Optimizing Traffic Count Program: A Methodology for Estimating AADT Volumes from Short-duration Counts

About the report:

Estimates of annual average daily traffic (AADT) volumes are important in the planning and operations of state highway departments. These estimates are used in the planning of new construction and improvements of existing facilities, and, in some cases, in the allocation of maintenance funds. It is, therefore, important that any method used in obtaining the estimates provide data of sufficient accuracy for the intended use. This importance of having reliable and current data on traffic volumes at hand is generally recognized, and over the years data collection programs have tended to expand. This expansion has led to huge amounts of money being spent annually for the collection and analysis of traffic data. Renewed efforts are, however, now being made to reduce the annual expenditure on traffic counts while at the same time maintaining the desired level of accuracy. A study is, therefore, being carried out by the Council to develop an optimal counting program for the state. This interim report presents the results of that portion of the study in which the feasibility of estimating AADT volumes from short counts was established. The procedure was first to use 1980 data for 16 continuous count stations to determine periods that are stable throughout the year for different short counts. It was found that stable periods for short counts occurred mainly on Mondays, Tuesdays, and Wednesdays, and expansion factors were then developed for short counts of different durations and different starting times for these days. The expansion factors were then used to estimate 1981 AADT's from short counts extracted from data obtained in 1981 continuous counts. The results indicate that relative errors of less than 10% were obtained for AADT's estimated from counts of 6-, 8-, 10-, and 12-hour durations on Mondays, Tuesdays, and Wednesdays-The results for Tuesdays and Wednesdays tended to be more accurate than those for Mondays, and counts taken between February and November tended to give more accurate results than those taken in January and December.

Authors