Use case

Lightning data for electric utilities.

When a breaker trips, the useful question is whether lightning caused it. Lightning API answers that with historical strike queries along a corridor, cloud-to-ground and in-cloud classification, and enough rows per call to review a whole feeder without paging through results all afternoon.

Correlating a trip with a flash

Protection operates, a recloser cycles, and someone has to write down a cause. Lightning is the most common suspect and the easiest one to confirm or rule out, provided you can ask for strikes in a corridor over a specific window rather than eyeballing a regional map.

That is a bounding box query with a time window. Pass the corridor, pass the minutes around the event, and read back the flashes with their timestamps and classification.

Strikes along a corridor around an event

curl "https://api.lightningapi.dev/v1/flashes" \
  -H "X-API-Key: lapi_your_key_here" \
  -G -d since_minutes=60 \
     -d min_lat=39.62 -d max_lat=39.88 \
     -d min_lon=-105.21 -d max_lon=-104.76

Cloud-to-ground versus in-cloud

For asset damage the distinction matters. In-cloud flashes inflate a raw count without ever reaching a conductor, so a corridor that looks busy may have had very few strikes that could plausibly have caused a fault. Classification is on every flash, so you can filter to cloud-to-ground before you draw a conclusion.

How much you can pull in one call

Row caps are per call and set by plan. They matter here more than on most use cases, because a storm across a service territory generates a lot of flashes and paging through them is what makes an analysis tedious.

PlanRows per call
Pro20,000
Ultimate50,000
Supercell200,000

Live monitoring, not just post mortem

The same data is available as a push stream, so crews can be held or released on current conditions rather than on a page refresh. Storm cell data is a separate endpoint if you want the shape of the system rather than individual flashes.

Streaming is an Ultimate feature. If you only need post-event analysis, a lower plan covers it, but the row cap is still the thing to check first.

Recommended plan

Ultimate

Fault review pulls a wide time window across a long corridor, and Ultimate raises the per-call row cap to 50,000, which is the difference between one query per event and twenty paged requests per event.

Compare plans

Frequently asked questions

How far back does the history go?

The dataset grows continuously and is broken down by state on the live data page. Check there for current depth before you design a workflow around a specific retention window.

Can I tell cloud-to-ground from in-cloud?

Yes. Classification is present on each flash, so you can filter to cloud-to-ground before correlating with a fault.

Is there a bulk export?

Access is through the API rather than file drops. The per-call row cap on Ultimate and Supercell is sized so that a corridor query for a single event returns in one request.

Do I have to talk to sales?

No. Plans are self-serve with published pricing. Enterprise terms are negotiated, but you do not need a call to start evaluating.

What coordinate format do queries use?

Decimal degrees for latitude and longitude, passed as a bounding box. The docs page has the full parameter reference.

Related

Last checked 2026-09-20

Coverage areas

Lightning data provided as-is; not for safety-critical use. Commercial use is permitted on every current plan. Read the EULA →