The Toppenish investigation was an attempt to study recurring light reports through field observation and a continuing database. David W. Akers’s original papers preserve both observations he considered unusual and reasons to be cautious about the results. This case covers the 1972–1975 programme, rather than treating each night or statistical update as a separate mystery.
The initial field study
Akers’s first report, submitted to J. Allen Hynek on 2 November 1972, concerns fieldwork near Toppenish from 20–31 August. With tribal and agency cooperation, he established portable observing positions, including locations near fire lookouts. One early account describes two reddish-orange, apparently round lights below a ridge, flickering and changing their relative positions. These are attributed descriptions, not measured dimensions of solid objects.
The instrument results were limited. One camera supplied the recorded photographic data; distance prevented useful spectra, and the motion-picture viewfinder was too dim for effective aiming. No abnormal magnetic disturbance was recorded. Akers considered the activity worthy of further investigation but did not claim conclusive evidence of an extraterrestrial source. His paper also entertained vehicle-light explanations for some observations.
The 1973 follow-up
The second paper, submitted on 8 April 1974, reports fourteen selected observations from 1973, eight followed up by Akers. Its categories comprise nine nocturnal-light reports, three close encounters without interaction, and two daylight observations. Reports with too little information or strong indications of an ordinary cause were excluded, so fourteen is not a count of every report made locally.
Akers looked for patterns in time, direction and place. He found no consistent direction of travel, but described groupings of observation locations near Toppenish, Zillah and Granger, and along the Satus–White Swan area. He cautioned that the sample was small. He also considered whether summer concentrations reflected the greater number of observers outside, noting more winter reports after active collection began.
The important distinction is between a pattern in collected reports and an established pattern in an unidentified physical process. A collection shaped by selection, follow-up and observer availability cannot answer both questions simply by being entered into a computer.
What the four-year database contained
In the statistical update, written after the 1975 reporting period discussed in its conclusion, Akers gives fifty-five encoded reports since the study began in 1972. Forty-three were classified as nocturnal lights, three as daylight objects, six as close encounters and three as other reports. Nine included photographs. The total refers to reports, not fifty-five photographed objects or fifty-five independent instrument detections.
His APDF system encoded information such as date, time, approximate location, conditions, appearance and observer reaction on punched cards. Its strangeness and probability ratings were subjective scales. Missing information could be left out, but reports lacking a date or approximate location could not enter the computer file.
Akers hoped that patterns might eventually help place observers and instruments more effectively. He warned against strong conclusions from the small sample, recognised possible observer effects on seasonal peaks, and reported that activity had declined sharply after May 1975. The update records an investigative method and its limits, rather than a demonstrated prediction of future sightings.
A later attempt to test a natural explanation
Greg Long’s 1990 monograph, Examining the Earthlight Theory, used interviews and the report collection to examine proposed links between unusual lights and tectonic strain. It discusses Michael Persinger and John Derr’s interpretation of statistical relationships with earthquake activity. Long criticised assumptions about screening, the reliance on heterogeneous witness reports and the gap between a correlation and a physical mechanism.
This is a substantive later argument, not a contemporary measurement of every 1972–1975 observation. Long’s criticism does not establish an alternative extraordinary cause. Equally, the existence of a natural hypothesis should not be presented as a completed identification of the entire collection. The relevant discussion is in the book’s chapters on tectonic strain and its evaluation; the book also ranges beyond the early programme considered here.
Assessment
The outcome is Insufficient evidence. The original study obtained limited physical data, and the follow-ups assembled a small, selected set of reports. Those records support a history of observations and investigation; they do not establish one cause for the programme as a whole. Individual observations may have different explanations.
Location, access and image
The map pin is an approximate Toppenish city reference from the US Census Gazetteer, not an observing station or a reported light’s position. The work covered a much larger area of south-central Washington. Its historical title uses “Yakima”; the community and government are referred to today as the Yakama Nation.
The Yakama Nation wildlife programme directs recreation enquiries to Tribal Forestry. The forestry page includes an older closure notice, which cannot establish current permission to enter a historic lookout. Public access to those positions has not been verified. This case is therefore not enabled as an accessible public place. Historical research cooperation gives no continuing visitor permission.
The original report’s photographs and illustrations carry restrictions and are not reproduced here. The credited Mount Adams photograph is separately released into the public domain and serves only as regional context. It does not show a reported object or establish where a witness stood.
Sources and further reading
Akers, 1972 report, annotated edition; 1973 observations, submitted 1974; four-year statistical update; Long, 1990, especially printed pages 87–99; 2025 Census Gazetteer, Toppenish city, GEOID 5371960. Source and access review: 13 November 2026.
For a separate instrument case with a different proposed explanation, see Fort Yukon, January 1965. Its single radar episode is not evidence for the Toppenish collection.