Integration and Visualization of Meteorological Data in Digital Twins: A Framework for Real-time Weather Simulation Using the AnywhereXR Platform

By Xiaoyu Yang

Abstract
This study investigates the integration of meteorological data within digital twin environments through a dual methodology addressing technical implementation requirements and perceptual evaluation dimensions. A framework was developed to facilitate the bidirectional communication between computational environments, implementing spatial interpolation methodologies to transform coarse-resolution meteorological data into fine-grained visualization. Comparative analysis revealed parameter-specific optimization patterns wherein bilinear interpolation achieved optimal performance for wind fields (R² > 0.99) while IDW demonstrated superior accuracy for precipitation field reconstruction (RMSE < 0.27 mm/h). Significant geographical sensitivity was observed across interpolation methodologies (p < 0.001), establishing location-specific optimization requirements. The perceptual evaluation component utilized a mixed-methods approach examining six distinct visualization scenarios (2 modalities × 3 geographical contexts) involving the July 9, 2024 storm event. In contrast to theoretical expectations, 2D visualizations elicited heightened risk perception compared to 3D alternatives, suggesting attentional distribution effects. Geographical context emerged as a critical determinant of both interpolation performance and risk perception outcomes, with significant medium × location interactions demonstrating context-dependent visualization efficacy. These findings establish methodological foundations for weather visualization in digital twin environments while advancing theoretical understanding of how visualization dimensionality influences risk perception and decision-making processes.

Keywords: Digital twins; Weather visualization; Spatial interpolation; Risk perception