A Geostatistical Investigation into the Effective Spatiotemporal Coverage of Road Weather Information Systems in Alberta, Canada

Abstract

Inclement weather acutely affects road surface and driving conditions and can negatively impact traffic mobility and safety. Highway authorities have long been using road weather information systems (RWISs) to mitigate the risk of adverse weather on traffic. The data gathered, processed, and disseminated by such systems can improve both the safety of the traveling public as well as the effectiveness of winter road maintenance operations. As the road authorities continue to invest in expanding their existing RWIS networks, there is a growing need to determine the optimal deployment strategies for RWISs. To meet such demand, this study presents an innovative geostatistical approach to quantitatively analyze the spatiotemporal variations of the road weather and surface conditions. With help of constructed semivariograms, this study quantifies and examines both the spatial and temporal coverage of RWIS data. A case study of Alberta, which is one of the leaders in Canada in the use of RWISs, was conducted to indicate the reliability and applicability of the method proposed herein. The findings of this research offer insight for constructing a detailed spatiotemporal RWIS database to manage and deploy different types of RWISs, optimize winter road maintenance resources, and provide timely information on inclement road weather conditions for the traveling public.

Development of a Novel Road Weather Information System Location Allocation Model Considering Multiple Road Weather Variables over Space and Time

“…The GAM function is formulated as m = β 0 + f 1 ( x 1 ) + f 2 ( x 2 ) + … + f i ( x i ) , where m = variable of interest, β 0 = intercept, f i ( x i ) = smooth function of predictor x i . The smooth function can be expressed as f i ( x i ) = ∑ n = 1 m s ( x n ) ( 16 , 17 ). Once the data has been processed, statistical analysis is performed on the processed data set to produce descriptive statistics and correlation analysis among the weather parameters.…”

“…In the field of transportation, the usage of spatiotemporal data has been noticed recently for evaluation of spatiotemporal outliers and the identification of erroneous sites ( 14 ); traffic accident prediction using a deep learning approach ( 15 ); and several others. One notable study related to the topic of interest is the investigation of spatiotemporal variability of road weather and surface conditions using RWIS data from Alberta, Canada ( 16 ). The output of this study provided both spatial and temporal features of the RWIS database.…”

Developing Statewide Optimal RWIS Density Guidelines Using Space-Time Semivariogram Models

“…Here, m is the variable 7 Journal of Sensors of interest, β 0 is the intercept, and f i ðx i Þ is the smooth function of predictor x i . The smooth function can be expressed as 15,34]. Descriptive statistics (means and standard deviations) revealed relatively less variation in average monthly temperatures in the midwinter months than in the shoulder months.…”

“…In this study, the authors developed space-time semivariogram models using RWIS measurements from Alberta in order to examine the applicability of the method. The study's findings provided the spatiotemporal feature of the RWIS database [15]. However, the dependency of the spatiotemporal feature of RWIS measurements on topography and weather severity has yet to be scrutinized.…”

“…Spatiotemporal regression kriging has been used to predict precipitation using Moderate-Resolution Imaging Spectroradiometer (MODIS) and Normalized Difference Vegetation Index (NDVI) data [14]. In a study more relevant to the topic of interest, Wang et al (2019) investigated the spatiotemporal variation of road weather and surface conditions using RWIS data from Alberta, Canada. In this study, the authors developed space-time semivariogram models using RWIS measurements from Alberta in order to examine the applicability of the method.…”