Analysis of Santa Monica Water Usage Data for Water Conservation

Historically, per capita water demand has tended to increase proportionately with population growth. However, the last two decades have exhibited a different trend; per capita water usage is declining despite a growing economy and population. Subsequently, city planners and water suppliers have been struggling to understand this new trend and whether it will continue over the coming years. This leads to inefficient water management practices as well as flawed water storage design, both of which have adverse impacts on the economy and environment. Water usage data, provided by the city of Santa Monica, was analyzed using a combination of hydro-climatic and demographic variables to dissect these trends and variation in usage. The data proved to be tremendously difficult to work with; several values were missing or erroneously reported, and additional variables had to be brought from external sources to help explain the variation. Upon completion of the data processing, several statistical techniques including regression and clustering models were built to identify potential correlations and understand the consumers’ behavior. The regression models highlighted temperature and precipitation as significant stimuli of water usage, while the cluster models emphasized high volume consumers and their respective demographic traits. However, the overall model accuracy and fit was very poor for the models due to the inadequate quality of data collection and management. The imprecise measurement process for recording water usage along with varying levels of granularity across the different variables prevented the models from revealing meaningful associations. Moving forward, smart meter technology needs to be considered as it accurately captures real-time water usage and transmits the information to data hubs which then implement predictive analytics to provide updated trends. This efficient system will allow cities across the nation to stay abreast of future water usage developments and conserve time, resources, and the environment.