Worth of short‐term rainfall forecasting for combined sewer overflow control

Abstract

Real‐time, short‐term rainfall forecasting is gaining recognition as a valuable input to more effective performance of a variety of urban water management activities, including controlling the incidence of untreated combined sewer overflows. An important question is, What levels of forecast error can be tolerated before it is better to abandon adaptive control policies utilizing forecast information in favor of simple reactive control methods? Experiments with an autoregressive‐transfer function model for shortterm forecasting are presented, utilizing the San Francisco North Shore Outfalls Consolidation Project as a case study. A split data technique is used to gain insight into expected forecast errors for selected overflow‐producing storms varying from high intensity‐low duration to low intensity‐high duration. These results are then compared with the performance of the planned system, utilizing automatically controlled gates in a large shoreline tunnel, for various levels of forecast error. The results of a limited number of simulation runs indicate that expected forecast model errors are generally lower than the error threshold above which reactive policies become more attractive.

FORECAST UNCERTAINTY IN WATER SUPPLY RESERVOIR OPERATION1

“…Therefore, decisions must be based on the best possible information (Yeh, 1985). Labadie, et a!. (1981), investigate the worth of shortterm rainfall forecasting for combined sewer overflow control.…”

A model to forecast flow with uncertainty analysis

“…The response time of a basin is the criterion for selection of the above elements. For example, a rainfall forecasting model is normally required acting in unison with a rainfallrunoff forecasting model to extend the forecast lead times when the basin has a rapid response (Labadie et al, 1981;Georgakakos, 1986aGeorgakakos, , 1986bBertoni et al, 1992;Lardet & Obled, 1994).…”

The Potential of Fuzzy Multi-objective Model for Rainfall Forecasting from Typhoons

“…In the hydrological approach, the stochastic method is commonly used to perform very short-term rainfall forecasting, based on the persistence of the rainfall process (Labadie et al, 1981;Obeysekera et al, 1987). For instance, Burlando et al (1993) applied an autoregressive-moving average process to forecast hourly rainfall.…”