Project Proposal

# Project Proposal Group 11

## Research objective

Air passenger traffic can be strongly affected by major external disruptions. However, disruptions from different domains may differ in both the magnitude and duration of the effects. This project compares three disruptions from different domains:
    1. Security/geopolitical: The 9/11 attacks
    2. Economic: Financial crisis of 2008/2009
    3. Pandemic: COVID-19

The objective of this project is to analyse and compare changes in air passenger traffic at Amsterdam Airport Schiphol during and after these three major disruptions.

### Research question
How did air passenger traffic at Amsterdam Airport Schiphol change around three major external disruptions: the 9/11 attacks, the global financial crisis, and the COVID-19 pandemic?

### Variables 
- Monthly total number of passengers
- Percentage change in passenger numbers
- Maximum relative decline
- Recovery time (Recovery time will be defined as the number of months required for passenger traffic to return to the pre-disruption baseline level)

### Temporal scale
- Approximately 24 months before each disruption
- The main disruption period
- Approximately 24 months after each disruption

### Geographical scale
Amsterdam Airport Schiphol, the Netherlands

## Datasets
### Primary dataset
The main dataset used in this project is the montly traffic dataset of Amsterdam Airport Schiphol. 
Source: https://www.schiphol.nl/nl/schiphol-group/verkeer-en-vervoer-cijfers/

The dataset contains monthly passenger data from 1992 onwards. 
For this project, the main variable of interest is the total number of passengers per month. 

### Additional dataset
CBS - Montly numbers of Dutch airports of national importance
Source: https://opendata.cbs.nl/#/CBS/nl/dataset/37478hvv/table

### Intended data analysis pipeline
From primary dataset (Schiphol):
- Load the data
- Clean the data so it's uniformly used across all needed data (2 years before, main disruption period and 2 years after). The file contains a lot of data so this is an important step
- Calculate the pre-disruption baseline for these periods, this is needed to set a standard to compare the 3 situations:
    - Normalize/index passenger traffic relative to each event's pre-disruption baseline, allowing the three disruptions to be compared despite differences in absolute passenger volumes over time
- Index the 3 situations to make the actual comparison
- Find the extremes per shock; what is the actual impact and recovery time
- Interpretation of the collected data; which shock had the most impact on air traffic

From additional dataset(s):
For now there is only one additional dataset since the primary dataset is very large. The CBS file itself is very small and only shows air traffic in and out. The CBS dataset will be used as a secondary source to validate passenger trends observed in the primary Schiphol dataset where comparable variables and time periods are available.

### Hypothesis
We expect COVID-19 to have the longest recovery time, 9/11 to cause a sharp immediate decline in passenger traffic, and the financial crisis to have a smaller but more gradual impact.

### Limitations
- We will look at the periods before, during and after the shock. We will however not consider any other circumstances in these periods, e.g. maximum capacity of Schiphol, volcanic eruption in Iceland (2010), which also could impact the amount of air traffic from Schiphol.
- We focus exclusively on passenger traffic at Amsterdam Airport Schiphol and do not compare Schiphol with other Dutch airports.
- Monthly passenger traffic is subject to seasonal variation. Differences between months may therefore not be caused solely by the disruptions under study.