Winter hoar frost conditions on Swedish roads in a warming climate

Abstract

As one of the most common reasons for slippery roads in wintertime, hoar frost can reduce surface friction and affect traffic safety. The risk of winter road hoar frost is subjected to changes in the warming climate. A better understanding of these changes could lead to improved forecasting of hoar frost risk and provide information to policymakers in making climate adaptation strategies. In this work, the decadal variation in winter road hoar frost risk between 2000 and 2016 in Sweden was examined by using in situ observations from 244 stations in the Swedish Road Weather Information System. Results show that hoar frost risks have decreased in the south of Sweden (south of 59°N), whilst increasing in central Sweden (approximately 59°–65°N). Hoar frost risk tends to increase (decrease) in regions where there is a relatively high (low) mean number of hoar frost risk days. Further analysis indicates that the strengthened winter North Atlantic Oscillation (NAO) over the last few decades, which resulted in warmer and wetter winters in Sweden, is the main cause of the changes. During strong positive NAO winters, the frequency of blocking anticyclones centred to the south‐west of Sweden significantly decreased and led to more warm and moist air from south‐west being transported to Sweden. The reduction in hoar frost risk in the southern part of Sweden is mainly due to an increase in road surface temperature, while the increasing hoar frost risk in central Sweden is dominated by the increase in relative humidity, which favours the occurrence of hoar frost.

On the relationship between the risk of hoar frost on roads and a changing climate in Sweden

“…Hu et al . () showed that the volume of water vapour transported from the southwest increased in a warming climate, which may result in more occurrences of warm air advection. However, based on the above analysis, some occurrences of warm air advection may not lead to an increase in temperature.…”

“…RWIS does not record the occurrence of hoar frost directly, so the above variables were used to infer an occurrence of hoar frost. For the purposes of this study, a hoar frost event occurs when RST ≤ 0 C and RST < T d (Takle, 1990;Karlsson, 2001) for at least 2 hr (no precipitation should be detected during the same period) same definition as the one used by Hu et al (2018). A day with a hoar frost event detected is called a hoar frost risk day.…”

“…Therefore, in this study, an occurrence of hoar frost is only recorded if RST is less than T d for at least 2 hr in order to cover all the possibly dangerous conditions caused by hoar frost thoroughly, same as the definition used by Hu et al . ().…”

Projected changes in days with zero‐crossings for Norway

“…Fewer zero-crossings may also lead to fewer days with slippery conditions, however, this connection is less certain (Šidlauskaitė and Kažys, 2019). Hu et al (2018) argue that increasing temperatures alone are not sufficient to assess such changes. In addition, humidity data (e.g., dew-point temperature) is required to assess changes in slippery conditions.…”

“…Projected changes in days with zero-crossings (DZCs) have been requested by several user groups in Norway (Hanssen-Bauer et al, 2017b). For certain applications, zero-crossings of the ground temperature are of interest; such applications include the development of icy conditions and hoar frost (Hu et al, 2018), and snow insulation effects on plants (Tierney et al, 2001;Kreyling and Henry, 2011). The topic of this study is, however, zero-crossings of the two-metre air temperature.…”