Geostatistical Innovation for Transportation Planning: The Role of Gaussian Geostatistical Simulation

This study applied a Gaussian geostatistical simulation (GGS) to enhance the estimation of the annual average daily traffic (AADT) on low-volume roads in Montana. Traditional methods such as short-term traffic counts, regression models, manual counts, engineering judgment, seasonal adjustment factors, ratio estimation, and the growth factor method often oversimplify spatial patterns and fail to capture the variability and uncertainty inherent in sparse datasets. The GGS model provides a robust alternative by incorporating spatial autocorrelation to capture the relationships across locations. Its ability to generate stochastic realizations enables a more comprehensive representation of the complex traffic dynamics. This study used data from 2009 to 2016 for AADT thresholds ranging from 400 to 2,000 vehicles per day by applying advanced geostatistical techniques. The approach uses R Studio for variogram modeling, simulation, and spatial visualization to effectively analyze traffic patterns. The results validated the hypothesis that GGS enhances predictive precision and captures spatial variability, which is closely aligned with the simulated and measured AADT values. The analysis of the dataset from Montana showed increasing variability at higher AADT thresholds, while maintaining consistent alignment across metrics. This confirmed the reliability of the model under stable conditions. This study also identified limitations, including underestimation of variability in extreme traffic conditions. These findings highlight the potential of GGS to enhance transportation and infrastructure planning by effectively addressing spatial uncertainties. In addition, it can improve data collection strategies and support informed infrastructure development decisions. Future refinements, such as adaptive modeling and real-time data integration, may further enhance the utility of GGS for traffic management and decision-making.