A well-developed water supply plan can be a very informative tool for water users to understand the ability to sustain their current and future water supply needs. However, the means to develop such a plan is relatively complicated and requires adherence to critical principles to ensure plan acceptance and value as a useful water management tool.

Use of Predictive Impact Models – So many times the models used for predicting impacts to the natural systems are not sufficient to determine the cause and effect relationships asserted by the plan developer. This is especially valid with regional groundwater flow models with limited regional datasets. It should be noted that models of this type are best representations of real world conditions and should only be used as a tool during hydrogeologic investigations as a means of interpreting hydrogeologic control parameters and for formulating ideas of groundwater flow dynamics. However, many plan developers overstep their ability to predict potential adverse impacts by placing absolute reliance on the predictive outcomes of regional groundwater flow models. Long-term predictive models for water supply plans should be used as a tool to identify evidence of a problem that may occur and monitoring should ensure versus the determination that a problem will definitely occur and premature restrictive actions must be placed into motion as a preventative strategy. Predictions far into the future should never be taken as definitive. As stated by Konikow and Bredehoeft (1992) regarding groundwater flow models:

"It is naïve to believe that we will somehow validate a computer model so that it will make accurate predictions of system responses far into the future. In a sense, emphasizing validation deceives society with the impression that, by expending sufficient effort, uncertainty can be eliminated and absolute knowledge be attained."

Importance of a Well Designed Data Collection Network – An accurate and comprehensive data collection program is the key to calibrating and evaluating a predictive regional groundwater flow model. As with all groundwater flow models, a degree of uncertainty exists and all modeling predictions should be continuously re-evaluated with any newly obtained data. Continual data collection allows the modeler to frequently modify, adjust, and/or recalibrate the model under a series of field conditions and have a better accounting for the cause and effect relationships of water withdrawals. The lack of an extensively regional and reliable dataset to calibrate a predictive groundwater or surface water flow model makes the model output less acceptable and subject to criticism by those who would disagree with the model results.

Input from User Groups – The water supply plan development process must be a very publically transparent and inclusive process that includes input from all user groups to fully understand the implications of the plan. This process will ultimately lead to the best plan recommendations and action items to avoid both adverse impacts to environmental receptors and the economic health of the water users.