NTSB Railroad Accident Report
National Transportation Safety Board 52
as previously noted, all the physical and documentary evidence shows conclusively that the
signal aspect was red.
The results of postaccident sight-distance tests demonstrated the difficulty in identifying
a red signal aspect at CP Topanga when viewed from the Chatsworth station. During that testing,
test participants on the station platform were able to identify green and flashing yellow signals
but were unable to reliably identify a red signal. The engineer who participated in the sight-
distance tests and who, by his own account, “knew where to look,” had to move his locomotive
almost 1,000 feet from the station before he had what he considered to be a view of the signal
adequate to positively identify the signal aspect when it was displaying red, or stop. At this
distance, some members of the observation group reported that they were still unable to reliably
identify the red aspect. These findings were not surprising given that the signal aspect is an 8
3/8-inch-diameter lighted disk almost a mile away being viewed in daylight.
Eyewitness reports of the signal indication at the time of the accident must be evaluated
in the context of the physical relationship between the signal and the station, the environmental
conditions, and the capabilities and known limitations of the human visual system. The visual
angle of the westbound signal light at CP Topanga when viewed from Chatsworth station is
essentially equal to the aperture of a single photoreceptor, which puts it near the limit of normal
visual function,
58
making it particularly difficult to identify signal color from that distance. An
observer with excellent vision who is familiar with the signal and the surrounding environment
may be able to see, and perhaps identify the color of, the signal light, but it is more likely that an
observer either will not be able to see a signal light at this distance or will misidentify its color.
As illustrated by the sight-distance testing results in this investigation, this is particularly true of
a red signal aspect because the human visual system is less sensitive to red light than green light.
Research on railroad signal detection has reported misperceptions of signal color at long
distances under certain conditions or by certain viewers.
59
However, that research was associated
with viewing distances of about 2,950 feet, which is a little more than half the distance from
Chatsworth station to CP Topanga. At longer distances, research showed that viewers were often
unable to accurately perceive signal color.
The human visual system has been found to be more sensitive to brightness than to color
when viewing very small stimuli.
60
That is, the stimulus may be above a person’s threshold for
perception of brightness but below the threshold for the perception of color, allowing a viewer to
58
Studies of the human retina estimate that at the center of the foveal region—which corresponds roughly to
the center of a person’s field of view—a single cone photoreceptor in a typical human eye integrates light through an
aperture approximately 2.5µm in diameter, equating to a visual angle of approximately 0.5 arcminutes or .008
degrees. (An arcminute is a unit of angular measurement equal to 1/60 of 1 degree.) The common visual acuity
reference of 20/20 is the ability to resolve a high contrast spatial pattern separated by 1 arcminute. For addition
information see (a) D. R. Williams. “Topography of the Foveal Cone Mosaic in the Living Human Eye,” Vision
Research 28 (1988): 433-454. (b) C. A. Curcio, K. R. Sloan, R. E. Kalina, and A. E. Hendrickson. “Human
Photoreceptor Topography,” Journal of Comparative Neurology 292 (1990): 497-523. (c) L. N. Thibos. “Formation
and Sampling of the Retinal Image,” in K. K. De Valois (Ed.) Seeing (San Diego, CA: Academic Press, 2000) pp.1-
49.
59
J. M. Wood, D. A. Atchison, and A. Chaparro. “When Red Lights Look Yellow,” Investigative
Ophthalmology and Visual Science. 46(11) (2005): 4348-4352.
60
P. E. King-Smith and D. Carden. “Luminance and Opponent-Color Contributions to Visual Detection and
Adaptation and to Temporal and Spatial Integration.” Journal of the Optical Society of America. 1976;66:709–717.