Data, Fire and the Fight to Keep California Insurable

Story by:

• Mika Ono- [email protected]

Published Date

A Call for Continued Partnership

"As Californians, none of us are strangers to natural disasters, from earthquakes to floods and, yes, wildfires, fires that are getting bigger, hotter, deadlier and more destructive with every passing year. Now, many of us know what it's like to drop everything at a moment's notice, to have to evacuate our communities and to wonder whether we'll ever see our homes again. And increasingly, families aren't just worried about fire; they're also worried about whether they can insure their homes. In 2026, we need to be doing much more to help communities prepare and to become a resilient before the next fire breaks out. That means investing in innovation, using better data, better technology and smarter planning to reduce risk at the community level… Meeting that challenge will require continued partnership between government, researchers, industry and innovators like you."

Beyond Historical Loss for Rate Setting

"About a year ago, the [California Department of Insurance] overhauled many of their regulations related to homeowners' insurance in the state, specifically as it relates to wildfire rates, and created a process for allowing for models to be used for those rates with a rigorous review process… Some of the things that these models can do that the previous system of historical losses could not [is] account for changes in land use, new developments on the landscape, changes in fuels, [and] changes in climate, with more fire weather days. … Replacement costs and rebuilding costs have increased over 60% since 2015, so if you're using historical losses directly, you're not accounting for those changes… You can account for large-scale mitigation, property-level mitigation… And we think that it's appropriate to be able to account for all these really important actions people are taking."

The Journey of a Catastrophic Event

"Let's talk about how [a catastrophic event] unfolds. It's immaterial what causes the fire - just the fact that it starts under extreme conditions. Drought and wind can enable that fire to burn into a community. Vegetation produces embers, and those embers can travel very long distances under high wind conditions. When they land in a community - whether it's on a structure or maybe passing through a vent and inside the structure - they can directly ignite that structure, leading to its loss, or they can land around the structure, igniting combustible items that surround it, and then those resulting fires can spread to the homes leading to loss in the start of that conflagration unfolding… Once a home ignites, there's a 93 to 94% chance it will be a complete loss. It goes without saying that we need to prevent the ignition, because there's a very low probability that it will survive once that occurs. We look at connected fuel pathways, and how we can break that chain so that the fire doesn't lead to uncontrolled spread."

An Economic Approach to Wildfire Risk

"We want to do two things [to manage wildfire risks]: we want to engage in what I would call physical protection - hardening assets, making things less likely to be destroyed - and we want those investments to pass a cost-benefit test. … Then we want to turn to insurance markets or other financial protection programs to make sure that residual risk is managed and that, when the worst inevitably happens, those unlucky people have resources available to get their lives back on track. It's obvious why insurance markets are so important for financial protection, but insurance markets can help us with physical protection as well. When we have well-functioning insurance markets, the premiums are going to send ex ante signals about the value of investing in cost-effective mitigation. Insurance has a lot of potential to help us manage these escalating risks."

The Problem of Home Retrofitting

"Our codes didn't come here all in one day. It was a series of events that slowly built those stricter codes into place. But the key point here is that free structures built before 2009 were built before the California Wildland-Urban Interface (WUI) codes, and those structures built after that point are built to those codes [which, as the previous speaker noted, have been staggeringly effective in reducing the loss of new homes]. … It's really easy to build to those codes for new builds. The challenge is: how do we retrofit those 2.1 million homes to lower the chance of them burning down in a wildfire? Under Title 24 Part 7 of the WUI code, an appendix outlines our home-hardening retrofit recommendations; hopefully, that can help the 2.1 million homeowners to start moving that needle forward."

Modeling Fire Behavior

"In developing our fire behavior models, we analyze a lot of experimental data that we get from our colleagues in the different agencies, the Forest Service for example, but we also collect a lot of data during small-scale controlled experiments… In one part of our work, we study embers under a microscope and some pycnometers to look at their density, shape, sizes, etc. … [Results] indicate brittle failure as the key mechanism of the origin of the origin of some particles. That led to our hypothesis that embers are not only dependent on the flow and the wind, but also on how the fuel [is burning] and what fuel is burning… We also track the flame height fluctuations and the heat release rate, and we were able to look at ember velocity and acceleration statistics. … Our hypothesis for what's happening with the embers that are traveling the very long distances, especially large embers, is that there is a sudden break in the material. That leads to a sudden burst of embers, and these bursts are coinciding with strong accelerations in the wind. That is the boundary condition for very long-range transport."

24/7, 360-Degree Data with ALERTCalifornia

"How do we get to a point where we can detect or at least confirm fires, monitor fires and help our friends at CAL FIRE, the fire departments, who run towards the fire when most of us would say, 'Let's retreat'? Our approach is really to look at how to use data, and not just any data, but better Big Data - data that can be turned into action for hazard resilience… The question is: how do we acquire the best possible data? How do we get the right sensor to the right location at the right time? How do we stream this data to where it needs to be? How do we curate it, document it, make it accessible? How, ultimately, do we analyze it, model it, simulate it, visualize it, engage with it? And how do we get the data into the hands of the first responders, the stakeholders, the public at large, to make informed decisions? That really is all at the heart of what ALERTCalifornia is all about. ALERTCalifornia is a state-scale sensor network with, currently, 1,230 monitoring cameras that provide continuous, 360-degree coverage across the state. … It's real-time data that's accessible to first responders. It's accessible to the public, so anyone, anywhere, at any time, can go to cameras.alertcalifornia.org and look at any of these data feeds."

The Drivers of Catastrophic Losses

"Catastrophe losses are increasing and increasing quite rapidly. Of course, everybody is concerned now and wondering what is going on. … Before we get to climate change, there are three other factors that are actually more important. One is where we build. … We build very close [to the water], in harm's way. The next thing is what we build. … We're building more amenities - not a one-and-a-half-bathroom-, but a three-bathroom house; not a one- or two-car garage, but a three- or four-car garage. … But the most important reason is the cost to build. If we look at construction costs over the past 20 years, they have increased by almost by almost double. … I visited the Altadena area yesterday, and I talked to a couple homeowners who have also become general contractors. This man was telling me that the cost to rebuild homes, the bare minimum, was $300 a square foot, going up to $1,000 a square foot. Now that means a 2,000 square foot house is going to cost $2 million to rebuild. … It comes out to about a 1% increase per year due to climate change, which we need to quantify. But it's relatively small, compared to the increasing construction cost, which is on the order of 5 or 6% per year, or even higher, today."

Risk Extends to the East Coast

"Over the last dozen years, California alone has experienced more than 90 wildfires, and most of these wildfires have resulted in … at least 10 structures burned, destroyed or damaged. This increasing issue is what prompted us to look at what's next. Almost 50% of the threatened kind of WUI environment is actually on the East Coast. We did some analysis recently for a customer advisory board, and just looking across the whole country, there is much more underappreciated fire risk on the East Coast. A lot of the attention, deservedly so, gets focused on the West Coast, but on the East Coast that risk is not to be undersold."

Time and Talent to Move California Forward

"I wanted to say 'hello' and congratulate our premier research university in California and all of you for coming together with your diverse backgrounds and expertise to share thoughts, best practices, knowledge and information about how to really put together a framework for data-driven insurability in times of disaster planning and response. We have a lot of work to do in California. We have to create a broader market to make it more competitive to deal with how we capitalize the liability of such a large state with such potential. How do we mitigate risk, and how do we use technology and data to make sure we're getting it right? You're already setting an example with your work and dialog today to help all of us in California work together to solve our biggest problems. Thank you so much for coming together and for investing time and talent in helping move California forward."

Moderators of panels at the event included: Neal Driscoll, PhD, principal investigator, ALERTCalifornia, UC San Diego, professor of geology and geophysics at Scripps Institution of Oceanography; Nathan Hui, PhD, research engineer, ALERTCalifornia; İlkay Altıntaş, PhD, chief data science officer, San Diego Supercomputer Center, founding faculty fellow, Halıcıoğlu Data Science Institute, UC San Diego; and Ramesh Rao, PhD, director, QI / CalIT2, UC San Diego, who also provided introductory and closing remarks.

Visit QI's YouTube channel to watch video recordings of the plenary sessions.

The February 12 event at the Qualcomm Institute, UC San Diego was the first of three CalIT2 symposiums, to be followed by UC Riverside's "Quantum Infrastructure from Academia to Community" and another at UC Irvine later this year.

Share This:

• Share this story on Linkedin
• Share this story on Facebook
• Share this story on Threads
• Share this story on Twitter
• Share this story via email