Assessing Climate Sensitivity of LEED Credit Performance in U.S. Hotel Buildings: A Hierarchical Regression and Machine Learning Verification Approach

Abstract

This study examines how climatic factors influence the predictive power of LEED credits in determining certification outcomes for hotel buildings across the United States. Using data from 259 LEED-NC v2009 certified hotels, project-level information was integrated with 30-year climate normals from the PRISM database and Building America climate zones. A three-step hierarchical linear regression was conducted to identify the LEED credits that most strongly predict total certification points while controlling for project size, certification year, and baseline climatic conditions, and to test whether climatic factors moderate these relationships. Regularized Linear Regression (LASSO) was then applied to address multicollinearity and assess model stability, followed by Support Vector Regression (SVR) to capture potential nonlinear relationships. This integrated methodological framework, combining hierarchical regression for interpretability, LASSO for coefficient stability, and Support Vector Regression for nonlinear verification, provides a novel, multi-dimensional assessment of climate-sensitive credit behavior at the individual credit level. Results show that energy- and site-related credits, particularly Optimize Energy Performance (EA1), On-Site Renewable Energy (EA2), Green Power (EA6), and Alternative Transportation (SS4), consistently dominate LEED performance across all climate zones. In contrast, indoor environmental quality credits exhibit modest but significant climate sensitivity: higher mean temperatures reduce the contribution of Increased Ventilation (EQ2) while slightly enhancing Outdoor Air Delivery Monitoring (EQ1). Cross-model consistency confirms the robustness of these findings. The findings highlight the need for climate-responsive benchmarking of indoor environmental quality credits to improve regional equity and advance the next generation of climate-adaptive LEED standards.

Citation

Goodarzi, M., Goodarzi, A. N., Naseri, S., Parsaee, M., & Abazari, T. (2025). Assessing Climate Sensitivity of LEED Credit Performance in US Hotel Buildings: A Hierarchical Regression and Machine Learning Verification Approach. Buildings, 15(23), 4382.