Data

Whenever an infectious disease poses the risk of spreading into a global pandemic, air transportation plays a key role. Modern air traffic connects every corner of our globe. In doing so, it also enables regionally confined outbreaks of infectious diseases to “jump” across country borders at an unprecedented rate.

Therefore, the core of the computational model is the worldwide air transporation network (WAN). This network has 3893 nodes (airports) that are connected by 51476 directed links (flight routes). Each link is weighted by the traffic flux between nodes, i.e. the average number of passengers that travel each route per day.

The WAN has been analyzed in many publications and used in a variety of computational models for the spread of infectious diseases.

**Worldwide air transportation network.** This network of global mobility is the foundation of the computational model that predicts import risks, most probable spreading routes, and the importance of specific airports in the global dissemination of COVID-19.

Worldwide air transportation network. This network of global mobility is the foundation of the computational model that predicts import risks, most probable spreading routes, and the importance of specific airports in the global dissemination of COVID-19.

Properties of the worldwide air transportation network

The WAN exhibits strong intrinsic structural heterogeneity – some airports are tiny, while others are huge international hubs. These distinctions can be observed by examining the wide-range of magnitudes of various properties of the WAN – a clear sign of network heterogeneity, summarized below.

**Statistics of the worldwide air transportation network.** In the first panel of the figure above, the traffic flux betwwen nodes or average daily number of passengers traveling on flight routes between airports (**link weight**) can vary from less than 1 to over 1000. The second figure shows the number of passengers that traverse an airport on a daily basis (**node flux**), which does not exceed 100 in some airports, while others accomodate over 10000 people each day. Similarly, in the last panel, the number of flight connections emanating from an airport (**out degree**) is approx. 13 on average, though some airports boast over 100 direct connections.

Statistics of the worldwide air transportation network. In the first panel of the figure above, the traffic flux betwwen nodes or average daily number of passengers traveling on flight routes between airports (link weight) can vary from less than 1 to over 1000. The second figure shows the number of passengers that traverse an airport on a daily basis (node flux), which does not exceed 100 in some airports, while others accomodate over 10000 people each day. Similarly, in the last panel, the number of flight connections emanating from an airport (out degree) is approx. 13 on average, though some airports boast over 100 direct connections.

**The WAN is dominated by big airports.** As a result of the network heterogeneity of the WAN, the model is predominantly driven by only a small number of nodes, i.e. the big regional and international hub airports, while the majority of the nodes are structurally trivial. For example, if the smallest half of airports the network (approx. 2000) were removed, only about 7% of the total traffic would be lost.

The WAN is dominated by big airports. As a result of the network heterogeneity of the WAN, the model is predominantly driven by only a small number of nodes, i.e. the big regional and international hub airports, while the majority of the nodes are structurally trivial. For example, if the smallest half of airports the network (approx. 2000) were removed, only about 7% of the total traffic would be lost.

Top 20 airports by traffic (excluding Mainland China)

Airport NameCodeCountryContinentLink Weight
FrankfurtFRAGermanyEurope579
AmsterdamAMSNetherlandsEurope551
Paris Charles de Gaulle AptCDGFranceEurope546
Moscow Domodedovo AptDMERussian FederationEurope502
IstanbulISTTurkeyAsia485
AtlantaATLUSAAmericas480
Chicago O’Hare International AptORDUSAAmericas474
MunichMUCGermanyEurope455
London Gatwick AptLGWUnited KingdomEurope451
DubaiDXBUnited Arab EmiratesAsia450
Beijing Capital Intl AptPEKChinaAsia445
Houston George Bush Intercontinental ApIAHUSAAmericas427
Rome Fiumicino AptFCOItalyEurope419
Dallas/Fort WorthDFWUSAAmericas415
BarcelonaBCNSpainEurope402
Newark Liberty International AptEWRUSAAmericas391
DenverDENUSAAmericas388
ManchesterMANUnited KingdomEurope380
Lester B Pearson IntlYYZCanadaAmericas373
BrusselsBRUBelgiumEurope367

Top 20 airports by degree (excluding Mainland China)

Airport NameCodeCountryContinentOut degree
FrankfurtFRAGermanyEurope293
AmsterdamAMSNetherlandsEurope278
Paris Charles de Gaulle AptCDGFranceEurope275
Moscow Domodedovo AptDMERussian FederationEurope251
IstanbulISTTurkeyAsia242
AtlantaATLUSAAmericas240
Chicago O’Hare International AptORDUSAAmericas238
MunichMUCGermanyEurope229
DubaiDXBUnited Arab EmiratesAsia228
London Gatwick AptLGWUnited KingdomEurope226
Beijing Capital Intl AptPEKChinaAsia222
Houston George Bush Intercontinental ApIAHUSAAmericas214
Rome Fiumicino AptFCOItalyEurope212
Dallas/Fort WorthDFWUSAAmericas206
BarcelonaBCNSpainEurope201
Newark Liberty International AptEWRUSAAmericas195
DenverDENUSAAmericas194
ManchesterMANUnited KingdomEurope190
Lester B Pearson IntlYYZCanadaAmericas187
BrusselsBRUBelgiumEurope187