NONLINEAR EFFECTS OF TEMPERATURE ON WATER DEMAND

Measuring the relationship between temperature, among other weather variables, and the demand for water is crucial for estimating the effects of climate change on water consumption. Water consumption data are available for individual residential, commercial, agricultural, and industrial premises for each monthly billing period. A simple way to model the relationship between water consumption and temperature for a given premise would be to compare the quantity of water consumed in a given billing period against some average temperature over the same interval. However, this method would not account for within-period temperature variation that affects the amount of water consumed from day to day. One solution is to construct a fine-scale weather data set, then compare water consumption to the full distribution of temperatures experienced in that time period. This allows for a non-linear relationship that better accounts for the variation of temperatures within a billing period. After estimating this relationship for various types of consumers, we can use it to estimate changes to water demand resulting from a variety of developmental and climate change scenarios.