Where the lightning density comes from

1. Three quantities, one loose name

Most of the confusion here is vocabulary, and it costs real accuracy because the three quantities differ from one another by factors that matter.

• N_T — total flash density. Every flash a detector sees, including cloud-to-cloud activity that never reaches the ground. This is what satellite instruments measure, because from orbit an intracloud flash and a ground flash both look like light.
• N_G — ground flash density. Flashes that reach the ground, per km² per year. This is the quantity most national maps and older textbooks publish, and the one the previous edition's method was built around.
• N_SG — ground strike point density. Distinct points struck on the ground, which is not the same as the number of ground flashes: one flash can have several strike points. This is what the current edition works from, and what a modern lightning location system reports.

The change is not cosmetic. The event rates in the assessment count how often something gets hit, and a flash with three strike points is three opportunities, not one. Taking a value labelled N_G off a map and using it where the method asks for N_SG is a quiet, systematic error running through every figure downstream.

2. The order of preference

Sources are not equivalent, and the standard is explicit that some are better than others. Work down this list and stop at the first one you can actually get.

1. A lightning location system covering your site. A ground-based detection network reports strike-point density directly, over a real observation period. IEC 62858 is the standard governing how that data is produced, and it is the reason the current edition can ask for N_SG at all — the quantity is only available because these networks exist. Many countries have one; in some it is a paid data product, and for a real project it is usually worth the fee.
2. A national or regional map published by a meteorological service or standards body. Usually N_G rather than N_SG, so it needs the conversion, but built from local observation over years and far better than a global average.
3. Satellite climatology. Global coverage, available anywhere, free — and the coarsest of the three. This is the fallback the standard's own annex describes, and it gives N_T, from which the ground quantities are derived by fixed conversion. Where no ground-based data exists, the standard takes the ground strike-point density as half the total flash rate.

3. What the satellite data actually gives you

The table below is computed by the same lookup the tool uses, over NASA LIS/OTD HRFC v2.3.2015 (0.5° satellite climatology) — the dataset the standard's own annex cites for this purpose. Values are per km² per year at the grid cell containing each city centre.

Two things are worth staring at. The range is enormous — roughly a thousandfold between the driest and wettest entries — so any habit of carrying a "typical" density between projects is indefensible. And the Gulf figures are low in a way that surprises people who work there: coastal desert genuinely produces very little lightning, while the interior produces noticeably more. Neither intuition nor a neighbouring country's value is a substitute for looking your own site up.

4. When not to trust the satellite value

This is the section that usually gets left out of tools offering a density lookup, and it is the one that matters, so here are the limitations of the data above stated plainly. They are properties of the dataset, not of any particular implementation of it.

• The cells are about 55 km across. Local terrain effects — a ridge, a lake, a coast — are averaged away entirely. A site on high ground in a mountainous region can see substantially more activity than its cell average, and nothing in the number tells you so.
• In low-flash regions the relative noise is worst. Look at the two UK entries above: they differ by roughly fivefold, and they are a couple of hundred kilometres apart in a country with little lightning anywhere. When the underlying count is small, sampling variation dominates. Treat a low value as "low", not as a precise figure.
• It is a climatology, not a measurement of your site. It describes a long-term average from an orbiting instrument, and it carries known sampling artefacts at high latitudes.
• Zero is ambiguous. In the source data, "no coverage" and "effectively no lightning" are stored identically. A returned zero should prompt a check, not a shrug.

None of this makes the satellite route useless — it is the standard's own fallback, and for a feasibility study or a site with no better data it is the honest best available. It makes it a suggestion, which is why our tool offers the lookup, shows its provenance, and never applies a value silently: the engineer signing the assessment should have chosen the density deliberately, and be able to say where it came from.

5. What to do on a real project

• Ask the client whether location-system data is available before anything else. On industrial, energy and telecom sites it often already is, because the operator buys it for other reasons.
• Check for a national map from the meteorological service or the local standards body. This is the commonest good answer and it is frequently free.
• Write down which source you used, and its date, in the assessment itself. This is the single input a reviewer is most likely to challenge, and "the tool suggested it" is not an answer you want to be giving eighteen months later.
• Where the value is marginal, test it. If a plausible alternative density changes the verdict, that is worth knowing and worth saying in the report — it is a much more useful finding than a single number presented with false confidence.