Advances in Regression Kriging-Based Methods for Estimating Statewide Winter Weather Collisions: An Empirical Investigation

Abstract

Winter conditions create hazardous roads that municipalities work hard to maintain to ensure the safety of the travelling public. Targeting their efforts with effective network screening will help transportation managers address these problems. In our recent efforts, regression kriging was found to be a viable and effective network screening methodology. However, the study was constrained by its limited spatial extent making the reported results less conclusive and transferrable. In addition, our previous work implemented what has long been adopted in most of conventional studies—the Euclidean distance; however, use of the road network distance would, intuitively, result in further improving kriging estimates, especially when dealing with transportation problems. Therefore, this study improves upon our previous efforts by developing a more advanced kriging model; namely, network regression kriging using the entire state of Iowa with the significantly expanded road network. The transferability of the developed models is also explored to investigate its generalization potential. The findings based on various statistical measures suggest that the enhanced kriging model vastly improved the estimation performance at the cost of greater computational complexity and run times. The study also suggests that regional semivariograms better represent the true nature of the local variances, though an overall model may still function adequately if higher fidelity is not required.

Should We Account for Network Distances or Anisotropy in the Spatial Estimation of Missing Traffic Data?

“…This state was chosen because of its higher density of counting stations and relative ease in acquiring data regarding its road network. In addition, the parameters of the semivariogram models, necessary for calculating the spatial estimates, may be heterogeneous across different geographic units (Wong and Kwon, 2021). Reducing the spatial coverage of the database, for instance AADT, ensures that the models capture local characteristics and, consequently, generate better estimates.…”

“…Thus, the contribution of anisotropy appears to be competitive with the use of network distances. As reported by some authors (Selby and Kockelman, 2013;Wong and Kwon, 2021), obtaining network distance matrices is computationally expensive, and also requires a perfectly connected network without topological faults to calculate the shortest paths. The results found in the present article, together with those of previous studies, suggest that, as the density of count locations increases, incorporating anisotropy may be a viable alternative to the use of network distances, if the main intention is only to provide accurate estimates.…”

“…Nevertheless, a recent study (Wong and Kwon, 2021) proved that network distances can, in fact, yield better estimates of winter weather collisions using Regression Kriging when compared to the Euclidean results. However, accidents can occur at any point along the road network, while traffic data assumes only one value in the length of a segment.…”

Analysis of Terrain Modelling Methods in the Coastal Zone

“…Figure 6 A graph showing the experimental semivariogram along with the fitted theoretical semivariogram[66] …”