In my last post about the 35W bridge and the traffic problems it caused, I promised that I would look for some traffic-related science. So here’s a collection of links I’ve found, plus some memories of older stuff that I’ve read that I haven’t taken the time to track down.
- First, I want to point out the link provided by well caffeinated in the comments on the last post. It refers to a phenomenon that we’ve all observed: a traffic backup that suddenly clears up, with no apparent cause for the jam in the first place. This is termed a “phantom traffic jam,” and is unavoidable on crowded streets. Human drivers never keep their speed constant, and they have finite reaction times to the car in front of them, which inevitably leads to slowdowns. Make sure you watch the video that contains a simulation of just such a phenomenon.
- The Physics arxiv Blog has a critical review of what I believe is the same study.
- I remember reading some time ago that Bogota, Colombia, has fewer traffic problems than American cities with similar sizes and other factors. A simulation was performed, and the conclusion is that the aggressiveness of Bogota drivers was the key. They accelerate faster and brake harder than their American counterparts, which then leads to fewer phantom traffic jams, and lessens the effects of backups caused by a real obstruction. Interestingly, their traffic mortality rate is comparable to American cities (although there is probably more property damage, but I’m just guessing there). The full scientific article can be found here. (Alternating between stomping on the gas and stomping on the brakes is certainly not a good way to improve gas mileage, however.
- Have you ever noticed how you always seem to end up in the slowest lane during a traffic jam? And when you switch lanes, your new lane becomes slow while your old one speeds up? It’s not just you. You’re simply a victim of statistics. The slow lane is slow because there are more cars in that lane. Simple probability states that you’re more likely to end up in the slow lane for that reason. As for why it gets slow after you change lanes, it’s the effect of everybody else who had the genius idea to change lanes farther up. When you merge into the fast lane, you slow it down. And if enough people merge into the “fast lane,” it’s inevitably going to bog down.
- How should you merge when a lane is closed up ahead? Your driver’s ed instructor probably gave similar advice to that found in this article. (Which also talks about phantom traffic jams and automated cruise control.) The conscientious, defensive driver should merge well ahead of the lane closure, unlike the jerks who fly past and try to merge at the last second. I have no studies to back up this claim, but I think this approach is wrong. The earlier you restrict the “pipe,” so to speak, the greater the backup you’re going to cause. It would be better to use all of the lanes for as long as possible to get the maximum traffic flow. Of course, there are limits. We can’t all merge in the last ten feet, obviously. But it also makes no sense to merge two miles ahead, effectively increasing the length of the lane closure, and causing greater traffic backup.
What are your thoughts about the science behind traffic? Any other cool science that I missed?