Flights per dayplotphoto
The flight schedule data in the last 91 days for Hong Kong International Airport (IATA code HKG) is scraped from open data API on its official website. Plots are generated for the daily average frequency of destinations/origins, of airlines, of hour of day in schedule, and of status of flight.
Note that over 92% of all passenger flights have only a single destination/origin, and over 57% of all cargo flights have only a single destination/origin.
For a departure flight with multiple destinations, the last is taken. For an arrival flight with multiple origins, the first is taken. (Another approach that is not taken would be treating these "direct" flights composed of multiple segments as separate flights by taking the first in destinations and the last in origins. Note that stopping in the middle for refueling is now less common for passenger flights.)
For a flight with multiple airlines in flight code, the first is taken. (It is assumed that the first in the list of airlines for each flight parsed is the operating carrier flying the code share flight.)
Passenger flights #
In the last 91 days, there were on average 100 passenger flights per day scheduled for departure from HKG.
- The most frequent destination was by far Taipei, followed by Manila, Singapore and Incheon, all of which were East or South East Asian cities. The most frequent non-Asian destinations were London Heathrow, Los Angeles and Sydney.
- Nearly half of all flights were operated by home carriers Cathay Pacific and Hong Kong Airlines.
- There were more passenger flights during the day starting from 9am than in midnight hours.
- There were on average 93 flights per day departed (DA) and 7 flights per day canceled (CX).
There were on average 94 passenger flights per day scheduled for arrival at HKG. The pie plots show very similar distributions.
Cargo flights #
There were on average 109 cargo flights per day scheduled for departure from HKG. This number is notably higher than that for passenger flights.
The distribution of destinations was significantly less top heavy than that for passenger flights. The more frequent destinations can be categorized into three regions:
i. Asia (including Taipei, Tokyo-Narita, Shanghai Pudong, Incheon and Singapore);
ii. United States (including Cincinnati-Kentucky and Los Angeles); and
iii. Middle East (including Dubai and Doha)
The distribution of airlines was also less top heavy than that for passenger flights. The most frequent airlines were still home carriers, including Cathay Pacific, Air Hong Kong and Hong Kong Air Cargo.
The scheduled departure times were almost evenly distributed throughout a day on average. There was an exceptionally high number of flights departing at 3 am, supposedly to fit cargo delivery operation schedules.
There were on average 102 flights per day departed (DA) and 7 flights per day canceled (CX). The cancelation rate was therefore similar to that for passenger flights.
In the same period, there were on average 106 cargo flights per day scheduled for arrival at HKG. The pie plots show similar and yet slightly different distributions.
Note on cargo flights with first destination/origin taken #
The distribution of destinations for cargo departure flights would be significantly different if the first destination is taken for flights with multiple destinations. The most frequent destination is by far Anchorage, which is a major cargo hub in Alaska, United States.
A similar plot can be obtained for distribution of origins for cargo arrival flights if the last origin is taken for flights with multiple origins.
Such types of plots presented above would be a good addition to the many statistical plots in Hong Kong International Airport Master Plan 2030 Technical Report (2011) although other metrics such as passenger volume and cargo volume may be more suitable for strategic analysis of airports rather than flights.
Further work #
It is suggested that other datasets be explored to perform similar data analysis on other airports.
It will be interesting to re-run my scripts for plots later and to observe the change in absolute number of flights, in cancelation rates, and in distribution of destinations/origins if any in the predicted post-Covid recovery in aviation.
It is also suggested that a predictor of flight status (departed or canceled) can be made based on features including day of week, hour of day, origin/destination, airline and flight number.
For some relevant illustration, some of my better photographs of aircrafts and aviation are included as follows.
A twin-engine aircraft passing by in the sky with green plants on the ground in Hong Kong (HKG) (2018) (f/8, 1/99, ISO 100)
An aerial view of Mount Fuji behind clouds through a window of an aircraft that departed from Tokyo (NRT) (2018) (f/6.625, 1/256, ISO 100)
- Feb 2021 Replaced plots with new versions generated with the latest data and re-exported in SVG format.
- Jan 2021 Added original aviation photography.
- Jan 2021 Merged plots of departure and arrival flights to an animated sequence for each category.