US data [21] show that about one in five crashes occurs in bad weather or on slippery road surfaces: 21% of all crashes, 19% of injury crashes and 16% of fatal crashes. Relatively most of these crashes involve a wet road surface (about 75%) and rain (about 50%). Snow/ice (about 15%), melting snow (about 10%) and fog (about 5%) are less common. The percentages mentioned are based on absolute numbers of crashes, they say nothing about risk: certain weather conditions occur less frequently than others and/or result in less mobility, and this partly determines the share of crashes in that type of weather.
A Finnish study [5] did look at crash risk during bad weather, in particular crash risk during different types of precipitation and at different precipitation intensities. As Table 2 shows, crash risk during snow is more than twice as high and during sleet almost one and a half times as high as average (= a relative risk of 1). In rain, crash risk hardly increases and with no precipitation crash risk is slightly lower than average. Looking at precipitation intensity, regardless of precipitation type, crash risk was found to be more than twice as high for heavy precipitation, one and three-quarters times as high for showery precipitation and almost one and a half times as high as average for slight precipitation. Furthermore, crash risk in bad weather conditions was found to be relatively higher on motorways than on non-motorways, while in good weather conditions crash risk was lower on motorways. Also, crash risk in bad weather was higher for single-vehicle crashes than for multiple-vehicle crashes.
Table 2. Relative risk at different types and intensities of precipitation in Finland (based on over 10,000 crashes on 43 major roads in Finland in 2014-2016). Source: [5].
