AbstractNetwork forecasting has traditionally been conducted using survey data collected by field-based surveys of the road network to model the microclimate surrounding the road network. A review of the literature in the subject area identifies that the current methods rely heavily on a field based survey approach. This research examines the premise that these field surveys replaced by modern GIS modelling methods. The research goes on to describe the development of a variety of GIS based methodologies and software for the calculation of geographical parameters for input into the Geographical Road Ice Prediction model and the dissemination of the resulting ice forecast through
a web based GIS system. The methods developed are then discussed in relation to a trial conducted in Hampshire in the winter of 2006/7, which test the methods in a real world scenario. The final element to this research is the exploration of a variety of different GIS datasets to test the methodologies developed and investigate the impact of varying the methodologies and data sources on the network forecast for a small locality in the north of Hampshire. Results suggest that a number of the methodological parameters developed could be adjusted without affecting the
resulting forecast and therefore improve the efficiency of the modelling process. It also identifies which of the GIS datasets available in the UK produce the best forecasting results. The research concludes that the methodologies developed have successfully helped predict ice formation on road networks without the use of any manual survey data. Moreover, the methods developed can be used in other research fields. In particular, this research has found a more accurate method for modelling building heights from medium resolution IFSAR data, which has resulted in an 11% improvement in building height estimations. Finally, this research goes onto discuss further research and how a more detailed assessment of the methodologies under different climatic and geographical
conditions the methods developed, have the potential to replace parts or all of the survey methods currently used in ice prediction surveys.
|Date of Award||Apr 2010|
|Supervisor||George Taylor (Supervisor), Mark Ware (Supervisor) & Gary Higgs (Supervisor)|
- Geographic Information Systems