Reshaping landscape factorization through 3D landscape clustering for urban temperature studies.

As urban populations grow and cities expand, the challenge of managing urban heat and its environmental impacts becomes increasingly critical. Traditional methods for analyzing urban temperature dynamics often fall short in precisely capturing the complexity of urban landscapes. This paper introduces the 3D Landscape Clustering (3LC) framework, a new tool designed to analyze urban temperature dynamics by factoring in landscape variables. It clusters landscapes into homogeneous groups using high-resolution 3D land cover maps. The 3LC adopts a clustering mechanism to enhance flexibility and objectivity in landscape categorization, thereby enhancing the depth and accuracy of urban climate studies and moving beyond traditional classification frameworks such as the Local Climate Zone (LCZ). Case studies demonstrate its capability to provide detailed insights into the relationships between urban landscape features and temperature variations. Additionally, the paper details how the framework can excel in multi-city analyses and outlines advanced analytical techniques. Promising research opportunities and limitations are identified. This research reshapes our approach to landscape categorization, advancing our understanding of the interactions between landscape and climate dynamics, and contributing to more sustainable, climate-resilient cities. • Clustering with detailed 3D land cover maps enhanced flexibility and objectivity. • Precisely identifies diverse landscape categories with distinct temperature profiles. • Supports quantitative-semantic querying, diverse sampling, and multi-scale analyses. • Multi-city case studies proved the framework's transferability and versatility. • Identified opportunities and challenges, advancing urban climate resilience. [ABSTRACT FROM AUTHOR]