This is regarding the article "Studying the Ebb and Flow of Stop-and-Go" by Alan Sipress [front page, Aug. 5]:

While the Los Alamos research is interesting theoretically, it is arguably of little practical importance. All models are wrong, some are useful. The Los Alamos model is not useful. Transportation models are useful only in the sense that they help people think when discussing projects.

Modeling traffic as a gas or something equally mindless leaves out the most important part of traffic: human beings. Driver behavior is the greatest stumbling block of all traffic models.

Even if by some fluke it does mimic traffic exactly, the microscopic regional nature of the Los Alamos model renders it a useless planning decision support system. The average transportation planner is usually forced to make politically correct decisions on a regional scale relying on a much coarser level of analysis. And seldom does he get the opportunity or budget to use a beast such as the one being created at Los Alamos.

Any available monies -- federal, state, local or private -- should be spent on updating and refining these existing models rather than on pie-in-the-sky ground-up ones such as Los Alamos.



The recent article on correlating traffic flow to natural phenomena is interesting, but modeling every car on the road to see how traffic reacts to certain stimuli will not solve the reality of the situation. More concrete solutions to traffic problems can be implemented without exhaustive analysis, with what I call common sense.

Traffic is like blood flow in the body, and when it clogs, the D.C. metro area has a "traffic stroke." We have all been there when the Wilson Bridge has been clogged or shut down. The result of this event is devastating. Slow traffic in urban areas wastes fuel and valuable time, increases pollution and causes untold frustration.

The computer modeling being developed at national labs could be used to control lights in response to various traffic situations. This type of system would be a major effort requiring the input of many traffic cameras and road sensors. Some of these inputs already exist at area traffic operations centers. What is needed is a feedback to area lights with the control of computer technology. Such a system would default to the present system when the computer goes down.

In addition to real metro traffic control, not mere observation, there are many low-tech solutions to our traffic problems. For instance, better planning by regional committees on growth. More and taller buildings go up in the suburbs, but the roads do not get any wider. This needs to change, or the traffic situation will not get better.

For major accidents, heavy-lift helicopters could be used to get overturned trucks off the road, especially if the truck is on a major bridge. The helicopter is worth the cost in terms of time wasted by area drivers sitting for hours on the Beltway. Installing signs that would activate and indicate alternate routes when major roadways are clogged would allow better traffic flow and prevent a traffic stroke.

Eventually, more roads interconnecting the suburbs may have to be built. Through thoughtful planning, and the implementation of high-tech and common-sense solutions, we might be able to get home in time to eat dinner at a decent hour.