I spotted two 80ish page USCG test reports on strobe effectiveness tonight. I only skimmed them, but thought they looked interesting. Links and abstracts below.

Suitability of Potential Alternatives to Pyrotechnic Distress Signals (2012)

Abstract : Purpose: To determine the potential suitability of electronic alternatives to pyrotechnic visual distress signals through the evaluation of the effectiveness of presently-available LED (and other) devices as visual distress signal devices (VDSDs). Methods: Requirements workshop, market research, field testing to assess visibility at different ranges, paired comparison testing to assess attention-getting characteristics, and ergonomic testing. Results: Lab test results predicting device visual detection range based on effective intensity compared well with results obtained from field testing. Light-emitting diode (LED) devices tested consistently better than incandescent or flashtube devices. Color and flash pattern (rapid flash rate or S-O-S characteristic) improved the perceptive performance of the devices. Conclusions: LED devices have potential as an alternative to pyrotechnic VDSDs. Desirable VDSD characteristics identified in this report can be used to inform future VDSD performance requirements development. Intensity profiles (omni-directional versus narrow beam) must be considered when comparing predicted visual detection ranges. Detection ranges predicted from laboratory-measured Effective Intensity of white VDSDs compared favorably with ranges observed during field tests. This indicates that Effective Intensity can be used in lieu of field tests to predict the visual detection range of VDSDs under specified meteorological conditions.

Alternatives to Pyrotechnic Distress Signals; Laboratory and Field Studies (2015)

Abstract : This report documents a multi-year project effort to develop a specification for a light-emitting diode (LED) signal characteristic as an alternative to pyrotechnic, maritime distress signal flares as visual distress signals. The report includes the methodology used in evaluating color, flash pattern, and intensity for an LED distress signal, conspicuous against certain lighting conditions, at six nautical miles, in 10 miles meteorological visibility. The effort included a literature review, measurement and quantifying different levels of background lighting to recreate their effect in a vision laboratory, a series of laboratory tests to determine signal conspicuity in a controlled environment, and a field test in the marine environment. The human-subject laboratory experiments determined relative LED signal conspicuity, based on subjects' accurate identification of a signal and the response time to make that identification. The lab results (conspicuous signal characteristics) were the basis for field testing. In addition to LED signals, the human-subject field test included two, commercial, off-the-shelf (COTS) signal devices and a handheld pyrotechnic flare in the signal evaluation. In four nights of testing, the experimenters produced a signal characteristic that was significantly more conspicuous than the flare and the two COTS devices.