Thesis subject

MSc thesis topic: Public Space in urban configuration | does a citypark count

The Greening of Urban area is a serious need. Many studies present a green index or rank of neighborhoods, quarters and cities. However what is taken into account to calculate he greening factor, and how does they count.

In different studies we notice that the spatial unit of a neighbourhood (wijk) is rather flexible and the variation of area size is highly variable. According to random selection of 123 out of 4486 Dutch neighbourhood areas range from 57ha to 5930ha. This variation is not strange at all because municipalities are free to delineate neighbourhood boundaries. It does seriously impact the greening rank of a neighbourhood due to the size and inclusion of specific facilities like outdoors sport facilities, embankments, utility gardens, wasteland, parks and recreation grounds.

To deliver a tool to make the green ranking of neighbourhoods a more fair practice a more precise definition of urban green is needed. In this study we like to figure out what exactly is meant by a (city) park, because it is not really clear what a park is by the geometrical delineation and their specific attributes, like the ration between paved, water, green (grass, herbs, shrubs and trees) and constructions as well the impact of these on the green ranking.

In this study we like to find the park definition as used by Dutch municipalities and based on that the extraction of park data to study the impact on city and neighbourhood green indexes and rankings.

Objectives

  • Review studies on urban park designs
  • Interview municipality staff on the definition of an urban park
  • Classify urban parks regarding the definitions - calculate impact of park characteristics on greening ranks/index of nieghbourhoods/cities
  • Assess impact on current green indices and rankings

Data

  • RGB data, PDOK data, Data of Municipalities

Methods

  • Park classification by Random Forest as well as deterministic approaches

The thesis will be supervised by GRS (primary contact Ron van Lammeren). There will be options to contact involved partners like WENR, COBRA and Stadswerk.

Requirements

  • Interest in urban configuration and especially city parks
  • Affinity with Random Forest approaches and traditional map algebra or overlay
    boolean

Theme(s): Modeling & visualization; Empowering & engaging communities