Technical Report No. 91
James E.T. Moncur
This paper investigates the usefulness of two procedures for analyzing water use time series data. A simple procedure for translating price data into a form suitable for a standard regression estimation of the demand function for water is first noted. Then a simple spectral density estimation model is applied to Honolulu water consumption data in search of identifying periodicities or cyclical regularities of water demand. Both procedures show promise of useful interpretation, though the conclusions are tentative in keeping with the scope of this project and the limited data available.