Patterns and Predictors of Residential Indoor Water Use Across Major US Cities

Abstract

This study investigates residential indoor water consumption variability across 39 US cities using data from 26,441 single-family smart water meters. Employing functional data analysis and mixed-effects random forest, we identified distinct usage patterns across city clusters, with 13 high and 6 low water-using cities (all in coastal California) differing significantly from 20 medium water-using cities. Shower and toilet use were primary drivers of indoor use differences between clusters, influenced by both behavioral and fixture efficiency factors. The presence of appliances, certain household features, and weather also affect indoor water use, with varying influence on indoor water use across clusters. Our findings highlight the effectiveness of state-level water efficiency interventions and emphasize the importance of considering both behavioral factors and appliance efficiency in conservation strategies, providing valuable insights for targeted water demand management in urban areas.