Effect of Weather on the Relationship Between Flow and Occupancy on Freeways

The relationship between flow rates and roadway occupancies on freeways has been investigated in several recent papers. However, all the data in those investigations have come from ideal conditions. The purpose ofthis paper is to investigate the same relationship under adverse weather conditions. An eart:^-¿--l-,4^---l ¿L^¿ -^!-^r--^^l 4-^------rtçr JfuuJ ruu[u uräf töttl lËuucËu rt'ccyväJ çapaut¿y, uut tl uru not address the effects of rain on the nature of the function relating the two variables. Three possible effects on the function are investigated, and it is found that, in essence, the slope of the line relating flow to occupancy (in the uncongested regime) decreases as weather conditions deteriorate. In other words, different parameters are needed for the function to describe the relationship under different weather conditions. This finding has some important implications for efforts to develop a new incident detection algorithm for freeways, based on the nature ofthe flow-occupancy relationship. This paper addresses the effect of bad weather on freeway traffic operations. In particular, does rainy weather change the nature of the relationships among speed, flow, and roadway occupancy? Although it might be objected that the existence of such an effect is obvious and hardly merits detailed attention, it turns out to be important to develop a quantitative description of the effect for use wÍth an automatic incident detection logic. This paper begins by briefly describing the background for the incident detection logic, in order to explain the rationale for developing a solid quantitative treatment of the effect of bad weather. The second section describes the data that were used for the analysis, with particular reference to weather conditions. The third section, dealing with the analysis, includes both a discussion of the methods used and a presentation of the results. The fourth section presents the conclusions that have been drawn from the investigation.