For better groundwater resource management, water quality protection, enhanced resiliency of urban water supply promoting a balanced groundwater extraction and recharge, and implementation of a robust operation and maintenance policy, the Silverdale Water District is interested to improve its existing Water Tracker. Silverdale Water District serves over 25,000 people in the greater Silverdale area using 13 above-ground reservoirs that provides nearly 8 million gallons of water storage for potable water supply, pressure equalization, and fire suppression. These reservoirs are filled from 12 permanent wells drilled in three distinct aquifers. The objective of this paper is to develop daily and monthly water demand prediction models to predict future water production using historical water production and weather information so that the developed models can be incorporated into the Water Tracker tool. Besides using the Water Tracker for determining optimal groundwater extraction, and overall system management, it will also be used to conduct simulations to determine the potential well failure rates so that appropriate O&M strategies can be implemented to enhance the water supply resiliency. To develop a monthly demand prediction model, the univariate time series approach was used using the historical 9-year monthly average water production data for model fitting and 4-years data for model testing. To develop the daily water production model, however, a multivariate time series approach was used using the daily water production, daily maximum temperature, average daily maximum temperature, and monthly and daily precipitation for the same period of data used to develop the monthly demand model.
