|M.Sc Student||Lina Shoukier Hadad|
|Subject||The Effect of Road Characteristics on Road Accidents on|
Local Interurban Roads
|Department||Department of Civil and Environmental Engineering||Supervisor||Professor Hakkert Shalom|
|Full Thesis text - in Hebrew|
This study investigates the level of road safety on low volume interurban roads in Israel, local, four-digit roads. In the first phase, the level of safety of all local interurban roads was compared with the level of all interurban roads.
An initial investigation established that data on many such local roads, are missing from the database of the Central Bureau of Statistics, including data on accidents, length of roads and traffic volumes. This made it difficult to establish the level of road safety for such roads. The comparison of levels of road safety indicated that serious accident rates per 1000 million vehicle- kilometers traveled were higher on the local roads than on the rest of the interurban network.
An important characteristic investigated in this study was traffic volumes. It was found that volumes on local roads in Israel are considerably higher than volumes on what, in the international press, is generally defined as local, low-volume, roads. It was found that 90% of local roads in Israel have traffic volumes of up to 15,000 vehicles a day.
In the second phase, a sample of 20 roads, which had particularly high accident rates, was selected. After the detailed investigation of the accidents, the roads on which they had occurred were surveyed and a number of road characteristics were measured, including vehicle speeds and traffic volumes.
From the speed survey conducted on these roads, the speeds measured were high in relation to the speed limits and in relation to the design speed. The 85-percentile speeds were generally higher than the speed limit and the mean speed.
One major aim of the study was to establish a relationship between the road characteristics and the accident rates. To establish such a relationship the various characteristics were combined and weighted, giving different weights to the characteristics. Three different weighting techniques were used. The first- GEO, assigned weights according to a formula that associates the variables with their contribution to the level of safety. Two other techniques were used, SCORES-GEO_CFA (a Common Factor Analysis) and a third one, termed SCORES-GEO_PC, assigned weights according to a Principle Component technique. According to these three techniques various relationships were explored, but none of the models produced consistent relationships.
The only relationships that were consistent were those between the geometric characteristics and travel speeds, and the relations between the geometric characteristics and traffic volumes.