On June 9, the sky over Lanesville, Indiana turned green, then black. More than eight inches of rain fell in a few hours. Laura Lin watched her yard disappear under rushing water while she helped a neighbor get out. The area’s flash flood warning system sent the “get on your roof” alert too late. Water had already trapped people inside their homes. “All the people in town were already flooded when they started sending out alerts,” Lin said afterward.
That lag, the gap between a storm turning dangerous and your phone actually buzzing, is what a new flash flood warning system built by NASA, UC San Diego, and the National Weather Service aims to close. Starting this fall, it’s headed to every NWS forecast office in the country.
A Tool Built to Buy You Minutes
The system is called TACLS, short for Transient Artifact and Continuous Learning System. NASA’s Jet Propulsion Laboratory and UC San Diego’s Scripps Institution of Oceanography built it with NOAA’s National Weather Service. Here’s the plain version: GPS-style satellite signals slow down slightly in humid air. TACLS reads that tiny delay in real time. A machine learning model, trained on three decades of atmospheric data, flags conditions that match past flash floods, and a forecaster reviews the flag within about fifteen minutes.
In tests against major storms from 2017 to 2023, the system caught 93 percent of the flash flood warnings forecasters had actually issued, often sooner than the original alert went out.
“Without TACLS, a storm moving into your area could kill some folks. With TACLS, you can put out the warning and save some folks.” (Ivory Small, science and operations officer, NWS San Diego)
NWS offices in Los Angeles and San Diego have run TACLS operationally for months. Now, per recent reporting, the agency is rolling an upgraded version out to all 122 forecast offices nationwide, landing in the second half of October. A family in rural Kentucky or a retiree in coastal Texas will soon have the same early-warning tool already protecting Southern California.
Why a Flash Flood Warning System Saves More Than Time
Fifteen minutes doesn’t sound like much until you need it. It’s enough time to move a car to higher ground instead of totaling a $25,000 vehicle. It’s enough time to get an older neighbor, a toddler, or a person who uses a wheelchair out the door before the street becomes a river. It’s enough time for a volunteer fire crew to stage a rescue instead of improvising one.
The stakes go beyond any single storm. A Joint Economic Committee analysis puts flooding’s yearly cost to the US economy between $180 billion and $496 billion, a bill that shows up later in insurance premiums, disaster aid, and local taxes. Shaving even a few minutes off warning times, across 122 offices, adds up to real savings.
A few numbers worth remembering before the next storm:
- Six inches of fast-moving water can knock an adult off their feet.
- Twelve inches can float a car.
- Two feet can carry away a pickup truck.
Floods are the second-deadliest weather event in the US and the deadliest worldwide. That’s why NASA built TACLS as open-source software any weather agency can adapt. Our earlier coverage of Missouri’s flood monitoring app and Google’s hurricane forecasting AI shows the same pattern elsewhere: better data, delivered faster, gives ordinary people more say over their own safety.
Before storm season reaches your area, check that your phone has Wireless Emergency Alerts turned on. Talk with your household about where you’d go if a warning arrived at 2 a.m. TACLS can buy you minutes. What you do with them is still up to you.