|Abstract:||Recently, cities across the world implemented Green Stormwater Infrastructure (GSI), a strategy that uses vegetation to manage stormwater. While evidence suggests that GSI provides ecological benefits to the urban environment, we know little about how GSI impacts human health and well-being. We know that urban nature, such as trees and parks, provides many human health benefits and is highly preferred. GSI, however, varies in shapes, sizes, and designs to support its unique stormwater functions. Its appearances greatly differ from the conventional urban landscape with mowed lawns and mature trees. Would these different forms of urban nature provides the same health benefits with trees alone? That is, to what extent do people prefer GSI as a part of urban landscape—and in extension, gain health benefits when they interact with GSI? This lack of knowledge prevents designers and planners from designing GSI that people prefer, which might reduce the effectiveness of urban nature. Furthermore, people are less likely to accept and contribute for the landscapes they do not prefer, which might affect the performance of GSI itself.
To address this research gap, I conducted four empirical studies linking GSI and human preference and well-being. First, I examined how people perceived and preferred different type of GSI using photo-questionnaire and factor analysis. Then, I examined how people preferred GSI landscapes with different vegetation density levels by analyzing three photo-questionnaire and Browndog’s Green Index, a recently developed tool that identifies vegetation density via machine learning and image processing. Third, I investigated the relationships between GSI density and two psycho-physiological measures: stress and attention. Finally, I examined how other urban contexts, such as perceived messiness, perceived levels of urban developments, and perceived safety influence GSI preference, and whether these contexts mediate the relationship between vegetation density and preference.
I found that different types of GSI are preferred differently, and messiness played a role in bio-retention preference. More vegetation density predicted higher preference, but an increase in vegetation density was associated with more dramatic changes in preference when the vegetation density level was low. I did not find a significant relationship between GSI, stress, and attention capacity. Finally, I found that perceived levels of urban developments and perceived safety predicted GSI preference, but did not mediate the relationship between GSI vegetation density and preference.
From our results, I suggested that designers should not be discouraged from applying GSI in the urban environment, especially around the areas with low vegetation density. Designers and planners should emphasize neatness and cues of care in GSI designs. They should find a way to minimize perceived urban landscape, that is: make the landscape appear more natural, and increase perceived safety in GSI implementation, among other landscapes. This research is important because it is one of the first studies that objectively examined how people prefer some types of GSI, such as bio-retention. It also is the first few to use a recent research technology, Browndog’s Green Index, to further built-environment research. I proposed future studies to further examine the relationship between GSI density and stress and attention capacity, while considering the urban contexts and stormwater capacity and management of the GSI. This dissertation and my future studies will help contribute to make cities across the world healthier for humans and the surrounding ecosystem, and to bring nature to every doorstep.