In fog, it is challenging to detect approaching vehicles, which may lead to risky pedestrian behavior. Previous studies have reported that in comparison to clear weather, fog produced greater distance and time-to-collision (TTC) estimates. However, these studies did not control the detectability of targets in fog. Thus, this effect could have been caused by contrast reduction of a visible target or by belated coming into view of an initially invisible target.
In two TTC experiments, we sought to disentangle whether contrast reduction or late visibility drive the overestimation of TTC in fog. Using a VR prediction-motion paradigm, participants estimated TTCs for vehicles approaching them at different constant velocities in clear and foggy weather. The vehicles remained visible until occlusion. Experiment 1 (n=20) revealed no significant main effect of the weather condition. In Experiment 2 (n=19), we additionally presented the vehicles in clear weather but such that all vehicles only came into view at distances of 41 m or less. Again, fog did not significantly increase TTC estimates, but participants estimated the TTC to be significantly longer when the visible portion of the car’s trajectory was shortened.
Our findings suggest that observers benefit from a larger spatio-temporal window when judging the TTC of an approaching object. The detrimental effect of fog may be reduced to this effect. Further research is needed to identify the interplay between contrast reduction and the spatio-temporal window.