Evidence-based thresholds for harmonizing vertical growth and mountain preservation in urban skylines: a case study of Fuzhou, China

Urban skylines with mountainous backgrounds sustain urban identity and public well-being, yet are challenged by urban vertical growth. Currently, a lack of scientific evidence undermines visual impact controls for such skylines. This study quantified public preferences through a photo-based survey on Fuzhou’s Fu Forest Trail (N = 450, 60 skyline samples), integrating correlation, Poisson LASSO and negative binomial regression, and decision tree analysis. The key findings reveal that: (1) visual preference is predominantly pleasure-driven, showing a strong positive correlation with pleasure scores (r = 0.664, p < 0.001) and a negative trend with arousal; (2) a core set of six physical parameters was identified and hierarchically validated, with building contour volatility (PVB) as the most influential, followed by the horizontal position of the tallest building (PHSR_2) and mountain contour volatility (PVM); and (3) the decision tree model uncovered critical parameter interactions and compensatory pathways, leading to specific, evidence-based threshold recommendations for planning: maintain PVB within [5.35, 12.37], PHSR_2 within [0.85, 1.62], and PVM within [2.78, 6.17]. These findings provide a scientific basis for data-driven skyline management, enabling policymakers to implement targeted design guidelines and control measures that balance urban growth with the preservation of natural landscapes and aesthetic quality.