Lightning as a Local Observation, Not a Rain Gauge

Lightning on a personal weather station (PWS) is an observation of electrical events, not a measurement of rainfall and not an official warning. The sensor may be an optical detector on the roof, a radio strike mapper, or a third-party network feed written into the same logger file as temperature. Those are three different instruments. A rain gauge is a fourth. TNET Weather is not a warning service. If thunder is in the area, follow NWS lightning safety and local emergency authorities—not this page and not a strike counter.

This URL is about lightning as a local observation. Combining lightning ticks with satellite or radar imagery is a different problem, treated at lightning.php.

Historical context

/weather-lightning.php lived in the observations cluster of the historical TNET Weather site, among Doppler, forecast, and “live” dashboard URLs. Recovered markup mixed lightning headings with GRLevel3 radar chrome—the same template family as other Mesa-era image pages. That mashup is not this article’s subject. The useful remainder is the station operator’s question: what is the lightning number next to today’s rain total, and what is it not?

Hobby stations of the 2000s added lightning because visitors asked, because monsoon afternoons in Greater Phoenix produce frequent storms, and because consumer optical and radio accessories had become cheap enough to hang beside a Davis or Oregon Scientific array. The original sensor model, feed vendor, and live map are not rehosted. This page does not display strikes.

Three ways a PWS “sees” lightning

Treat the source as part of the observation. A count without a source is an unlabeled derived badge.

Optical sensor

An optical lightning detector looks for the light pulse of a flash—typically a photodiode or similar receiver with filtering against sunlight and man-made flicker. It reports events at the instrument. Range, if the manufacturer states one, is a property of that device and its sky view, not a mapped strike location. Nearby security lights, welders, and corona on power gear are known nuisance sources. An optical count is not a cloud-to-ground pinpoint, and it is not rainfall.

Radio strike mapper

A radio-frequency (RF) detector listens for the electromagnetic pulse of a discharge (sferics). Direction-finding or time-of-arrival designs can estimate a bearing and, with a network or a second station, a location. A single backyard RF board is still a local instrument: its range, noise floor, and discrimination of in-cloud versus cloud-to-ground flashes are hardware and siting problems (metal roofs, amateur radio, switching power supplies). Plotting every trigger as a street-level strike is a misuse of the observation.

Network feed

Many PWS pages never had a sensor. They displayed a third-party lightning network—commercial or volunteer—through an image, XML, or JSON include. That is someone else’s observed catalog, ingested at a poll interval, not a station measurement. The feed has its own detection efficiency, latency, and license. Hotlinking it under a TNET or station logo does not make it a rooftop observation. If the historical site used a network and the contract is unclear, do not restore the tiles.

A satellite optical mapper (for example NOAA’s Geostationary Lightning Mapper) is a fourth family: observed lightning from orbit, at satellite resolution, not a PWS accessory. It belongs in an imagery discussion, not in the rain-gauge column. See the Cumulus satelight note for that split; do not collapse GLM into this URL.

Why it is not a rain gauge

A rain gauge measures liquid (or melted) precipitation depth at a point, in millimetres or inches, over a stated interval. Its error modes are wetting, evaporation, wind undercatch, splash, and clogging.

A lightning observation measures electrical discharges the instrument or network was designed to detect. Its error modes are optical clutter, RF noise, missed in-cloud flashes, false bearing, and feed outages.

Lightning and rain often occur in the same storm. They are not interchangeable quantities. You cannot convert a strike count into a rain total, and you cannot treat a dry gauge as proof that lightning did not occur (virga, anvil crawlers, and storms that have not yet reached the funnel all exist). Putting both numbers on one “storm intensity” bar without documenting the derivation produces a derived index pretending to be two observations.

Label them separately:

Rain: gauge type, height, interval, units, time.
Lightning: source class (optical / RF / network), what is counted (flashes, strokes, or “events”), time system, and any stated range. If cloud-versus-ground is unknown, say so.

What to put on the observation, not on a warning badge

A defensible lightning line on a PWS looks like a citation:

Optical detector at the station, 18:12–18:27 UTC, 14 events. Not a mapped strike location. Not an NWS warning.

Or:

Network feed X, last poll 18:27 UTC, 3 cloud-to-ground detections inside a 20 km radius of the station. Source is the network, not this logger. Official hazards: weather.gov.

Do not:

Color the whole county red because the optical sensor clicked.
Animate stale strikes on a current webcam and call it live.
Use lightning rate as a tornado, hail, or flash-flood proxy. Those are other observation and warning families.
Hide the clock. A count without a time is not an observation.

Safety: the 30-30 rule, conceptually

Lightning is a life-safety hazard. NWS’s current public rule is direct: if you hear thunder, lightning is close enough to strike you; move immediately to a substantial building with wiring and plumbing, or to an enclosed metal-topped vehicle; stay there at least 30 minutes after the last thunder (NWS lightning tips; last checked 13 August 2026). No place outside is safe when thunderstorms are in the area.

The older 30-30 rule is a counting aid, not a replacement for that guidance and not a TNET product. Conceptually:

The first 30: if the time from flash to thunder is 30 seconds or less, the storm is close enough that you should already be moving to shelter (at roughly 1 mile per 5 seconds for sound, 30 seconds is about 6 miles; NWS notes lightning can strike on the order of 10 miles from the parent storm).
The second 30: wait at least 30 minutes after the last thunder before resuming outdoor activity.

NWS has also stated that the flash-to-thunder count is for good visibility with a clear view of the storm. If you hear thunder at all, you are already in the risk area. A PWS strike map that still shows “0” does not overrule thunder in your ears. A PWS strike map that shows many events does not create an NWS warning.

TNET does not issue lightning alerts. A local observation can tell you that this instrument detected events at this time. It cannot tell a family when to leave a pool. That decision follows NWS and local emergency authorities.

How a station page should treat NWS watches, warnings, and advisories—link them, do not paraphrase them—is hazards.php.

Distinct from the imagery URL

/lightning.php is the historical imagery mashup: satellite and radar pictures with lightning-adjacent navigation. This page is the in-situ and feed observation. Do not copy captions between them. A GRLevel3 reflectivity frame is a radar product. An optical click is a local event. Drawing them in one browser window does not make them one measurement.

Modern relevance

Convective structure and electrical activity are atmospheric evidence. They remain observed only when source, time, and instrument class travel with the number. TNET’s public grouping of weather inputs among other evidence families is the evidence families behind a connection outlook. That page does not ingest a PWS lightning counter as an official network, and it does not treat lightning as a promised change in connectivity.

This URL does not display live lightning.

Sources

NWS lightning safety (last checked 13 August 2026)
NWS lightning tips (last checked 13 August 2026)
NWS Shreveport: 30/30 lightning safety rule (last checked 13 August 2026)
weather.gov
TNET: evidence families
TNET observations hub