Skip to content

Important notice

The building where our contact centre is based was evacuated on 13 June. Our contact centre and emails are up and running again, but please be patient as we have limited support available. We’ll do our best to help you if you need to contact us. For more information, read our latest media release.

Access keys for nzta.govt.nz

  • h Home
  • m Menu
  • 0 Show list of access keys
  • 2 Skip to content
  • 3 Skip to top
Back to Resources

Research Report 477 Modelling crash risk on the New Zealand state highway network

Published: | Category: Safety, security and public health , Research programme , Research & reports | Audience: General

This report presents an updated statistical analysis of data relating to crash rates on New Zealand roads. The research was carried out during 2007-2009 and it precedes the changes in 2010 to the New Zealand T10 specification. The refinements presented are associated with accounting for differences between the local and the general (ie design) speed environment, crash severity and interactions between curvature and roughness. The addition of these refinements will extend the present model's usefulness for guiding safety initiatives and providing economic justifications.

The regression model used in the analysis assumes that crashes are statistically independent and the number of crashes in each 10m segment of road follows a Poisson distribution. Inputs to the model include the average daily traffic (per side) and is a linear combination of the road characteristics, being transformations of terms that include factors such as gradient, curvature, out-of-context-curve effect, skid site classification, skid resistance, region and an urban/rural classification.

There is still more variability in the data than the Poisson model would predict. However, the results indicate the availability of a robust crash prediction model that takes into account both road condition and road geometry, allowing proactive identification of existing engineering-related road safety deficiencies and more importantly, the ability to quantify the potential for improvement.

Publication details

  • Author:
  • Published: March 2012
  • Reference: 477
  • ISBN/ISSN: ISBN 978-0-478-39422-1 (electronic)
Top