An empirical study on gap time and its relation to the fundamental diagram of traffic flow

The relation between flow rate and traffic density is of fundamental importance to traffic studies and is graphical representation is often called the fundamental diagram of traffic flow. Various shapes of the fundamental diagram have been suggested in the past. Despite numerous debates on which ones fit experimental observations better, it has not been well explained why the observed fundamental diagrams have different shapes. This study introduces gap time as a key parameter to decipher drivers' behavioral differences at different locations and under different traffic conditions, relating it to the shapes of the fundamental diagram. From I-80 freeway event detector data, it is shown that gap time follows a certain probabilistic distribution and its mean value varies over locations. It turns out that drivers take different gap times for different travel speeds. Three different types of gap time-speed diagrams are identified and matched to Greenberg, reversed-lambda, and inverted-V types of fundamental diagrams, respectively.