Development and Verification of a Numerical Forecast Model for Road Meteorological Services

Accurate road meteorology forecast and road traffic information are very important to road transportation security. Road surface temperature is a crucial parameter in traffic weather forecast. Now there are three main kinds of road surface parameters forecast model: Statistical model, GIS-based model and physical model. Physical model is widely used and it mainly considers the road surface energy balance model and the effect of anthropogenic heat. In 2008, based on the rapid update cycling forecast system (BJ-RUC), the road weather information system is developed and run operationally by the Institute of Urban Meteorology. Since 2007, Beijing Meteorological Bureau has established 18 weather stations along the express way using the apparatus manufactured by ROSA Vaisala in Finland, and established 8 visibility observation stations using the digital visibility sensor. These all make the fine traffic weather forecast and operational run possible.A fine numerical model for urban road surface temperature (RST), snow depth and ice depth prediction (BJ-ROME) is developed based on Common Land Model (CoLM). The model is developed according to characteristics of the road surface, and based on the data of express way weather observation and fine land surface data of Beijing. The model is forced by the meteorological data output from BJ-RUC, and the forecast and update time span is 24 hours and 3 hours, respectively. The model is validated using in-situ observation data measured by the ROSA road weather stations of Vaisala Company, Finland. The sensitivity analysis is also implemented.Nine sites are chosen to validate the RST prediction results of BJ-ROME. The validation time is during 9—24 Aug 2009, when the RST is very high in Beijing. Four sites, i.e., Xihongmen, Wenyuhe, Xiguan and Lugouqiao are chosen to validate the snow depth prediction results of BJ-ROME. The validation time is during 2—5 Jan 2010, when a big snowfall happens in Beijing. The validation results indicate that BJ-ROME can successfully simulate the diurnal variation and maximum value of RST both under clear-sky and rainfall conditions. The validation results also indicate that BJ-ROME can successfully simulate the accumulation time and the variation and maximum value of snow depth. The results of sensitivity analysis indicate that road surface evaporation and the anthropogenic heat are very important in road surface temperature forecast. The forecast results of BJ-ROME can be used as an important reference to take measures by traffic administration department and road administration department.In the near future, BJ-ROME will be coupled in double direction with BJ-RUC to improve the forecast. The anthropogenic heat (AH) should be parameterized more precisely, and the variational assimilation algorithm should be used to assimilate the RST observations. The predicting performance of water depth of BJ-ROME should also be validated.