Daily news explainers: How storm alerts reach you

What to take away

Daily news explainers should distinguish who issues a warning from how it reaches people.

The alerting chain from forecast to phone

A storm alert begins with a forecaster, not a phone company. The National Weather Service office responsible for a county issues a warning product, and software pushes it into FEMA's Integrated Public Alert and Warning System. From there the message fans out to phone networks, broadcasters and other paths.

State and county emergency managers sit alongside the Weather Service in that chain. They can originate their own alerts through the same federal pipeline when a hazard is not weather related, such as a chemical release or a wildfire evacuation. One storm can produce messages from two or three originating agencies.

The chain breaks at the last mile. Cell coverage gaps, muted phones, sleeping residents and old receivers all interrupt delivery. That is why emergency managers treat no single channel as sufficient, and why a reporter should name the channel that carried a message.

Timing decides what a warning is worth. A warning issued with lead time lets people act. A warning issued after the fact is documentation. When you read a weather alert, check the issue time against the event time before judging whether the system worked.

Who holds IPAWS alerting authority

IPAWS consolidates federal, state, territorial, tribal and local alerting into one pipeline. FEMA runs the technical platform and the application process. The originating agency owns the message and the decision.

Who holds that authority varies by state. A state emergency management agency almost always has it. Many county emergency management offices do too. Some large cities, and in a few states campus or utility operators, have obtained access for specific hazards. A police chief cannot log in and alert a metro area.

That structure explains a common complaint after disasters. The federal system worked, but nobody with authority pulled the trigger, or the message was drafted too slowly. Reviewing who held authority and when they acted is the core accountability question in any alert postmortem.

IPAWS supports several message types, including presidential alerts, imminent threat alerts and public safety messages. Storm alerts usually use the imminent threat category, which requires a hazard posing an immediate risk to life or property. For the reporting standards involved, see this weather reporting guide.

Weather Service warning products and their triggers

Weather Service products are the weather-specific layer. Forecasters at local offices write them against defined criteria, which is why the vocabulary travels.

A watch means conditions are favorable for a hazard within a defined area and time window. A warning means the hazard is occurring, imminent or highly likely. An advisory means a nuisance hazard is expected that is inconvenient but not life threatening. Those three tiers are the grammar of storm coverage.

Triggers are specific. Severe thunderstorm warnings generally require hail of a set size, wind gusts above a threshold, or both. Tornado warnings rest on radar signatures, spotter reports or both. Flash flood warnings rest on rainfall rates, gauge readings and soil conditions. Heat, winter storm and wind chill products carry their own numeric criteria.

Each product also carries quotable structure: a headline, a hazard description, an impact statement, timing and a list of affected counties. The impact statement is often the most useful part for readers, because it translates meteorology into consequences such as downed trees or flooded underpasses.

Because the criteria are public, a warning is falsifiable. If a storm never met the threshold, that is a legitimate story. If it exceeded the threshold and no warning came, that is a bigger one. A worked example of that analysis is available in this daily news explainer.

Wireless Emergency Alerts, EAS and NOAA Weather Radio compared

Wireless Emergency Alerts are the short messages that appear on a phone screen with a distinctive tone and vibration. The Federal Communications Commission sets the rules for message length, alert categories and which alerts a user can turn off. Presidential alerts cannot be turned off.

The Weather Service explains how wireless emergency alerts reach devices: participating carriers broadcast to all compatible phones in a geographic area, so delivery does not depend on a phone number or an app. That is how the alert reaches a visitor whose number is from another state.

The Emergency Alert System is the older broadcast path. It interrupts radio, television and cable programming with audio and text, reaching people who are not looking at a phone. Broadcasters must still carry required tests and alerts.

NOAA Weather Radio is a dedicated network of transmitters carrying continuous weather information and tone-activated warnings. A receiver with the right feature stays silent until a warning for its county triggers it, which helps overnight and in areas with poor cell service.

Every channel has a blind spot. WEA misses people without compatible phones or with alerts disabled. EAS misses people who are not watching or listening. Weather radio misses anyone without a receiver. Opt-in systems miss anyone who never signed up. For a side-by-side look at how these products differ, see this health alert reporting guide.

State and county alert tools

County mass notification services send text, email, voice call or app push to residents who register. Registration is voluntary, so coverage is never complete, and officials know it.

These systems carry local granularity. A county can warn one neighborhood about a water main break, a road closure or an evacuation zone that does not match a Weather Service polygon. Weather warnings are drawn by meteorology. Local alerts are drawn by street.

State emergency management agencies often run their own layers on top. Some maintain a statewide opt-in system, some push alerts through a public safety app, and some coordinate with counties so residents do not register in five places. Practice varies widely, which is why a national story about alerts can mislead local readers.

Large jurisdictions add sirens, highway message signs, social media accounts and reverse 911 calls to landlines. Each has a distinct reach and a distinct failure mode. Sirens are outdoor-only and are not designed to wake people indoors.

For desk editors, the question is which local system a county uses and whether it is opt-in or automatic. That detail changes what a reader should do after the storm. This daily news explainer problems covers what to confirm before publishing.

What each trigger means for residents

A watch is a prompt to check supplies and plans. A warning is a prompt to act now, whether that means sheltering, evacuating or moving to higher ground, depending on the hazard.

Wireless Emergency Alert messages are deliberately short, so they cannot carry full instructions. Treat the phone alert as a signal to seek more detail from a weather radio, a local broadcast or a county feed.

An evacuation order is different from a warning. A local official issues it, usually through county tools and IPAWS, and it carries legal and practical weight. A warning describes a hazard. An order tells you to leave. Confusing the two is the most consequential error a reader can make.

Silence is not safety. If a phone did not buzz, the alert may still have gone out through EAS or weather radio, or the phone may have been outside the target polygon. Absence of a WEA message is weak evidence that no warning existed.

Here is how to read the common triggers:

A checklist for a storm alert day:

Reading an alert story without conflating systems

Most errors in alert coverage come from treating four systems as one. A headline saying the government failed to warn readers may be describing a county opt-in system, a Weather Service product or a federal pipeline, and the accountability differs in each case.

Start by identifying the originating agency. Then the channel. Then the trigger. A story that names all three is almost always accurate. A story that names none is usually vague in a way that hides the real failure.

Be careful with activation language. Saying IPAWS was activated implies federal action, but the message may have come from a county emergency manager using federal plumbing. Attribute the decision to the agency that made it, not to the platform.

Avoid implying that a phone alert is the only warning. Broadcast, weather radio and local systems may have carried the same message earlier or later. A reader who learns that one channel failed may wrongly conclude that no warning was issued.

Corrections matter more on this beat than on most. If an early story misstates which system fired, the correction should name the system explicitly rather than saying the alert was sent.

The Weather-Ready Nation initiative is a useful frame for this coverage because it treats public readiness, not just forecasting, as the goal. Alerting is the handoff between the two, and that handoff is where most storm-day confusion lives.

Common questions

Who actually sends a tornado warning?

The local National Weather Service forecast office issues the warning product. It may reach phones through Wireless Emergency Alerts and broadcast through EAS at the same time.

Can a county send its own wireless emergency alert?

Yes, if the county has alerting authority through FEMA's Integrated Public Alert and Warning System. Many counties do, but the authority is granted, not automatic.

Why did my phone not get an alert during a storm?

Common reasons include being outside the target area, having alerts disabled, an incompatible device, or coverage gaps. Another channel may still have carried the warning.

What should editors verify before publishing an alert story?

Confirm the originating agency, the channel used and the exact trigger language. Name all three, and correct any early report that conflated them.