United State flight delays and performance data set is from RITA. The clean data set for 2008 is download from here. Flight cancellation and delay information in 2008 are explored in this report.
The flight cancellation rate (number of cancelled flights over number of total flights) for each month in 2008 is shown as stacking bar in Figure 1. We can see that January to March and December have high cancellation rate comparing to other months. In addition, we also can see the flight cancellation rate by extreme weather are also high in these months. It makes sense that the high cancellation rate in winter season could be caused by the snow.
However, the cancellation caused by weather in September is also high. In order to find the reason of high cancellation in September, flight cancellation heatmap for each day in 2008 is shown in Figure 2. We can see that Sep. 12 to Sep. 14 have over 1000 flights cancellation each day and it is hurricane Ike after I check the weather history.
How are cancellation categories defined
The average flight delay time is calculated over the hour of day as in Figure 3. The flights scheduled at 8 pm to 9 pm have the highest average delay time 17 minutes. The average delay time increases from 5 am to 8 pm. Then it decreases after 8 pm. At 3 am, the average delay time is 11 minutes which is significantly higher than other time in the early morning.
Figure 4 shows the average delay time over hours and days by week. Monday morning at 3 am have long delay time as well as Firday night 6 pm to 8 pm.