Research Paper Sample: Urban Green Space and Perceived Air Quality

Model Answer

Abstract

This study investigates the relationship between urban green space provision and residents’ perceived air quality across three neighbourhoods in Sheffield, and compares those perceptions with objectively measured nitrogen dioxide (NO&sub2;) concentrations. A cross-sectional survey of 150 residents, evenly distributed across a high-provision, medium-provision, and low-provision neighbourhood, recorded self-reported air quality perception (0–10 scale) alongside distance to the nearest publicly accessible green space. NO&sub2; concentrations were independently measured using passive diffusion tubes at five fixed sites within each neighbourhood over a four-week period. Perceived air quality was significantly higher in the high-provision neighbourhood (M = 7.4, SD = 1.3) than the low-provision neighbourhood (M = 4.8, SD = 1.6), one-way ANOVA F(2, 147) = 28.94, p < .001, a pattern that closely tracked measured NO&sub2; concentrations (high-provision M = 18.2 µg/m³; low-provision M = 31.5 µg/m³), F(2, 147) = 42.71, p < .001. Distance to the nearest green space correlated negatively with perceived air quality, r(148) = -.34, p < .001. These findings suggest urban green space is associated with both real and perceived air quality benefits, with implications for planning policy and environmental justice in UK cities.

Introduction

Air pollution is estimated to contribute to around 36,000 excess deaths annually in the UK, with nitrogen dioxide and particulate matter from road traffic identified as the principal urban pollutants of concern (Public Health England, 2019). Exposure is not evenly distributed: residents of denser, lower-income urban neighbourhoods typically experience higher ambient pollution alongside lower provision of green infrastructure, a pattern that has drawn growing attention within environmental justice research (Mitchell and Dorling, 2003). Urban green space, parks, street trees, and other vegetated land, has been proposed as a partial mitigant, both through direct pollutant deposition on leaf surfaces and through its role in shaping traffic patterns and land use around residential areas (Nowak, Hirabayashi, Bodine and Greenfield, 2014).

Alongside these physical effects, green space provision may also shape how residents perceive their local environment, including their subjective sense of air quality, independent of objectively measured pollutant levels. Perceived environmental quality matters in its own right: it is associated with wellbeing, outdoor activity levels, and residents’ broader satisfaction with their neighbourhood (White, Alcock, Wheeler and Depledge, 2013), and perception gaps between residents and objective monitoring data can complicate public health communication and local policy responses (Brody, Peck and Highfield, 2004).

This study examines both dimensions in a single UK city, asking whether neighbourhoods with greater green space provision differ in objectively measured NO&sub2; concentrations, whether residents’ perceived air quality mirrors this objective pattern, and whether individual-level proximity to green space predicts perceived air quality after accounting for neighbourhood. It was hypothesised that both measured NO&sub2; and perceived air quality would differ significantly across the three neighbourhood types in the expected direction, and that residents living closer to green space would report better perceived air quality than those living further away.

Literature Review

The capacity of urban vegetation to physically improve air quality is well established, if modest in magnitude at the neighbourhood scale. Nowak et al. (2014), modelling pollutant removal by urban trees across major US cities, estimated that tree cover reduced NO&sub2; concentrations by up to 2–3% locally, an effect small in isolation but potentially meaningful when combined with green infrastructure’s indirect influence on traffic calming, building density, and the buffering of major roads. UK-focused work has reached broadly similar conclusions: Tallis, Taylor, Sinnett and Freer-Smith (2011) modelled pollutant deposition across Greater London and found that strategically located urban trees could achieve locally meaningful reductions in particulate matter, though effects varied considerably with species, planting density, and street geometry, cautioning against treating green infrastructure as a substitute for reducing emissions at source.

Separately, a body of environmental psychology research has examined how green space shapes perceived, rather than measured, environmental quality. White et al. (2013), using large-scale panel survey data, found that residents reported significantly higher life satisfaction in the years they lived in greener areas, with perceived environmental quality identified as a partial mediating pathway. More directly relevant to the present study, Bell, Phoenix, Lovell and Wheeler (2018), working with UK survey data, found that self-reported neighbourhood environmental quality, including air quality, was significantly associated with both objective green space provision and self-reported frequency of green space use, suggesting perception is shaped by both exposure and direct engagement with local nature.

A smaller but important literature has examined the alignment, or misalignment, between residents’ perceptions and objectively monitored pollution levels. Brody, Peck and Highfield (2004) found only a weak correlation between US residents’ perceived and measured air quality, attributing the gap partly to the invisibility of key pollutants such as NO&sub2; and partly to residents relying on visible proxies, such as traffic volume or industrial activity, rather than direct sensory information. This raises an open empirical question for the present, UK-based study: whether green space provision, as a highly visible neighbourhood characteristic, produces perceived air quality patterns that track objective monitoring more closely than has been found in contexts where green space itself was not the variable of interest. Addressing this question in a single UK city, using matched objective and perceptual measures across neighbourhoods differing systematically in green space provision, is the specific contribution of this study.

Methodology

Design and sample. A cross-sectional survey design was used, combining resident self-report data with independent environmental monitoring. Three neighbourhoods within Sheffield were purposively selected to represent high, medium, and low green space provision, based on Ordnance Survey greenspace layer data on the proportion of land within 400m of each neighbourhood’s centroid classified as public green space (high: 38%; medium: 19%; low: 6%), while remaining broadly comparable on housing density and dominant housing type to limit confounding. Fifty residents were recruited from each neighbourhood (N = 150) via a combination of door-to-door leafleting and on-site intercept surveying at accessible public locations, with a target of achieving a broadly even spread of ages and household types within each site.

Measures. Perceived air quality was measured with a single item, “How would you rate the air quality in your local area?”, scored 0 (very poor) to 10 (excellent), alongside a short set of items capturing perceived greenness, general neighbourhood satisfaction, and demographic characteristics. Distance from each respondent’s home postcode to the nearest publicly accessible green space entrance was calculated in metres using GIS network analysis. Objective air quality was measured using NO&sub2; diffusion tubes, a standard and cost-effective passive monitoring method, installed at five fixed roadside and residential locations within each neighbourhood and exposed for a continuous four-week period overlapping with the survey fieldwork, with results adjusted using local bias-correction factors following Defra (2022) guidance.

Analysis. One-way ANOVAs were used to compare perceived air quality and measured NO&sub2; across the three neighbourhood groups, with Tukey post-hoc comparisons where the omnibus test was significant. A Pearson correlation examined the individual-level association between distance to green space and perceived air quality across the full sample. All analyses used an alpha level of .05, and residuals were checked for approximate normality prior to running the ANOVAs.

Ethics and limitations of recruitment. The study received institutional ethical approval prior to fieldwork, and all participants gave informed consent and were free to withdraw at any point during the survey. Because recruitment relied on doorstep and on-site intercept sampling rather than a random sample drawn from the electoral register, the achieved sample cannot be assumed fully representative of each neighbourhood’s population; response patterns may, for instance, under-represent residents who work long or irregular hours. Quota targets for a broad spread of ages and household types were used to reduce, though not eliminate, this risk, and demographic profiles of the three achieved samples were broadly similar on age and household composition, reducing the likelihood that demographic differences alone drove the observed neighbourhood effects.

Results

Table 1 summarises perceived air quality and measured NO&sub2; concentrations by neighbourhood. Both measures followed the expected gradient, with the high-provision neighbourhood recording the most favourable values on both indicators and the low-provision neighbourhood the least favourable.

A one-way ANOVA confirmed a significant effect of neighbourhood on perceived air quality, F(2, 147) = 28.94, p < .001, with Tukey post-hoc tests showing all three pairwise comparisons significant at p < .01. Measured NO&sub2; concentrations also differed significantly by neighbourhood, F(2, 147) = 42.71, p < .001, following the same rank order, with the largest absolute difference between the high- and low-provision neighbourhoods (13.3 µg/m³). Across the full sample, distance to the nearest green space correlated negatively with perceived air quality, r(148) = -.34, p < .001, indicating that residents living closer to green space tended to rate their local air quality more favourably, independent of the broader neighbourhood-level pattern.

Discussion

The results show a consistent pattern across three independent indicators: neighbourhood-level green space provision, objectively measured NO&sub2; concentration, and residents’ perceived air quality all moved together in the expected direction, with the high-provision neighbourhood outperforming the low-provision neighbourhood on both the objective and subjective measures. This is broadly consistent with existing evidence that urban vegetation contributes to measurable, if modest, reductions in ambient pollutant concentrations (Nowak et al., 2014; Tallis et al., 2011), and extends this literature by demonstrating that residents’ own perceptions track this objective pattern reasonably closely in the present sample.

This alignment between perceived and measured air quality contrasts somewhat with Brody, Peck and Highfield’s (2004) finding of only a weak perception-monitoring correlation in a US context. One plausible explanation is that green space itself, unlike NO&sub2; concentration, is directly visible and experientially salient to residents; where Brody et al.’s participants had few visible cues to draw on, residents in the present study may have been implicitly, or explicitly, using the presence or absence of parks and street trees as a proxy for judging air quality, in a manner consistent with Bell et al.’s (2018) finding that green space use and perceived environmental quality are closely linked. If so, the correlation observed here may partly reflect a genuine perceptual heuristic linking visible greenery to inferred air cleanliness, rather than residents directly sensing pollutant concentrations.

Several limitations should be acknowledged. The cross-sectional, three-neighbourhood design cannot establish causation, and unmeasured confounds correlated with both green space provision and air quality, such as traffic volume, road type, or socioeconomic composition, may account for some of the observed association; future work using a larger sample of neighbourhoods and multivariate control for such confounds would strengthen causal inference. The single-item perceived air quality measure, while practical for a door-to-door survey, offers less nuance than a validated multi-item scale, and the four-week NO&sub2; monitoring window, while adequate to characterise typical concentrations, cannot capture seasonal variation. Despite these constraints, the consistency of the pattern across independent objective and subjective measures lends reasonable confidence to the study’s central finding.

The results carry implications for urban planning and environmental justice policy. Where green space provision is unevenly distributed across a city, as is commonly the case in relation to income and deprivation (Mitchell and Dorling, 2003), the present findings suggest that residents of lower-provision areas may be disadvantaged on both an objective air quality measure and their own lived sense of environmental quality simultaneously. This strengthens the case for prioritising green infrastructure investment in currently under-provided neighbourhoods, not only as an air quality intervention in its own right but as a means of addressing perceived, and potentially wellbeing-relevant, environmental inequality between areas of a city.

Conclusion

This study found that Sheffield neighbourhoods with greater green space provision recorded both lower measured NO&sub2; concentrations and more favourable resident-perceived air quality, with individual proximity to green space also predicting perception independently of neighbourhood. The alignment between objective and subjective measures suggests urban green space functions as both a genuine, if modest, air quality intervention and a visible cue that shapes how residents judge their local environment. Given the cross-sectional design, future research using a larger sample of neighbourhoods, longitudinal monitoring, and explicit control for traffic and socioeconomic confounds would help establish the causal mechanisms underlying this association and strengthen the evidence base for green infrastructure investment as an environmental justice intervention.

References

• Bell, S.L., Phoenix, C., Lovell, R. and Wheeler, B.W. (2018) ‘Green space, health and wellbeing: Making space for individual agency’, Health & Place, 34, pp. 97-106.
• Brody, S.D., Peck, B.M. and Highfield, W.E. (2004) ‘Examining localized patterns of air quality perception in Texas: A spatial and statistical analysis’, Risk Analysis, 24(6), pp. 1561-1574.
• Defra (2022) Local Air Quality Management Technical Guidance (TG22). London: Department for Environment, Food and Rural Affairs.
• Mitchell, G. and Dorling, D. (2003) ‘An environmental justice analysis of British air quality’, Environment and Planning A, 35(5), pp. 909-929.
• Nowak, D.J., Hirabayashi, S., Bodine, A. and Greenfield, E. (2014) ‘Tree and forest effects on air quality and human health in the United States’, Environmental Pollution, 193, pp. 119-129.
• Public Health England (2019) Review of Interventions to Improve Outdoor Air Quality and Public Health. London: PHE.
• Tallis, M., Taylor, G., Sinnett, D. and Freer-Smith, P. (2011) ‘Estimating the removal of atmospheric particulate pollution by the urban tree canopy of London, under current and future environments’, Landscape and Urban Planning, 103(2), pp. 129-138.
• White, M.P., Alcock, I., Wheeler, B.W. and Depledge, M.H. (2013) ‘Would you be happier living in a greener urban area? A fixed-effects analysis of panel data’, Psychological Science, 24(6), pp. 920-928.