Which weather stations send storm radar alerts to your hub?

A smart home weather station for storm radar alerts has to answer a less glamorous question than “how far away can it detect lightning?” The useful question is: when the sky goes bad, does the alert leave the station, app, or display and arrive somewhere your house can act on it?

That means separating two alert layers that often get blended together in product copy. One layer is local lightning detection: a sensor near your house hears the electromagnetic signature of a strike and reports distance or strike status. The other layer is cloud weather information: radar maps, vendor app alerts, and National Weather Service watches or warnings. A backyard station is not scanning Doppler radar from your roof. If radar or NWS warnings matter, that layer is coming from a cloud service or public alert feed, not from the pole-mounted sensor.

The compatibility table that matters before you buy

Status checked for Q3 2026. “Hub-visible” here means there is a practical path for the alert or alert state to show up outside the vendor’s own screen or app. Vendor behavior below is treated as documented, not hands-on confirmed.

The protocol reality check

If you came in hoping to buy a storm station the way you buy a motion sensor — Matter badge, Thread radio, add to Apple Home, done — the verified options do not line up that way. Eve Weather is a real Matter-over-Thread weather device, but its job is outdoor temperature, humidity, and barometric pressure. It is not a lightning detector and not a storm radar alert device [9].

So the protocol question shifts. Instead of asking whether the station speaks Zigbee, Z-Wave, Thread, or Matter, ask where the alert state becomes visible. WeatherFlow becomes interesting because Home Assistant can see local WeatherFlow data. Ambient becomes interesting because AWN can send cloud alerts and support account-based routes. AcuRite becomes less interesting for automations because the alert lives on the console in the base setup.

That distinction is also why severe-weather automations deserve a different mental model than comfort monitoring. If you want the broader architecture problem, NestGrid’s guide to smart weather monitoring in extreme events covers the same split between local sensors and external alert feeds.

WeatherFlow Tempest: the cleanest local path into Home Assistant

WeatherFlow is the setup that best matches the “show me the alert inside my hub” test. The Tempest system’s lightning detection is local to the station, and WeatherFlow documents a detection range up to 40 km [1]. More important for smart-home use, Home Assistant’s WeatherFlow integration uses the local UDP data broadcast from the WeatherFlow hub, so the station data does not have to travel through a vendor cloud before Home Assistant can see it [2].

That gives you a verifiable installation path. After pairing the station and adding the integration, you can look inside Home Assistant for the WeatherFlow entities and build an automation from what is actually present. The practical advantage is not philosophical purity about local control. It is that you can test the route without waiting for a vendor push notification to prove it exists.

For the second layer — watches and warnings — Home Assistant is where the WeatherFlow path becomes stronger than the product box suggests. The NWS integration is the official Home Assistant route for National Weather Service data, and the community nws_alerts project is another commonly used Home Assistant path for NWS alert entities and automation work [3][4]. That keeps the two jobs separate: Tempest supplies the local lightning signal, and the NWS feed supplies the broader public warning layer.

WeatherFlow’s own lightning behavior still matters. Vendor documentation describes validation/confirmation logic around lightning detection and notes that the Tempest app avoids repeatedly pushing every strike; push behavior is limited around the first strike in a period or a closer strike event, rather than behaving like a raw strike-by-strike siren [1]. Tempest’s public explanation of lightning detection also frames false-strike filtering as part of the detector’s signal-processing problem, which is exactly the kind of detail that should be tested in your own environment before tying it to loud alarms or shutdown routines [10].

A sensible Home Assistant automation would treat lightning as one input, not the whole emergency system. For example, a hypothetical automation could flash indoor lights when a nearby strike appears, pause a robot mower, and then use an NWS warning entity to decide whether to send a stronger household notification. The exact entity names and thresholds should come from your own installed integration, not from a copied automation that assumes every installation exposes identical data.

Ambient Weather: capable, but the cloud is part of the route

Ambient is the more app-and-cloud-shaped answer. The WS-2000 is a serious station platform, and the Ambient Weather Network is where the radar and alert story lives. AWN’s free tier lists map layers for Radar, Temperature, Wind, and Lightning, along with email and push alerts; SMS is reserved for AWN+ [5]. The WS-2000 product page also makes the lightning piece an accessory story: lightning detection requires the optional WH31L sensor [6].

That can be perfectly acceptable if your requirement is “tell me on my phone and let me wire cloud services around it.” It is weaker if your requirement is “make the strike state locally visible in my hub without depending on a vendor account.” The Ambient path is not bad because it uses the cloud. It becomes a problem only if the buyer thought “smart home compatible” meant the station would behave like a local sensor on the LAN.

There is also a difference between Ambient as a weather station and Ambient as an Alexa automation endpoint. Wirecutter’s 2026 weather-station guide treated the WS-2000 as a strong accuracy pick, but the Alexa-skill concerns in that coverage are based on user-review corroboration around Alexa+ changes, not a lab-confirmed finding that every Ambient Alexa setup is broken [11]. That is still enough reason to test the exact route you plan to use before making Alexa the only way a storm alert reaches the house.

For Google Home and Alexa households, Ambient may be the more natural fit than WeatherFlow if you want vendor app alerts first and smart-speaker routines second. For Home Assistant households, it is the less satisfying route unless you are deliberately comfortable with a cloud-mediated chain.

AcuRite Atlas: a real lightning alarm that mostly stops at the console

AcuRite Atlas is where the word “alert” can mislead a smart-home buyer. The product page for the Atlas package with lightning detection lists a detection range of 1 to 25 miles and describes audible lightning alerts from the HD display [7]. That may be all some households need: a display in the kitchen beeps, someone brings the patio cushions in, and the system has done its job.

But that is not the same as a hub route. In the base materials used here, the alert lands on the display console rather than in Apple Home, Home Assistant, Alexa, or Google Home. If you were planning to use the alert to close shades, announce through speakers, stop irrigation, or flip lights, the console-only path is the limitation to notice before the box arrives.

Ecowitt WH57: local detection, gateway dependency, cloud email

Ecowitt’s WH57 is easy to misread because the lightning sensor itself sounds self-contained. Ecowitt lists detection up to 25 miles / 40 km and a 79-second transmission interval [8]. The alert path, though, depends on what receives that data. For cloud email alerts, Ecowitt says the WH57 must be used with a compatible gateway such as the GW1100/GW1200 class [8].

That makes Ecowitt a split system: the sensor is the local detector, but the buyer still has to provide the gateway path if they want remote notification. Ecowitt also publishes lithium-battery guidance in connection with false-positive behavior, which is a reminder that lightning detectors are not magic boxes; power choice, placement, and electrical noise can affect whether the system is trustworthy enough for automation [8].

Ecowitt may be a good fit for a tinkerer who already understands the gateway ecosystem and only needs email-style notification. It is not the clean answer for a buyer asking, “Will this show up as a storm alert in my hub the way a leak sensor does?”

Do not confuse comfort weather with storm alerting

Eve Weather deserves a quick mention because it is exactly the sort of product that appears to solve the protocol problem. It supports Matter over Thread, which is what Apple Home, Google Home, Alexa, and SmartThings users increasingly want to see. But the device measures temperature, humidity, and barometric pressure. Its product materials do not make it a lightning station, radar-warning device, or NWS alert bridge [9].

That does not make Eve Weather irrelevant. It is a clean outdoor comfort sensor. It can help with HVAC context, garden automations, or “is it getting humid outside?” routines. It simply belongs in a different purchase category than storm alerting.

The same caution applies to safety wording. Consumer home-weather guides rightly distinguish condition-based station alerts from emergency alerting; these systems can support household automation, but they should not replace official emergency alert channels [12]. If you want to test severe-weather devices as a system rather than as product claims, NestGrid’s severe-weather smart-home testing guide is the more appropriate next stop.

Which one fits your ecosystem?

For Home Assistant, WeatherFlow Tempest is the clearest fit. It has the most direct, status-verifiable route for local station data because Home Assistant can consume WeatherFlow’s local UDP broadcast [2]. Pair that with an NWS alert feed inside Home Assistant, and the two storm layers can be handled where your automations already live [3][4].

For Apple Home, the honest answer is less tidy. None of the verified storm-station options here behaves like a native Matter lightning-and-radar station. If you already bridge Home Assistant into Apple Home, WeatherFlow can still make sense as part of that larger architecture. If you want a direct Apple Home accessory with storm alerts, Eve Weather is not it; it is comfort weather only [9].

For Alexa or Google Home, Ambient is the easier cloud-first candidate, especially if you like the AWN app and accept push/email alerts as part of the route. Just verify the exact voice-assistant or automation route you intend to rely on. A marketing phrase like “works with Alexa” is not the same as “my lightning or radar alert reliably appears as a trigger I can automate.”

For a household that only needs a local audible alarm, AcuRite Atlas may be enough. A console beep is not elegant, but it is at least clear where the alert goes. For an Ecowitt buyer, the pre-purchase checkpoint is the gateway: without the compatible receiver path, the WH57’s detection claim does not automatically become a remote alert path.

Buy by alert path first, then by detection range. A 25-mile or 40-kilometer claim is useful only after you know where the warning lands. If it lands in Home Assistant, you can automate from it. If it lands in AWN, you are buying into a cloud route. If it lands on a console, someone has to hear the console. And if the plan depends on Matter or Thread solving storm alerts neatly in 2026, save the money until the product category catches up.

References

1. Lightning Detection, WeatherFlow Support.
2. WeatherFlow, Home Assistant.
3. National Weather Service (NWS), Home Assistant.
4. nws_alerts, GitHub.
5. Ambient Weather Network, Ambient Weather.
6. WS-2000 Smart Weather Station, Ambient Weather.
7. AcuRite Atlas Weather Station with Gray HD Display and Lightning Detection, AcuRite.
8. WH57, Ecowitt.
9. Eve Weather, Eve.
10. How Does A Lightning Detector Work?, Tempest.
11. The Best Home Weather Stations, Wirecutter.
12. Best Home Weather Stations, The Spruce.