How Lightning Detection Works

The lightning dots on a live map seem almost magical: a flash happens somewhere and, a few seconds later, it appears on your screen at the right spot. Behind that is a network of sensors listening for the radio signal every lightning flash emits. Knowing how that network works helps you understand what the dots can tell you, and what they can't.

Lightning as a radio transmitter

A lightning flash moves an enormous electric current in a fraction of a second, and that current broadcasts a burst of radio energy across a wide range of frequencies. If you've heard crackling static on an AM radio during a thunderstorm, you've heard it. Scientists call these bursts sferics, short for atmospherics. The strongest emissions are at very low and low frequencies, which can travel hundreds or even thousands of miles by bouncing between the ground and the upper atmosphere.

Locating a strike

A single sensor can tell that lightning happened, but not exactly where. Networks combine many sensors spread across a region, each with a precise clock synchronized by GPS. When a sferic arrives, each sensor records the time to within a millionth of a second or better. Because the signal travels at nearly the speed of light, it reaches a nearer sensor slightly sooner than a farther one. A central server compares the arrival times from several sensors and calculates the one location that fits them all. This is called the time-of-arrival method, and it's the same basic idea GPS uses in reverse.

Some systems also measure the direction the signal came from at each sensor, which adds accuracy. And because the waveform of a cloud-to-ground stroke looks different from an in-cloud discharge, many networks can estimate the type of flash and its peak current and polarity.

Who runs lightning networks

• Commercial networks, such as the National Lightning Detection Network used by the National Weather Service, airports, utilities, and insurers, use professionally installed and calibrated sensors. They typically detect well over 90 percent of cloud-to-ground flashes across the continental U.S., locating them to within a few hundred meters.
• Community networks, such as Blitzortung.org, which RadarBug uses, are built by volunteers who install low-cost receivers at their homes and share the data. Thousands of stations worldwide feed a central processing system. Where stations are dense, as in much of the U.S. and Europe, detection is good and locations are generally accurate to within about a kilometer, often better. Where stations are sparse, fewer strikes are detected and positions are less precise.
• Satellites. The Geostationary Lightning Mapper on NOAA's GOES weather satellites watches the Americas from orbit, detecting the flash of light from lightning, including in-cloud lightning, day and night. Its resolution is coarser (several miles), but it sees over oceans and remote areas where ground networks are thin.

Cloud-to-ground vs. in-cloud lightning

Most lightning never touches the ground. In-cloud and cloud-to-cloud flashes typically outnumber cloud-to-ground flashes several to one. The ratio matters for two reasons:

• In-cloud lightning is often an early sign. A rapid jump in total lightning, especially in-cloud flashes, often comes before a storm produces severe weather, because it reflects a strengthening updraft. Forecasters watch these "lightning jumps."
• Only cloud-to-ground strikes hit people. But a storm producing lots of in-cloud lightning can start producing ground strikes at any moment.

Ground-based networks are best at detecting cloud-to-ground strokes; how many in-cloud flashes they catch varies. On a community network map, it's best to treat every dot simply as "lightning occurred here" rather than trying to read too much into individual dots.

How accurate is the map?

Some limits to keep in mind:

• Location error. A dot may be off by a few hundred meters to a kilometer or more depending on sensor coverage. Use the dots to judge how close a storm is, not to pinpoint whether a particular tree was hit.
• Missed strikes. No network detects every flash. A gap in dots doesn't guarantee there's no lightning, especially in areas with few sensors or when a storm is just starting.
• Delay. Data reach the map seconds after the flash, which is fast enough for awareness but not for reacting to a single bolt. By the time you see a dot near you, the storm is already close enough to be dangerous.
• Display limits. RadarBug draws recent strikes inside the area of the map you're viewing. During very active weather, it caps how many strike markers it draws at once to keep the map responsive. Zoom in on your area during big outbreaks.

Using RadarBug's lightning layer well

The lightning layer and lightning alerts are most useful as an early, objective signal that lightning is getting close, so you can act before you hear thunder or when you can't hear it (indoors with headphones, at a noisy event, or near the water).

• Pick an alert radius that gives you time to act. Lightning can strike 10 miles or more from its parent storm. A radius of around 10 miles gives you warning while there's still time to get to shelter. A small radius may only alert you when a strike is already dangerously close.
• Set a home location to measure alerts from your house or a place you're responsible for, such as a ball field or a pool, rather than from your phone's position.
• Understand the digest. The first nearby strike raises an immediate alert. Further strikes are rolled into a "storm update" at the aggregation interval you choose until a quiet window passes, so you're not flooded with notifications.
• Use the 30-minute rule for the all-clear. A quiet map for a few minutes doesn't mean the storm is over. Wait 30 minutes after the last strike or thunder before resuming outdoor activities. See Lightning Safety Explained.

Lightning and radar together

Radar and lightning data complement each other. Radar shows where precipitation is and how intense it is (see How to Read a Weather Radar). Lightning shows which of those storms are electrically active. A large area of green rain on radar with no lightning is usually a steady rain; a small, intense cell with a cluster of lightning is a thunderstorm, even if it looks small. Lightning also often appears ahead of the heaviest rain, under the storm's spreading top, which is why strikes can hit where it isn't raining yet.