A winter weather index for estimating winter roadway maintenance costs in the Midwest

摘要

Winter roadway maintenance budget data for the state of Iowa have been combined with available climate data for a 6-yr period to create a winter weather index that provides a useful assessment of winter severity. The weather index can be combined with measures of transportation department infrastructure within a region to estimate expenses for a given time period in the region. The index was developed using artificial neural network techniques that are nonlinear and perceive patterns in the input data. Winter weather severity as diagnosed by the index correlates well with Iowa Department of Transportation roadway treatment expenses. The neural network-based index is shown to perform better than the Strategic Highway Research Program (SHRP) index and an index developed using linear regression techniques.