Statistical Modeling: Why Do People Move to a City?

A project by TH Lübeck, in collaboration with Wirtschaftsförderung Lübeck (Economic Development Lübeck), uses ten laser counters installed on the Old Town Island by the Wirtschaftsförderung in 2021. They record the flow of people in the Old Town, as well as certain information on the weather, public transportation and parking space utilization.

The data sources used provide heterogeneously structured data of varying frequency, quality, and detail. Professor Thomas Romeyke has therefore developed an IT system for the project that retrieves the regularly generated data from its sources, transforms it in a target-oriented manner and places it in a database for analysis. This project is like many other 'big-data' projects, he says. Real-world data from a variety of settings tend not to match, and a meaningful transformation is almost always necessary before a reliable analysis can be made."

Professor Karen Cabos then uses the generated data to identify the factors that affect the number of people in the downtown area. I want to be able to show which underlying conditions make the inner city attractive to people, Cabos says. In order to do so, she uses a structural regression model to determine the extent to which different variables affect visitor numbers.

Cabos considers factors like parking, public transportation, time of year and time of day, weather or major events and other seasonal influences in her statistical model. Defining the influencing factors is as important as the dataset. Currently, we have reliable data gathered for just over a year, Cabos explains, based on which we can already draw some preliminary results. It is already possible to make a fairly reliable statement that the parking situation in Kanalstrasse influences the public traffic in Hüxstrasse more than in Breite Strasse. Or that months with a high volume of tourists show an above-average demand for parking spaces in the city center. Specific effects, such as the influence of Corona measures or the 9-euro ticket, however, cannot yet be quantified using this dataset. That will only be possible in the coming years. The flow of people and their underlying causes are strongly related to the traffic situation, but they also have a major economic impact. What makes commercial spaces attractive? Which streets and stores are particularly busy during the Christmas season? What do we need to do to attract people to the city center?