In her office at Stanford University, Corrine Bowers takes a break from her research. Photo by Xavier Martinez
Many residents of the Bay Area spent early January assembling barricades of sandbags outside their homes or stockpiling food and candles, anticipating a myriad of impacts from the region’s prolonged atmospheric river. But risk analysis expert Corinne Bowers suggests that the best disaster preparation can be done year-round, using new and clear terminology, data from insurance companies, and a whole lot of statistical modeling.
Bowers, a fifth-year civil engineering PhD student at Stanford University, studies the impact of extreme weather events. Atmospheric rivers, narrow and intense weather patterns that carry an amount of water roughly equal to that found at the mouth of the Mississippi River, have been a particular interest of hers. Bowers’ research takes place on complex software systems, where she can access vast amounts of data and run simulations to predict the damage caused by these storms.
Her first project focused on building a “top-down” approach to model the impact of an atmospheric river. When an atmospheric river occurs, it brings with it rain, which then can lead to flooding, causing damage that culminates in a monetary loss for people and businesses, and a physical loss for the environment. Using meteorological data about the characteristics of the weather event, she used statistical programming to predict the impacts, culminating in an expected loss expressed in dollars.
“Starting from one storm, I modeled all of the possibilities of the loss that could have occurred from that storm, if the soil moisture had been different, or if the houses had been a little older, or if there were some changes in the exact plant cover next to a river,” Bowers said.
Her second item of business reversed that process. Using data on the number of insurance claims in the days following an atmospheric river from the National Flood Insurance Program, Bowers employed machine learning to identify which characteristics of the natural and built environments contributed to the impacts of the storm.
Bowers’ most recent work is perhaps her most relevant, as it relates to the changing composition of atmospheric rivers. She is completing her research by studying “sequences,” a term she coined to refer to multiple atmospheric rivers that occur one after the other. By examining atmospheric trends that occur in the time before a sequence of atmospheric rivers, she hopes to be able to predict the impacts of these storms– which occur at the strongest intensity.
The January 2022 storms are a perfect example. Storms occurred in quick succession, with actual rainfall exceeding the predicted amount by over 50 percent. Some communities in the Bay Area suffered millions of dollars in damages.
“This [storm] in January, you could kind of hear in the news coverage,” Bowers said. “People had some sense that it was going to be a ‘one after another’ situation.”
Bowers assesses risks by considering the exposure and vulnerability of a structure or environment, along with the hazard posed by a weather event. Diagram courtesy of Corinne Bowers
Research with a wide reach
While some academic research never reaches the public eye, Bowers believes that her findings can help those living in the areas where atmospheric rivers wreak havoc. Her first target for sharing these results is with the State of California itself. Whether it is buying back wetlands as a preventative measure or strategically protecting infrastructure to limit damage, Bowers sees a plethora of uses from her research.
Bowers’ work is in line with a larger trend of using the results of data-heavy projects to make informed regulatory decisions. California’s Forecast-Informed Reservoir Operations uses weather forecasts to strategically release water, which scientists hope will reduce waste and combat droughts. Under the program, a dam will hold more water if the projected weather conditions are dry. It will release water to make room for more if a storm– like an atmospheric river– is in the forecast.
Smaller governments may receive the greatest benefit from this research, Bowers hopes. A 2018 study by the National Institute of Building Sciences found that every one dollar spent by government agencies toward disaster mitigation saved society six dollars, a margin that may incentivize local officials to take actions toward mitigation. Bowers says this expected impact encourages her to pursue her research.
“I research these things because I want to reduce the effects on the communities that are exposed to these hazards. The way that I see my research being used is sort of at the community decision maker stakeholder level.”
Bowers’ research on the drivers and consequences of atmospheric rivers can be of use to the general public, too. Her modeling is specific enough that it can identify the optimal sandbag height for a given residence or offer a homeowner insight into the benefits of raising their home (hint: it is almost always financially beneficial).
But the research has an even larger impact than concrete steps that officials and the public can take to prepare for a disaster. In creating the term “sequences,” Bowers believes that she can increase the public’s awareness of these storms, maintaining their relevance even during the summer months.
It’s true that the conversation around these storms have grown in recent years. The term “atmospheric river” was only established in the late 1990s, despite the fact that the weather patterns have always existed. Bowers hopes that, by tying the term “sequence” to the discussion around atmospheric rivers, her research can help strengthen the tie between scientific findings and the public’s awareness of that science.
“There are a lot of factors at play that make it very attractive to ignore the reality that we are in this climate that is much more prone to extremes. And there is, on the science side, a lack of a connection of relevance to decision makers,” Bowers said.
Bowers examines California trends of rainfall and flood insurance claims in her office at Stanford University. Photo by Xavier Martinez
Academics like Bowers demonstrate the importance of continuous preparation outside of the time directly before a disaster. In a state like California with polarizing weather extremes, the mitigating effects of this preparedness can help save thousands of dollars – and lives. Big data and clear communication, such as the research that Bowers produced, can help residents prepare for the worst.
“Everyone tends to forget that it does flood in California whenever we have a drought,” Bowers said. “We need to always be preparing for floods. There are things you can do with 24 hours’ notice, but it's really better to start as far in advance as you can.”
---
Xavier Martinez is a senior at Stanford who is passionate about incorporating big data into everyday journalism. Originally from Wenatchee, Washington, he enjoys learning about and reporting on issues like agricultural labor, immigration, and rural governance. Outside of the classroom, Xavier likes to try new restaurants in the Bay Area and explore the Santa Cruz mountains.
Comments