Data-Driven Bioclimatic Zoning in Sri Lanka: PCA and Clustering Analysis

Abstract

Driven by evolving lifestyles and the escalating demand for thermal comfort, Sri Lanka faces a critical absence of climate zone classifications necessary for constructing energy-efficient and climateresponsive buildings. This study addresses this gap by implementing bioclimatic zoning using a comprehensive 31-year weather dataset of 25 locations across the country. By applying Principal Component Analysis and Hierarchical Clustering to the 31-year weather data, Sri Lanka was classified into three distinct bioclimatic zones: Z1 (Hot and humid), Z2 (Cool and humid), and Z3 (Warm and humid). Bioclimatic potential analysis for each zone reveals natural ventilation as the most effective passive design strategy, demonstrating potential percentages of 64 ± 13%, 63 ± 10%, and 83 ± 4% in a year for Z1, Z2, and Z3, respectively. These findings underscore the crucial role of bioclimatic zoning in guiding the design of energy-efficient buildings in Sri Lanka. The approach contributes significantly to achieving national energy efficiency goals by leveraging climate-specific passive design strategies and reducing reliance on energy-intensive cooling systems. Moreover, the study not only classifies Sri Lanka into three bioclimatic zones but also emphasises the broader impact of implementing such strategies on sustainable construction practices. This research, therefore, stands at the intersection of bioclimatic zoning, sustainable building practices, and the evolving energy landscape.