How early warning systems help us predict flooding

A stream gauge is a humble thing.

But the Bay Area’s vast network of gauges, strategically placed and linked to powerful telecommunication lines and computers, offer early warning of rising waters, helping protect lives and property from flood damage.

“We can tell first responders, with some certainty, that our gauges are saying it’s going to be a flood-prone area in one hour and 45 minutes,” said Kevin Murray of the Palo Alto-based San Francisquito Creek Joint Powers Authority. “That gives them some time to get ahead of the disaster.”

This week, with a series of storms just past and more on their way, experts patrolled rivers and creeks to inspect and recalibrate their precious tools to ensure the accuracy of water data. After last Saturday’s fierce storm, they ask: Is the equipment still working? Have flow conditions changed due to downed trees, clogged bridges, eroded banks or blocked storm drains?

“We’ve been out with replacement equipment, assessing problems, assembling the pieces and doing repairs so that everything is operational,” said geologist Scott Brown with Berkeley-based Balance Hydrologics, a consulting firm that helps Bay Area agencies manage flood warning systems.

“With additional storms coming, we need to mobilize fast in order to get everything back up and running again,” he said.

Even as climate change increases the risk of floods, protection strategies are shifting away from structural measures, such as walls, gates and levees. The role of prediction, and warnings, is growing.

The gauges can provide days or hours of advance notice, giving residents time to sandbag, move cars, help the homeless, lift electronics and precious antiques off the floor — or seek higher ground.

On New Year’s Eve, a gauge on the Stanford campus alerted authorities at 8:43 a.m. that San Francisquito Creek, between Santa Clara and San Mateo counties, had reached “flood monitoring stage” and was continuing to rise. At 9:06 a.m., officials issued a public flood advisory. At 9:30 a.m., residents were alerted that within a half-hour, flooding was likely at Palo Alto’s narrow Pope-Chaucer Bridge. Water rose to about 21 feet, frighteningly close to overflowing the creek banks.

Other watersheds may allow more time, and earlier warning, said Jack Xu, senior engineer at Santa Clara Valley Water District. For instance, flooding along San Jose’s Guadalupe River can be forecast two to three hours in advance. Coyote Creek flooding may take a day. But in urban concrete channels, water may need only 30 minutes to flow from the mountains to downtowns.

Surveillance starts two to three weeks in advance of a storm, said Xu.

Weather is notoriously tough to predict in the Bay Area, a mountainous landscape perched on the edge of the cold ocean. Low and fast-moving jets of moist air scatter into a patchwork quilt pattern of precipitation, affecting locales differently.

Looking ahead at long-range weather forecasts, Xu said, “We see: ‘Is it a wet pattern? Is it a dry pattern?’ When you see back-to-back storms, you know you have to plan resources and think, operationally, about how you’re going to move water, if you need to.”

Plans get more targeted as time draws near, because forecasts improve within 10 days of a storm. “But if the forecast is inaccurate, everything downstream is wrong,” said Xu.

Using computer models, they calculate how much of the predicted rainfall will become runoff, based on past history, soil saturation and local geology.

The big decisions start five days before a storm, Xu said. There are regional conference calls with PG&E, other utilities and various cities, where they ponder: Should they release water from reservoirs to make room for additional runoff? Do they need to boost staffing? “We weigh the consequences with the risk,” Xu said.

Crews on the ground assess a creek’s condition, looking for debris and other risks. This effort is intensified if winds are forecast.

Once the storm lands, the role of rain gauges becomes critical. Perched in distant mountains, these gauges — pipes with a funnel, bucket and tipping mechanism at the top — measure precipitation. They tell officials what to expect.

Stream gauges have a membrane that precisely measures the depth of water and converts it into a flow rate. They transmit every hour and send a packet of four 15-minute time stamps. When a certain threshold is reached, the sensors can send data every five minutes.

Contra Costa County Flood Control & Water Conservation District manages 32 rain gauges and 16 stream gauges. The Santa Clara Valley Water District has a network of 70 stream and rain gauges. Alameda County has about 90 rain and stream gauges.

These gauges, powered by solar panels, send electronic signals to data loggers via radio, landlines, cell phone signals or satellites. With ever-increasing computer power, software processes the many signals into a computer database, which monitors the information as it is received. It triggers a warning when certain thresholds — say, water filling 80 percent of a creek’s capacity — are reached.

When waters run high, officials must decide whether to issue an electronic flood alert to local residents downstream. Because different locations have different flood risks, the warnings can be localized.

A revolution in technology is allowing for better forecasting. Over the past decade, the National Weather Service has developed weather models that are better suited to the West Coast, and widespread availability of cell services means it’s easier to transmit real-time information, Xu said. Supercomputing allows more advanced calculations, under many different scenarios.

New “X-Band Radar” is helping estimate rainfall in specific communities. In Contra Costa County, radar was installed last month on Rocky Ridge near Las Trampas Regional Preserve, according to hydrologist Mark Boucher. Radar for the Santa Clara Valley Water District sits atop the Penitencia Water Treatment Plant facility. They are part of a future regional network of radars — including Marin, Sonoma, San Francisco and Santa Cruz — that will provide local data about atmospheric rivers.