From Grass Fields to Parks: Identifying and Classifying Urban Green Space for Sustainable City Planning in Utrecht

Abstract

This research analyses the location and quality of Urban Green Space (UGS) in Utrecht, aiming to categorize them based on quality factors. A combination of satellite image classification and qualitative assessments has been used. Different image classification methods were evaluated to determine the most suitable approach. The study further assesses the quality of green spaces by considering both objective (data-driven) and subjective (community-based) factors. The objective approach uses a multi-criteria analysis with factors determining green space quality. The subjective approach analyses surveys filled out by citizens of Utrecht. The final step involves creating a typology of UGS in Utrecht based on these quality assessments. The study found that classifying an RGB aerial image with an object-based model was the most appropriate model with an accuracy of 90%. The green space data was compared with various ambitions and it showed that Utrecht has enough green space per household throughout the city to meet their target, but the green area is unevenly distributed. In 43% of neighbourhoods, there is enough green space compared to the ambition. In 71% of neighbourhoods, the proportion of public green space within total green space is large enough. Finally, it was found that no correlation exists between objective and subjective green space quality. In some parts of the city, the proportion of green space in the area is very high in number and quality. Other parts are grey and the green space that is there in these parts is usually of lower quality, which means that more and higher quality green space is needed in those places. These findings provide insights into the distribution and characteristics of urban green areas, offering valuable recommendations for sustainable city planning.