First comparisons of ball-lightning report website data with lightning-location-network data
Karl D. Stephan · Richard Sonnenfeld · Alexander G. Keul
Summary and citation · read the original at the source · none found
In one page
Ball lightning has a data problem, and Karl Stephan, Richard Sonnenfeld and Alexander Keul set out to fix it. Almost everything known about the luminous ball comes from people who happened to be looking — untrained witnesses, no instruments, no second channel. So in June 2020 the team put up a website that collects detailed, structured reports, and then did the thing that turns an anecdote into a measurement: they matched those reports against a professional lightning-location network. Earth Networks logs where and when every cloud-to-ground stroke lands, so a report can be checked against a machine record of the storm around it. Seven cases are worked through in the paper, and in every one the network data substantiates the witness. The physics that falls out is immediately useful — the ball does not necessarily form where the stroke hit the ground — and the witnesses keep reporting a behaviour theory rarely models: passing through a closed glass window.
Why it matters hereChapter 1 is about the evidence ladder, and this paper is a rung being built in public: a phenomenon that lived on testimony now has a second, independent, instrumented channel to be checked against. Chapter 9 treats the self-holding luminous ball as a real object to be engineered, and every constraint that survives this cross-check is a constraint a bench plasmoid has to reproduce.
What it claims
01The evidential starting point is stated plainly: nearly all data available for ball lightning comes from eyewitness accounts, from untrained observers who see the object fortuitously. That is the condition the paper exists to change, and naming it is what makes the rest of the work a measurement programme rather than a collection.Article highlights, first bullet; abstract, first two sentences
Published and peer-reviewed02The authors built the collection instrument themselves. In June 2020 they established a website designed to gather detailed reports of ball-lightning sightings from eyewitnesses — a structured questionnaire rather than free-form correspondence, so that individual reports can be compared with each other and with external records.Abstract, second sentence
Published and peer-reviewed03The method is a coincidence test against an independent machine record. Cases from the questionnaire were linked to an Earth Networks database of cloud-to-ground strokes by selecting strokes close to the incident both geographically and in time. The lightning-location network knows nothing about the witness and the witness knows nothing about the network, which is what makes the agreement worth something.Abstract, third and fourth sentences
Published and peer-reviewed04The result is a clean seven for seven. The paper presents seven individual cases, and the location-system data substantiates details of the ball-lightning reports in every case; all seven fit the previously established characteristics of ball lightning well. A small sample, but a first, and it is the first time website-collected reports have been reinforced with objective instrumented data.Article highlights, second and third bullets; abstract, fifth sentence
Published and peer-reviewed05One physical finding comes straight out of the cross-check: ball lightning does not necessarily form at the lightning ground-strike location. The ball can appear at a distance from where the stroke terminated, which constrains any model that requires the object to be born in the strike channel itself.Abstract, fifth sentence
Published and peer-reviewed06What to watch: the reports emphasise features that theoretical treatments do not always consider, the authors singling out the object’s ability to pass through closed glass windows, and the paper closes with recommendations for both theorists and observational scientists. Correlating reports with location-system data may lead to new insights into the nature of the phenomenon — the next measurement is simply more cases, and a report caught early enough for instruments to be pointed at a ball while it is still flying.Abstract, sixth and seventh sentences; article highlights, fourth bullet
What to watch
The way in
https://doi.org/10.1016/j.jastp.2022.105953LICENCE CHECKED. Published as Journal of Atmospheric and Solar-Terrestrial Physics, volume 240, article 105953, November 2022. The only licences Crossref carries for the DOI are Elsevier’s text-and-data-mining licence and its STM article-sharing policies, none of which is a licence to readers; OpenAlex and Unpaywall both report the record closed with no repository copy on 2026-09-08. No text of the paper is reproduced here. SOURCE READ. The ScienceDirect page returns 403 to an automated reader, and the authors’ own postings could not be reached from this machine — Karl D. Stephan of the Ingram School of Engineering, Texas State University, San Marcos, keeps his papers on a departmental page that did not resolve on 2026-09-08. This sheet is therefore written from the authors’ complete abstract together with the four published article highlights, both retrieved in full through the CoLab mirror of the Crossref record on 2026-09-08, and from the bibliographic record. Locators say which of the two a statement comes from. Richard Sonnenfeld works at New Mexico Institute of Mining and Technology and its Langmuir Laboratory for Atmospheric Research; Alexander G. Keul is at the University of Salzburg. When the full six pages can be read, this sheet should be rewritten from them.
How to cite it
Karl D. Stephan, Richard Sonnenfeld, Alexander G. Keul (2022) First comparisons of ball-lightning report website data with lightning-location-network data. doi:10.1016/j.jastp.2022.105953
Where it sits in the curriculum