Effects of weather and weather forecasts on driver behaviour

Winter-time slipperiness is a considerable source of elevated road accident risk, especially in northern countries such as Canada, Finland and Sweden. The national road administrations often offer a service to inform drivers of forthcoming weather and driving conditions in different regions. This study addressed the effects of adverse weather and traffic weather forecasts on driver behaviour in Finland. Drivers (n = 1437) answered a questionnaire on perceptions of weather, self-reported driving behaviour, pre-trip acquisition of weather information, and possible travel plan changes. The questionnaires were distributed and instantaneously collected in rural service stations in different weather and driving conditions. Data from traffic weather forecasts, automatic traffic counters and weather measurement stations concerning the same area (and road) were also collected. Acquisition of weather information for the trip was associated with low recent driving experience, increasing age, female gender, long trip in question and very poor (local) conditions perceived by the driver. Drivers who had acquired information had also made more changes to travel plans, but information acquisition did not have an effect on their on-road driving behaviour. However, they estimated prevailing risks higher than those who did not acquire weather information. Drivers generally considered the driving conditions better than the forecast, but significantly less so in darkness than in daylight or civil twilight. Leisure trips were clearly underrepresented during very poor driving conditions forecasts, suggesting that some trips are postponed as a result of adverse weather conditions or forecasts thereof. Drivers reported various kinds of compensatory behaviour during adverse conditions, including a 6–7 km/h target speed decrement. This corresponded to traffic flow speed measurements. The results suggest that the on-road driving behaviour is predominantly affected by the prevailing observable conditions, rather than traffic weather forecasts. It is suggested that if administrators wish to help drivers in adjusting their on-road behaviour in adverse weather conditions, the methods should be more local and technical by nature.