ANALYSIS OF ROUTINE TRAFFIC COUNT STATIONS TO OPTIMIZE LOCATIONS AND FREQUENCY

This report describes a grouping of statewide permanent and key traffic counters on the basis of their geographic variations in traffic flow. Several factors were considered including the distance between clusters and urban versus rural areas. Traffic counts for a 3-year period were grouped into clusters by highway functional class for each individual parish. A computer program examines the maximum distance within a cluster and the maximum, average, and minimum distances within and between clusters. Count stations are arranged in clusters or groups of comparatively like counts. The cluster groups were examined by reviewing Louisiana parish maps that show the station locations. A computerized cluster analysis of all districts for 1977-1979 was reviewed according to highway functional classes 6, 7, and 8. An indepth review of the various cluster arrangements indicated the possibility of estimating the average annual daily traffic at some locations from sample traffic counts. A total of 111 stations could be measured less frequently by subjectively reviewing the location and proximity of the 2,290 stations, whereas a total of 1,246 stations could be read less frequently based on the objective but insensitive computerized cluster analysis. This observation confirms the need for further analysis by taking into consideration factors such as seasonal variations, geographic distribution of stations, and the number of stations existing in each cluster.