 x 10^3 movements/day
normal-traffic equivalent
~= 7.0 x 10^4 missing movements / 1.2 x 10^3 movements/day
~= 6 x 10^1 days
~= 8 weeksA good calibrated answer is therefore about 70,000 missing aircraft movements, equivalent to roughly 55 to 60 days of DXB's normal flight activity, or about eight weeks.
For a passenger-hours extension, if the average disrupted itinerary represented 4 to 5 hours in the air:
missing passenger-hours
~= 1.45 x 10^7 passengers x 4 to 5 hours
~= 6 x 10^7 to 7 x 10^7 passenger-hoursPost-check reflection
Matt's reflection
Looks like I misread the prompt and assumed we were looking at stats to date rather than just the first half of the year, so that is one issue. I also estimated a per-flight passenger count that was lower than the realistic estimate, 150 versus about 210.
Otherwise, my math and results were not terrible. That is a pretty considerable disruption in traffic. I also was not familiar with the terminology "aircraft movements"; if that is more common for this type of problem, I should probably keep track of it.
Recommended memory peg
For airport problems, remember aircraft movement = one takeoff or one landing. For a major widebody-heavy international hub, ~200 passengers per movement is a useful starting peg.
Reader results
Bars show how submitted estimates sort into the answer choices from the gut-check prompt.