Abstract
Climate zoning for buildings (CZB) is one of the basics in building design. A well-defined CZB enables buildings to be optimized according to the specific demands of each climate region. Available CZB methodologies, however, are becoming highly heterogeneous with respect to available input data and classification techniques. This review provides the most comprehensive synthesis of CZB to date, analyzing 111 key scientific publications and 44 national codes. Unlike existing CZB reviews, which categorize studies using a broad inconsistent mix of climate variables and classification schemes, this work directly addresses two fundamental questions of any climate zoning: “what” is classified (input data types) and “how” the classification is performed (approaches and algorithms). By linking input data to classification logic and quality objectives, this review provides a clearer and more consistent basis for analyzing and comparing heterogeneous CZB practices. The analysis revealed substantial variability in current CZB approaches, with predefined interval methods dominating the field, accounting for 66.2% of the approaches, followed by ML classification at 32.4%, and statistical techniques, which are minimally represented at 1.4%. Despite the increasing adoption of performance-based CZB, it remains limited, with approximately 12% of cases. Besides, six major challenges in achieving adequate CZB mapping quality were indicated. Few emerging trends and improvement options were identified. Ultimately, a standardized machine learning–based framework was proposed, providing a structured approach for effective and resilient CZB mapping.