Pessimism or optimism? A tri-spatial analysis of public risk perception bias and its influencing mechanisms: evidence from 31 major cities in China

Background: Public risk perception bias (PRPB) is a critical psychological and social determinant that substantially shapes public health outcomes and the effectiveness of safety governance-particularly during large-scale crises like the COVID-19 pandemic. Amid increasingly complex urban risk landscapes, traditional single/dual-spatial frameworks fail to capture PRPB's multi-source and interactive nature. Methods: This study develops an Extended Tri-Spatial Prism Model to measure PRPB across 31 major Chinese cities (2017-2023) using multi-source data and explores its spatiotemporal dynamics and multi-level influencing mechanisms via a mixed-methods approach (Delphi technique, entropy weighting, GIS spatial analysis, multiple regression, mediation analysis). Results: (1) PRPB is dominated by pessimism with marked tri-spatial heterogeneity-physical and cyber spaces show pronounced pessimism, while social space remains cognitively neutral; (2) PRPB exhibits a "north-optimistic, south-pessimistic" spatial pattern and a COVID-19-disrupted V-shaped temporal trend (2017-2023); (3) Individual and urban factors shape PRPB through the mediation of psychological stress, social capital, media rationality, and government trust. Conclusion: This study advances the theoretical understanding of PRPB by systematically delineating its tri-spatial characteristics and underlying mechanisms. These findings provide theoretical insights and practical guidance for optimizing public risk communication systems and urban safety governance frameworks.