In 2003, Dr. Neal Driscoll and his family awoke to a billowing, pyrocumulonimbus cloud that stretched 20,000 to 30,000 feet overhead. Minutes later, seemingly impenetrable, black smoke encircled their home to the point that all visibility was decimated. What would later be called the Cedar Fire — a massive wildfire that burned over 270,000 acres — led to Driscoll’s young son being diagnosed with asthma while the rest of the family suffered with a smoke-related disease for 6 to 8 months after the event.
“At that time, we didn’t know where to go or what to do,” Driscoll said. “Back then, we had no actionable, real-time data — only a few cameras and network nodes used to communicate between firefighters and an emergency command center.”
The wildfire pushed Driscoll to his current calling as director of the ALERTCalifornia camera system led by the University of California at San Diego, where he is a professor.
ALERTCalifornia is a multi-hazard public safety platform used to understand natural disasters and determine short- and long-term impacts on people and the environment to prompt informed management decisions. The system integrates and expands on what was formerly the ALERTWildfire camera network and is composed of just over 1,000 pan-tilt-zoom, high-definition, near-infrared cameras. The ALERTCalifornia system is the third generation of UC San Diego’s wireless network and was made possible through a funding grant from the National Science Foundation.
From the initial onset of a wildfire, first responders, California residents and even the public can confirm ignition through the network using strategically placed sensors in various locations across the state.
Perhaps one of the best capabilities of the network is the value it adds to post-fire research efforts using the real-time data collected during fire events.
“After wildfires occur, we go back and study the data in search of prepositioning assets,” Driscoll said. “This includes analyzing burn scar severity and how the event impacted sediment erosion, debris flows and revegetation.”
This type of data mining is vital since the health of a forest plays a crucial role in the occurrence and severity of wildfires. In a healthy forest ecosystem, plants can naturally regulate their own processes, including managing fuel loads, maintaining proper moisture levels and reducing the risk of ignition. But when a forest is experiencing factors such as drought, insect infestations, disease outbreaks, and human activities like logging or improper land management practices, it weakens the forest's resilience and creates favorable conditions for wildfires.
“Using data compiled from the ALERTCalfornia camera network, we make models of the typography of vegetation and collect what’s called multispectral data,” Driscoll said. “That way, we can look at the health of the forest to characterize the carbon mass and moisture content of the fuels to help us manage and plan. If we can look at areas that are more stressed from extremities such as recurring heat waves and extreme drought using the thousands of scans from millions of dead trees in the forest, we can make critical decisions during an event.”
As the system continues to expand with new cameras being added each year, one aspect the ALERTCalifornia team and first responders often consider is what areas could most benefit from the cameras. This decision is made based on a plethora of factors related to sight value.
“After you reach 1,020 cameras, you want to make sure each new camera adds value,” Driscoll explained. “We look for mountaintops and try to find regions that increase our viewshed or provide really important linkages.”
Just this year, new cameras were added in Los Angeles, Santa Barbara and Madera counties.
“Sometimes we have potential sights where the viewshed coverage is superior but backhaul capability — the ability to bring information from other remote sensors that have a line of sight back to the incident command center — is also needed. So, there are essentially two criteria to meet: viewshed coverage and the ability to backhaul information to emergency command centers,” Driscoll added.
The system has undoubtedly had a noticeable impact on wildfire detection and evacuation times. During the Kincade Fire in 2019, over 180,000 people were evacuated without any loss of life, according to Driscoll. He believes this is largely due to the system’s ability to support collaborative life-saving measures.
“We’ve built a village,” he said. “It’s coordination between the counties, sheriffs, police, first responders, firefighters and professors here at UC San Diego. We all work together using the ALERTCalifornia network.”
The camera system has also helped improve bottleneck issues for firefighters attempting to reach wildfires quickly.
“There have been fire incidents where people were trying to escape the area and unconsciously blocked the exiting roadways, so firefighters were unable to access the impacted area,” Driscoll said. “This is why these coordination efforts, made possible by our sensors and communication network, are critically important.”
Yet, in the process of building and maintaining such an expansive camera network, there have been hindrances to overcome. Beyond the challenges posed by developing relationships and trust among stakeholders, Driscoll explained that permitting was another significant hurdle for the team to overcome. It's worth it, he said, when the work translates to potentially life-saving, quick responses.
“What I hear back from firefighters is that our biggest successes are the ones that you never hear about because of this technology,” Driscoll shared. “Five years ago, firefighters would need to send a battalion or rotary aircraft to confirm ignition. It could be 20 minutes, or it could be hours if the fire was tough to access. Now, within minutes they can have eyes on the problem area and scale their response to keep small fires from growing.”
The ALERTCalifornia system is more than cameras perched on mountaintops as tools for emergency personnel. The public can visit alertcalifornia.orgto visually access every camera in the network. Through the website, they can see where each camera is located via a digital map and click on their specific area for a 360-degree view.
“What they see can be an anomaly, smoke or even just a lightning strike, but they can pull up street maps and see exactly where threats are,” Driscoll expressed. “This empowers them to have knowledge about their situation and make decisions about whether to evacuate or not.”
Driscoll encourages residents to go online and train themselves on the ALERTCalifornia camera system before a wildfire or other natural disaster takes place.
“You don’t want to wait until there’s an order from the sheriff or fire department or police. You want to be prepared. These cameras offer another layer, another tool, to help increase your preparedness for events,” he said. “Our central goals in creating the ALERTCalifornia network is for us to prepare, manage, respond and recover so that we can empower first responders and the public to make decisions that will consequently save lives.”
