Forecast and Planning Tools

Case Study: Seattle Public Utilities

The Seattle Public Utilities (SPU) water system experienced different types of droughts and potential shortages in 1987-88, 1992, and 1997-98. SPU’s responses to these three events illustrate the capability of an institution to learn from and successfully respond to adverse hydrologic watershed conditions, due in part through a progressive use of weather and climate-based forecasts and computer tools.

Drought of 1987-88

Summer 1987 began well with full reservoirs, but a hot dry summer followed by late fall rains (possibly related to the developing 1987-88 El Niño) created a serious water supply situation for the City of Seattle and its customers. Raw water quality declined, water use curtailments were implemented, instream flows for fish were reduced, and an emergency temporary pumping station was installed in the City’s main mountain reservoir to access the low water level in the reservoir. In response to this experience, the City developed a Water Shortage Contingency Plan (WSCP) to provide a plan of action for future droughts and potential water shortages. According to the WSCP, the City would respond to a potential or actual shortage in four stages based on increasing severity as progressively more serious conditions warrant:

SPU also began developing a state-of-the-art reservoir management and streamflow forecasting computer model system (SEAFM) for use in real-time water management decision-making and long range planning. The forecast model was designed to continually simulate the current hydrologic state of the watershed, and to use the latest climate forecasts from the National Weather Service to produce probabilistic streamflow forecasts up to 12 months out. Particularly during potential low water years, the SEAFM forecast model has become a very valuable tool for providing decision-makers with in-depth analyses and insight for risk management.

Drought of 1992

Another drought occurred in 1992 (an El Niño year) for somewhat different reasons and with a rather different response. Following a winter that produced a below normal mountain snowpack but in which SPU had followed standard flood-control rules by spilling water from their reservoirs, the normally rainy spring season produced below normal precipitation. SPU’s mountain reservoirs remained low during the typical spring refill period due to the small snowpack and low spring inflows, requiring SPU to progress rapidly to the mandatory stage of the WSCP by mid-May.

Throughout the hot and dry summer, raw water quality again declined sharply, this time prompting a decision to begin building a costly ozone-purification plant. In collaboration with other resource management agencies, SPU studied and evaluated methods and procedures to operate its mountain reservoirs and dams using dynamic flood-control rules that reflect actual watershed conditions and allow risk and uncertainty to be managed. SPU also worked with the Natural Resources Conservation Service to establish strategically located cooperative SNOTEL sites in the mountainous portions of the watershed to provide real-time snow and climate data. The SNOTEL network, along with SPU’s cooperative USGS real-time stream gaging network and SEAFM forecast model, has enabled the utility to successful implement this reservoir management approach.

The Strong El Niño of 1997-98

By 1997, early research from the Climate Impacts Group (CIG) on El Niño/Southern Oscillation (ENSO) impacts on Pacific Northwest (PNW) temperature and precipitation, combined with the June 1997 forecast for a strong El Niño over the winter of 1997-98, led to SPU’s incorporation of the El Niño forecast into reservoir management decisions. An El Niño Task Force was established within the agency which met twice monthly during the winter and spring as the impacts of the El Niño conditions in the tropical Pacific played out in the PNW region (A. F. Chinn, Seattle Public Utilities Water Management, personal communication, 1998).

Winter snowpack turned out to be slightly below normal in 1997-98 and a hot dry summer followed. But SPU had aggressively educated employees and a number of measures were implemented to conserve water or otherwise increase supply. For example,

Measures like these allowed the drought to pass with the public experiencing no water shortage. Adequate advanced planning – due in part to SPU’s continued monitoring of the NOAA and CIG climate forecasts – and improved systems operations were sufficient to address the potential shortage that occurred.

SPU Today

In integrating weather and climate-based forecasts into its operations, SPU is an uncommonly adaptable resource-management agency. When CIG first interviewed senior managers in spring 1996, they were aware of the National Centers for Environmental Prediction ENSO forecasts but did not find them very useful. By spring of 1998, after CIG worked closely with SPU, SPU’s water management office reported, “…[our] staff regularly obtains and uses NOAA’s Climate Prediction Center forecasts and various other NOAA weather and climate products made available via the internet websites” (A. F. Chinn, Seattle Public Utilities Water Management, personal communication, 1998). Each year now, SPU produces probabilistic streamflow forecasts and water supply analyses based on ENSO conditions. SPU continues to work closely with CIG researchers to explore and learn more about how to incorporate climate forecasts based on other patterns of climate variability, such as the Pacific Decadal Oscillation and the Madden-Julian Oscillation.