Introduction

This pilot research web site shows modelled and observed road surface conditions (road weather) as well as road dust emission data at over 400 road weather and air quality stations in Norway. In addition, from November 2023, modelled and observed runway surface conditions at 47 airports in Norway are presented. The long term aim of this research is to provide support for road and runway weather forecasting, winter road maintenance activities and air quality management. Its development is supported by the Norwegian Meteorological Institute.

This pilot version is intended for research and development purposes only and is hosted by the Division for Climate Modelling and Air Pollution (FoU/KL). Any comments or queries can be directed to Bruce Rolstad Denby . The solution has been developed by Eivind Grøtting Wærsted, Augustin Mortier and Heiko Klein (FoU/KL).

User guide

The web site is intended to be intuitively easy to use. The user is presented with a map containing all the available road weather measurement sites as well as air quality measurement sites that are near roads. These are colour coded for easy identification. Click on the button "Legend" in the lower right corner of the map to view an explanation of these colours.

Clicking on a site, or typing in the name, on the map will provide the user with the modelled and measured road weather information available at that site. By default the time period will be 5 days, starting around 2.5 days before the current date and covering up to 2.5 days into the future. In this way a recent comparison with observations can be made to help assess the quality of the forecast. The model time series shown, prior to the current time, are constructed from the forecast hours 6 to 12 of the previous forecasts (since the forecast is made every 6 hours) and correspond to the same times shown in the air quality forecast. Times are given in UTC.

A drop down menu is available in the top right corner of the map. The start date and period in days can be changed by the user here. In addition a number of filters are provided to help with station selection:

Under the heading "Datatypes", the user can choose to only show road weather sites that have a certain type of observational data available for the selected time period. The option "all" means that all stations are shown, which is the default.

Under the heading "Categories", the user can select which types of stations should be shown on the map. By default, road weather sites and air quality sites are shown, but it is also possible to show traffic counting stations. The option "Bridges" will make the map show air quality and road weather sites located on bridges even if no other categories are active. The options "Airports" and "Custom" are not yet implemented.

Under the heading "Cameras filtering", the user can activate "Show cameras" to enable viewing the latest picture from roadside cameras.

The menu also provides the option to switch the map symbols from "Station type" to "Forecast symbol". This will cause each station on the map to be shown with a symbol indicating the modelled road conditions now. Traffic count sites will not be shown in this mode, though.

On the time series plots it is possible to hide values by clicking on the name in the legend. Images can also be downloaded by selecting from the drop down menu in the top right corner of each time series plot.

The meteorological data provided comes in four separate tabs and one additional tab showing air quality when an air quality station is selected. There is also an additional tab for road maintenance activities. If a traffic count site is selected, none of these tabs are available and there is only a single tab showing modelled and observed traffic volumes.
  • Temperature and Precipitation
    • Road surface temperature (°C), modelled and observed (if available)
    • 2 m dew point temperature (°C), modelled and observed (if available)
    • 2 m air temperature (°C), modelled and observed (if available)
    • Precipitation (mm w.e.), modelled and observed (if available)
  • Surface conditions
    • Road surface water (mm w.e.), modelled and observed (if available)
    • Road surface ice (mm w.e.), modelled and observed (if available)
    • Road surface snow (mm w.e.), modelled and observed (if available)
  • Meteorological information
    • 2 m relative humidity (%), modelled and observed (if available)
    • 10 m wind speed (m/s) and direction, modelled and observed (if available)
    • Surface energy balance (W/m2), modelled only
  • Road dust emissions and loadings
    • Hourly traffic volume for heavy and light weight vehicles. These traffic volumes are not real time but are predetermined based on historical data. Information on studded tyre share and speed limit is also provided. These modelled traffic volumes are also shown together with measured traffic counts when traffic counting sites are chosen.
    • Emissions of PM10 (Particles less than 10 um in diameter) calculated by NORTRIP in g/km/hour. Emissions are split into direct emissions from road, tyre and brake wear, suspended emissions from dust on the road surface, exhaust emissions and wind blown emissions (not currently active)
    • Dust, salt, dust binder and sand loading on the road surface in g/m2 calculated by the model. Only particles less than 200 μm in diameter are shown. Roughly 30% of these are estimated to be < 10 μm in diameter.
    • Suspension potential calculated by the model. This reflects the surface conditions and the ability to emit particles from the surface. Dry conditions allow suspension and direct wear emissions whilst wet conditions suppress suspension and direct wear emissions.
  • Air quality
    • Modelled and measured PM10 concentrations including source apportionment based on the model calculations
    • Modelled and measured PM2.5 concentrations including source apportionment based on the model calculations
    • Modelled and measured NO2 concentrations including source apportionment based on the model calculations
  • Temperature and precipitation NA (Nordic analysis)
    • The information presented here is the same as in the first tab however the model results are based on the Nordic analysis data. When running NORTRIP, every forecast is pre-prepared by running the last 6 hours on Nordic Analysis data. This ensures that the initial conditions for the model at every forecast startup are initialised using these data. This tab does not provide a forecast but is a cross check showing the actual initial conditions for a forecast calculation.
  • Activities
    • This tab shows the road maintenance activities (road salting, sanding, ploughing and cleaning) that are predicted by the model, based on a set of rules (see the "Modelling" tab). "Salt1" is normal salt (NaCl), applied to prevent icing, while "Salt2" is a dust-binder (MgCl2 or CaCl2) used to reduce the suspension of road dust from the surface. Both of these are given as dry weight. "Wetting" refers to the water that is added to the road together with salt. Ploughing and cleaning is given as a number between 0 and 1, according to the efficiency at removing snow/dust that is assumed in the model.
    • In addition to modelled activities, the activities registered by road maintenance personnel are also indicated, labelled "Zeekit" as these data are provided via an API from this company. Currently, these registrations are only available in Trondheim municipality (covering 8 stations) from November 2020 as a trial. Ploughing and cleaning is always 0 or 1, since data on efficiency is not provided.

From the perspective of road weather it is interesting to see the possibility of ice forming on the road surface. In the model ice can be formed by either freezing water already on the surface, from rain that can freeze directly on a cold surface or from the deposition of ice on the road surface by condensation from the atmosphere. Dew point temperature is important in this regard and is therefore presented in the same graph as the surface temperature. If the dew point temperature is higher than the road surface temperature then condensation of water or deposition of ice can take place.

In the model the application of salt will reduce the vapour pressure of the snow/ice/water and will allow a certain amount of the ice/snow to melt. Whilst in reality salting of roads will often remove all snow/ice, the amount of ice/snow that melts in the model reflects the vapour pressure equilibrium and does not properly include mixing by traffic and other effects. For this and other reasons not all ice /snow may disappear after salting in the model.

All road weather stations measure road surface temperature and air temperature. Some stations also measure wind, humidity and precipitation. A very few stations also measure the road surface conditions.

Modelling

Modelling of road weather is carried out using the NORTRIP model, see list of references below. The model has been operational since November 2018 at the Norwegian Meteorological Institute and is used for calculating emissions of road dust as part of the Norwegian air quality forecasting system. The model is also used for calculating historical emission data for the years 2016-2021 as part of the Air quality expert user service.

The NORTRIP model is a combined road dust and road weather model. In order to model road dust emissions it is essential to also model road surface conditions, since the state of the road surface (water, ice or snow) determines if these emissions occur. A number of physical processes, in regard to road surface conditions, are described by the model and include:

  • Water mass balance on the road surface including source and sink terms such as precipitation, evaporation/condensation, drainage and vehicle spray
  • The impact of salt application to the road on the surface vapour pressure and freezing point
  • The impact of dust binding chemicals on the surface vapour pressure and freezing point
  • Energy balance of the road surface that takes into account the incoming and outgoing long and short wave radiation, turbulent latent heat flux through evaporation/condensation, turbulent sensible heat flux and heat fluxes beneath the surface. The main result of the surface energy calculation is the prediction of surface temperature and surface evaporation/condensation.

In forecast mode, used for the air quality forecasts, the NORTRIP model is run 66 hours ahead in time. This limit is imposed by the length of the meteorological forecast, which NORTRIP uses as input. The NORTRIP and the air quality forecast is updated every 6 hours. In order to provide NORTRIP with the most up to date meteorological information to start the calculation then the calculations are started 6 hours prior to the current meteorological forecast. During this 6 hour period use is made of Nordic analysis data. This is a dataset updated every hour that combines both measurements and modelling results to give the best representation of temperature and precipitation for Norway at 1 km resolution. Calculating in this way means that the initial state of a NORTRIP forecast is always based on Nordic analysis data.

Currently model output for road weather data is limited to the roads closest to the measurement sites. 369 road weather and 74 air quality station sites are modelled in this way and are available through this web solution. For the air quality forecast NORTRIP calculates emissions at approx. 720 000 road links. Eventually road weather information will also be provided for all these road links.

The application of salt, sand and dust binder in the model is based on a set of rules and will only partially reflect real world activities. Salting is dependent on meteorological conditions, as is dust binding and sanding to a lesser extent. However, each municipality in Norway has a number of additional rules concerning which road types and which traffic volumes are required to allow these activities to take place. Snow removal occurs automatically on all roads when snow depth is greater than 3 mm.w.e. but not more than once every 3 hours. The maximum allowable water film depth on a road is 0.6 mm. Only when snow or ice have accumulated on the road surface can the water depth become larger.

In addition to the NORTRIP calculations, model calculations for air quality are also provided when an air quality station is selected. These are the same data used for the air quality forecasts and can be retrieved using the air quality forecast api. More detailed information concerning the air quality forecasts and how they are made can be found here.

Published model description

  • Denby, B.R., Sundvor, I., Johansson, C., Pirjola, L., Ketzel, M., Norman, M., Kupiainen, K. , Gustafsson, M., Blomqvist, G., Kauhaniemi, M. and Omstedt, G., 2013. A coupled road dust and surface moisture model to predict non-exhaust road traffic induced particle emissions (NORTRIP). Part 2: surface moisture and salt impact modelling. Atmos. Environ. 81, 485-503. DOI: http://dx.doi.org/10.1016/j.atmosenv.2013.09.003
  • Denby, B.R., Sundvor, I., Johansson, C., Pirjola, L., Ketzel, M., Norman, M., Kupiainen, K. , Gustafsson, M., Blomqvist, G. and Omstedt, G., 2013. A coupled road dust and surface moisture model to predict non-exhaust road traffic induced particle emissions (NORTRIP). Part 1: road dust loading and suspension modelling. Atmos. Environ. 77, 283-300. DOI: http://dx.doi.org/10.1016/j.atmosenv.2013.04.069

Other NORTRIP related publications available

  • Denby, B.R. , M. Ketzel, T. Ellermann, A. Stojiljkovic, K. Kupiainen, J.V. Niemi, M. Norman, C. Johansson, M. Gustafsson, G. Blomqvist, S. Janhäll, I. Sundvor, September 2016, Road salt emissions: A comparison of measurements and modelling using the NORTRIP road dust emission model, Atmospheric Environment Volume 141, Pages 508-522, ISSN 1352-2310, DOI: http://dx.doi.org/10.1016/j.atmosenv.2016.07.027.
  • Norman, M., I. Sundvor, B.R. Denby, C. Johansson, M. Gustafsson, G. Blomqvist, S. Janhäll, June 2016, Modelling road dust emission abatement measures using the NORTRIP model: Vehicle speed and studded tyre reduction, Atmospheric Environment, Volume 134, Pages 96-108, ISSN 1352-2310, DOI: http://dx.doi.org/10.1016/j.atmosenv.2016.03.035
  • Kauhaniemi, M., Stojiljkovic, A., Pirjola, L., Karppinen, A., Härkönen, J., Kupiainen, K., Kangas, L., Aarnio, M. A., Omstedt, G., Denby, B. R., and Kukkonen, J., 2014, Comparison of the predictions of two road dust emission models with the measurements of a mobile van, Atmos. Chem. Phys., 14, 9155-9169, DOI: http://dx.doi.org/10.5194/acp-14-9155-2014
  • Ana Stojiljkovic, Mari Kauhaniemi, Jaakko Kukkonen, Kaarle Kupiainen, Ari Karppinen, Bruce Rolstad Denby, Anu Kousa, Jarkko V. Niemi, and Matthias Ketzel 2019 The impact of measures to reduce ambient air PM10 concentrations originating from road dust, evaluated for a street canyon in Helsinki. Atmos. Chem. Phys., 19, 11199–11212, DOI: https://doi.org/10.5194/acp-19-11199-2019
  • K. Kupiainen, B.R. Denby, M. Gustafsson, C. Johansson, M. Ketzel, J. Kukkonen, M. Norman, L. Pirjola, I. Sundvor, C Bennet, G. Blomqvist, S. Janhäll, A. Karppinen, M. Kauhaniemi, A Malinen, A. Stojiljkovic 2017 Road dust and PM10 in the Nordic countries: Measures to reduce road dust emissions from traffic. Nordic council of Ministers NR Nr. 2016:790. ISBN (Trykt) 978-92-893-4800-3. ISBN (Elektronisk) 978-92-893-4801-0. DOI: https://www.norden.org/no/node/7384

Measurements

Road weather stations measure a number of meteorological parameters and provide an overview of the weather conditions along the road network in Norway. Currently there are 379 road weather stations in Norway maintained by the Norwegian Public Road Authorities (Statens Vegvesen, www.vegvesen.no). All road weather stations normally measure road surface temperature and air temperature. Some stations also measure wind, humidity and precipitation. A very few stations also measure the road surface conditions. The data collected are:

  • Air temperature (° C)
  • Road temperature (° C)
  • Relative humidity (%), which gives dew point temperature (° C)
  • Precipitation amount (mm), precipitation intensity (mm / time) and type of precipitation (drizzle / rain / hail / snow)
  • Wind speed (m / s), max. gust of wind last 10 minutes (m / s) and wind direction
  • Road surface conditions (water/ice/snow)

An overview of the sites and their requirements can be found here.

Data from the weather stations is collected by the Norwegian meteorological institute every 10 minutes and stored in the database along with other meteorological observations and is accessible through the Frost api. A limited amount of automatic quality control is carried out on the data.

In addition to road weather measurements air quality observations are also included when air quality stations are selected. Air quality measurement sites are owned and maintained by both municipalities and the road authorities (Statens Vegvesen). These data are collected and quality controlled by the Norwegian National Reference Laboratory for Air and further distributed through the NILU api.

The road authorities also measure traffic volume at several thousand locations. Hourly data on the number of vehicles of different lengths that pass in each direction and lane are publicly available through an API. When the "traffic count sites" switch is on, the observed and modelled traffic can be compared at all these locations.

Acknowledgement

Coming soon.

FAQs

Comparison with measurements in an earlier season 2015-2016 have shown that a typical error for the forecast, out to 1 day, is around 1.3 C (Mean absolute error). However, this varies from station to station and from period to period. The quality of the surface temperature forecasts are directly dependent on the quality of the meteorological forecasts.
Currently there are very few measurement sites available to assess the quality of the surface condition modelling. This assessment is in progress. Previous calculations with the NORTRIP model have shown that the model correctly predicts dry or wet road conditions around 80 - 85 % of the time, but this does not say the exact state of the road surface is correctly calculated.
Measurements also have quality problems. Though there are some automatic checks to assess these it is quite possible for measurements, particularly road surface measurements, to give inaccurate results, particularly the amount of water/ice/snow may be poorly distributed. More on this to come …