How accurate and biased is slope perception of snow-covered mountains?

Estimating the slant of a snow-covered mountain slope is a difficult task. Human perception is tailored for accurate perception of cardinal orientations but is less precise for oblique angles. Indeed, previous studies found that humans overestimate the steepness of slopes (Proffitt et al., 1995, 2001). Accurate estimation of the slant of snow-covered mountains is key to identify avalanche release zones, i.e. snow-covered slopes between 30 and 45 degrees that can be start zones for avalanches. We here investigated the accuracy of slope estimation among backcountry skiers varying in experience (study 1, Kattfjordeidet study) and whether training and feedback by using clinopoles can improve accuracy (study 2). In study 1, participants estimated the slant as line-of-sight upwards or sideways from a parking lot. We found that most backcountry skiers overestimated the slant, but the overestimation was less pronounced the more experience the skier has. The overestimation was strongest for shallower slopes. In study 2 we equipped backcountry skiers with clinopoles and asked them to estimate and thereafter to measure slants during their trip, either as line-of-sight upwards, sidewards, downwards, or as contact. Here, we did not find overestimation for shallow slopes but slight underestimation for slopes above 30 degrees. Participants’ estimates improved with repeated practice, particularly for contact measurements but less so for line-of-sight measures. Confidence in one’s own estimate did not change with practice and was rather idiosyncratic. Our data replicates the classic overestimation bias when skiers use a static estimation method and line-of-sight upwards estimation. This bias was most pronounced for shallower slopes (below 30 degrees) and reduced (study 1) or reversed (study 2) for slopes over 30 degrees. We discuss implications for avalanche safety and education.