Effects of urban green spaces and related urban morphology parameters on urban sound environment

Margaritis, Efstathios (2017) Effects of urban green spaces and related urban morphology parameters on urban sound environment. The University of Sheffield, Doctoral Thesis.

Record type: Thesis (Doctoral)

Abstract

Urban morphology in combination with soundscape planning and design are important parameters towards the development of sustainable cities. Towards this direction this study primarily investigates the effect of urban morphology and green-space related parameters on traffic noise in different analysis levels. Secondly, it complements this first objective approach with a subjective one, investigating peoples’ perceptual attributes using auditory and visual stimuli. Both approaches aim at merging the gap between acoustics and planning on the grounds of the new holistic approach of urban sound planning. At first, a triple level analysis was conducted including case study cities across Europe with a view to understand to what extent greener cities can also be quieter. The analysis was conducted using GIS tools and noise data from European databases combined with land cover parameters. Results were scale-dependent with lower noise levels to be achieved in cities with a higher extent of porosity and green space coverage. A further cluster analysis combined with land cover data revealed that lower noise levels were detected in the cluster with the highest green space coverage. At last, a new index of ranking cities from the noisiest to the quietest was proposed. Using the findings concerning green spaces and traffic noise from the previous study, a second analysis was conducted focused on eight UK cities. The green space variables were adjusted to incorporate also parameters related to spatial pattern and smaller ontologies, such as vegetated backyards or front yards. Parameters related to urban morphology, such as buildings and roads were also investigated. The analysis was conducted in a macro, meso and micro scale using regression models and GIS tools. Cities were divided in two types of settlement forms (linear, radial) and results showed that the latter were associated with a higher green space ratio. Green space and morphological parameters managed to predict the Lden levels in two cities with an explained variance up to 85%. Results suggested that urban green space variables combined with other features of urban morphology conduct a significant role in traffic noise mitigation and can be used as a priori tool in urban sound planning. The third part of the study focused particularly on the effects of vegetation and traffic-related parameters on the sound environment of urban parks. The sound environment was evaluated using both simulated traffic data and in situ measurements from mobile devices inside the parks. Results showed that simulated noise distribution in the park scale varied between 43 and 78 dB(A) with a maximum range of 9 dB(A) per park and higher noise variability for LA10. Two groups of parks were identified according to the distance from the international ring road. For measurement data, LA90 and LA10 were higher outside the parks with differences up to 6 dB(A) for LA90 and up to 14.3 dB(A) for LA10. Additional correlations were also detected between noise levels and morphological attributes, while slightly higher noise levels were detected in areas covered with grass compared with tree areas. The previous objective findings were combined with a perceptual study on the transition from prediction to soundscape and design implementation. In this study the relationship between land use and sound sources was explored. The stimulus material was based on binaural recordings and 360°-videos. Participants were required to assess the dominance of sound sources and the appropriateness of land use and socio-recreational activities. Results showed that the activity-based environment can be explained by two main Components. The green space coverage and the proximity to roads were the most significant parameters in the prediction of these two components. In the final stage, a multivariate analysis (MANOVA) was used in order to identify significant variations for the land use activity variables in the three urban activity profiles. The whole process emphasized on the importance of linking urban planning and design with soundscape from the land use activity viewpoint. In the final stage, two of the previous UK case study cities were selected in order to develop a mapping model to aid soundscape planning with parallel implementation and assessment of its effectiveness. Ordinary Kriging interpolation was used in both cases to simulate the predictive values in unknown locations. In Sheffield, the soundscape model was based on the prediction and profiling of sound sources, while in Brighton in the prediction and profiling of perceptual attributes. The cross-validation process in both cases presented small errors with slightly underestimated prediction values. The outcomes from both case studies can be applied in environmental noise management and soundscape planning in different urban scales.

This record has no associated files available for download.