OBJECTIVE METHOD FOR IMPROVING THE OPERATIONAL PERFORMANCE OF A ROAD ICE PREDICTION MODEL USING INTERPOLATED MEOSCALE OUTPUT AND A TEMPLET FOR CORRECTING SYSTEMATIC ERROR

Two techniques for improving the operational performance of numerical road ice prediction models such as MORIPM are presented. The output of the meoscale model has already been tailored to provide an input on an experimental basis. As the meoscale model grid-points are not exactly coincidental with specific forecast sites, a linear combination with four grid-point values is used to give better data for a specific site. Eight of the 11 sites chosen for the investigation showed improvements in the prediction of overall temperatures. The model was found to have a cold bias which may be due to simulation and physical processes and unpredictable influences of topography, altitude, traffic and thermal properties of the road construction. The averaging templet over a period of 3, 5 and 7 days was used to correct for the systematic prediction error. Too short a period may be unable to smooth out the necessary range of road surface temperature variations and a 7-days averaging period was preferable to take into account the weekly cycle of traffic. The templet option could be deselected if weather conditions change significantly from the previous few days.

  • Availability:
  • Corporate Authors:

    Her Majesty Stationary Office

    49 High Holborn
    London WC1V 6HB,   England 
  • Authors:
    • THORNE, J E
    • Shao, J
  • Publication Date: 1992-9

Language

  • English

Media Info

  • Features: References;
  • Pagination: p. 197-204
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00675141
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • Files: ITRD
  • Created Date: Mar 28 1995 12:00AM