Date of Award
Master of Science (MS)
R Scott. Summers
Water demand projections are crucial for efficient management of water utilities, especially given the stress from socio-economic growth and supply reductions and variations under climate variability. Currently, many water utilities rely largely on statistical models for average demand incorporating socio-economic and limited climate information. Water managers require projections of different demand attributes such as peak demand, demand exceeding desired thresholds, sustained period of exceedances, etc. There is increasing recognition that climate plays a substantial role in demand modulation and that it needs to be accounted better in water demand models. Two key gaps are identified with the state of knowledge that motivated this research - (i) the need to incorporate all attributes of weather and climate and (ii) the need to provide tools to model a variety of demand attributes that are simple and robust and can incorporate climate and socio-economic factors. This research makes the following contributions. (1) A suite of weather attributes are identified that are strongly related to attributes of water demand, these include hot/dry and wet/cold spells and average precipitation and temperature; (2) A Generalized Linear Modeling (GLM) framework to model monthly average, monthly peak and number days of demand exceeding a threshold, incorporating the weather attributes; (3) Introduction of Extreme Value Analysis (EVA) to model monthly peak demand and demand exceeding thresholds, which also uses climate attributes to model nonstationarity of the water demand variability (4) demonstration of the models from EVA to make projections of water demand extremes under climate change. All of these are developed and demonstrated for water demand data from Aurora Water, the water utility of Aurora, CO. These methods are simple and robust and can easily be modified to incorporate other covariates such as social and economic. Furthermore, they can be applied with limited resources and thus, provides an effective set of tools to most water utilities.
Haagenson, Erik Clifford, "Statistical Methods for Water Demand Modeling to Help Mitigating Climate Induced Risk" (2012). Civil Engineering Graduate Theses & Dissertations. 251.