US Domestic Flight Cancellation, Delay and Diversion Analysis (2015)

(Tableau Project)

Introduction

This is a Tableau Project published on my Tableau Public (Click here to view).
The goal of this project is to analyse the reasons of US flight cancellation, diversion and delay in 2015.

Data

This data comes from a Kaggle dataset, it tracks the on-time performance of US domestic flights operated by large air carriers in 2015.

• flights.csv: 274,964 flight records (rows) * 33 columns
• airports.csv: airports’ information
• cancel_reason.csv: full-name of cancel reasons
• airlines.csv: full-name of airlines

Data Analysis

Cancellation

Cancellation Reasons

There are three reasons for flight cancellations:

1. Weather (54.07%)
2. Airline/carrier (28.42%)
3. National Air System (17.51%)
   Although there is another reason, security, mentioned by the Kaggle website, it does not appear in the row data.

The top 3 airlines that have the most cancellations are:

1. Southwest Airlines Co.
2. Atlantic Southeast Airlines
3. American Eagle Airlines Inc.

Cancellation Rates

However, the top 3 airlines with the highest cancellation rate are a bit different:

1. American Eagle Airlines Inc. (5.11%)
2. Atlantic Southeast Airlines (2.94%)
3. US Airways Inc. (1.95%)

Bad Weather Cancellations

Bad Weather has led to most cancellations, but the row data does not show whether they were affected by the weather of the departure or arrival airport.

This map only shows the departure airports of those cancelled flights. We can find that the flights departure from

1. Dallas/Fort Worth International airport and
2. Chicago O’Hare International airport
   has been cancelled most, which were 231 and 217 in the whole year respectively.

Diversion

Airlines of Diverted Flights

Southwest Airlines Co. had the most diverted flight (181), which is almost two times than the second most Atlantic Southeast Airlines (99).

However, the diverted rates of most airlines are similar.

Destinations of Diverted Flights

The flights towards

1. Hartsfield-Jackson Atlanta International Airport and
2. Dallas/Fort Worth International Airport (mentioned above as having the most cancellations due to bad weather)
   were diverted most, 57 and 56 respectively.
   
   

Delay

Total Delayed Time

1. Southwest Airlines Co. had the longest delay time, which was 289,992 minutes, followed by
2. Atlantic Southeast Airlines (187,908mins) and
3. Skywest Airlines Inc.(166,148mins)
   Delayed for Weather Reasons
   The top 3 airports that had the longest delay time due to weather reasons were:
4. Chicago O’Hare International Airport (24,381mins)
5. Dallas/Fort Worth International Airport (16,276mins)
6. Hartsfield-Jackson Atlanta International Airport (15,987mins)
   
   

Delayed for Airline Reasons

1. Southwest Airlines Co. had the longest delay time due to airline reasons (182,670mins), followed by:
2. American Airlines Inc. (133,674mins) and
3. Delta Air Lines Inc. (125,178mins)
   
   

Delayed for Weather Reasons

1. Chicago O’Hare International Airport
2. Dallas / Fort Worth International Airport
3. Hartsfield-Jackson Atlanta International Airport
   
   

Delayed for Security Reasons

The top 3 airport that had the longest delay time due to security reasons were:

1. Dallas/Fort Worth International Airport (374mins)
2. Phoenix Sky Harbour International Airport (362mins)
3. Los Angeles International Airport (348mins)
   
   

Great Months for Flying

The trends were similar for both of them. There were the least cancellations, diversions delayed time in September 2015, followed by April. February and June were bad travel time in 2015.

Great Days of Week for Flying

Saturday would be great for travel, while those who flighted on Monday had a larger chance to have bad experience in 2015.

Design

In terms of maps, I found it might not be easy for readers to quickly get information regarding the numbers, but only using bar chart would lose important geographic information (for example, the map of cancellations for weather is similar to the one of destination of diverted flights). Thus, I put the bar chart at the bottom of each map for users’ convenience. The bar chart and map share a filter.

Thought bar chart is more accurate, I found it would be boring to always use it. Thus, I used bubble chart to show less important information (e.g. diverted rates).

5 reasons of delay are shown in row data. While weather and security delays might tell us which airport we could use less, the airline reasons would help us choose flight when planning a trip.

Some fly-in-and-fly-out persons fly on Monday and come back on Friday. Thus, I decided to explore whether there was a rule of the delayed flights during a week. Line Chart could provide a clear view of this rule.