How does three-dimensional landscape pattern affect urban residents' sentiments

Abstract

The impact of urban street landscapes on residents' sentiments is a critical concern. However, the current representation of street landscapes through landscape pattern two-dimensional metrics (LP2DM) derived from remote sensing images neglects the perceptibility of residents' visible environments at the eye level. To address this gap, we developed a novel landscape pattern three-dimensional metric (LP3DM) to quantitatively represent landscape perceptibility based on four individual perception dimensions: green space, gray space, openness, and crowding. We then investigated the relationships between LP3DM and residents' sentiments using Baidu street view images and Weibo social media textual big data in Beijing, China. Our results demonstrate that LP3DM is more significant correlated with residents' sentiments than LP2DM (average contribution, ACLP2DM=0.025, ACLP3DM=0.054). Notably, the greenness metric exhibited the highest contribution (AC=0.12), with the greenness three-dimensional metric showing a positive correlation (r = 0.15, p < 0.01) with residents' sentiments, while grayness exhibited a slightly negative correlation (r = −0.087, p < 0.1). Our study highlights the importance of considering the perceptibility of natural landscape elements in addition to their quantity during urban construction to enhance residents' sentimental well-being. Overall, our LP3DM framework offers a promising approach to capture residents' landscape perceptibility and inform urban planning and design decisions.