HR: 14:20h
AN: H13K-03 INVITED     [Abstracts]
TI: A Statistical Model-Based Decision Support System for Managing Summer Stream Temperatures with Quantified Confidence Analysis
AU: * Neumann, D W
EM: david.neumann@colorado.edu
AF: Center for Advanced Decision Support for Water and Environmental Systems (CADSWES), Univ. of Colorado, UCB 421, Boulder, CO 80309 United States
AU: Zagona, E A
EM: zagona@colorado.edu
AF: Center for Advanced Decision Support for Water and Environmental Systems (CADSWES), Univ. of Colorado, UCB 421, Boulder, CO 80309 United States
AU: Rajagopalan, B
EM: Rajagopalan.Balaji@Colorado.edu
AF: Univ. of Colorado, Dept. of Civil, Environmental, and Architectural Engineering, UCB 426, Boulder, CO 80309 United States
AB: Warm summer stream temperatures due to low flows and high air temperatures are a critical water quality problem in many western U.S. river basins because they impact threatened fish species' habitat. Releases from storage reservoirs and river diversions are typically driven by human demands such as irrigation, municipal and industrial uses and hydropower production. Historically, fish needs have not been formally incorporated in the operating procedures, which do not supply adequate flows for fish in the warmest, driest periods. One way to address this problem is for local and federal organizations to purchase water rights to be used to increase flows, hence decrease temperatures. A statistical model-predictive technique for efficient and effective use of a limited supply of fish water has been developed and incorporated in a Decision Support System (DSS) that can be used in an operations mode to effectively use water acquired to mitigate warm stream temperatures. The DSS is a rule-based system that uses the empirical, statistical predictive model to predict maximum daily stream temperatures based on flows that meet the non-fish operating criteria, and to compute reservoir releases of allocated fish water when predicted temperatures exceed fish habitat temperature targets with a user specified confidence of the temperature predictions. The empirical model is developed using a step-wise linear regression procedure to select significant predictors, and includes the computation of a prediction confidence interval to quantify the uncertainty of the prediction. The DSS also includes a strategy for managing a limited amount of water throughout the season based on degree-days in which temperatures are allowed to exceed the preferred targets for a limited number of days that can be tolerated by the fish. The DSS is demonstrated by an example application to the Truckee River near Reno, Nevada using historical flows from 1988 through 1994. In this case, the statistical model predicts maximum daily Truckee River stream temperatures in June, July, and August using predicted maximum daily air temperature and modeled average daily flow. The empirical relationship was created using a step-wise linear regression selection process using 1993 and 1994 data. The adjusted R² value for this relationship is 0.91. The model is validated using historic data and demonstrated in a predictive mode with a prediction confidence interval to quantify the uncertainty. Results indicate that the DSS could substantially reduce the number of target temperature violations, i.e., stream temperatures exceeding the target temperature levels detrimental to fish habitat. The results show that large volumes of water are necessary to meet a temperature target with a high degree of certainty and violations may still occur if all of the stored water is depleted. A lower degree of certainty requires less water but there is a higher probability that the temperature targets will be exceeded. Addition of the rules that consider degree-days resulted in a reduction of the number of temperature violations without increasing the amount of water used. This work is described in detail in publications referenced in the URL below.
UR: http://cadswes.Colorado.edu/publications/journal_articles.html
DE: 1847 Modeling
DE: 1860 Streamflow
DE: 1878 Water/energy interactions (0495)
DE: 1880 Water management (6334)
DE: 6309 Decision making under uncertainty
SC: Hydrology [H]
MN: Fall Meeting 2005