October 5, 2007

Exit 22 on the Henry Hudson Parkway -- Bronx NY

The yellow arrow signs are hit regularly by cars that overshoot the short exit ramp and accidentally cross the divider. Notice the skid marks on the curb and one sign lying down on the ground.

Driveby 8 July 2007. Click here to see the movie, a 658kB .wmv file. In this clip only the left sign is standing. The other two are down.

To see an overhead vew of Exit 22 click on this link to Google Maps.
The last time I looked at the Google Maps image there were no sign shadows visible. The lighting and resolution in the image are appropriate to making sign shadows visible. The street light pole shadows are visible. So I guess there were no signs up when the Google Maps image was photographed. There is a slight chance the signs are lit edge-wise and therefore wouldn't leave shadows prominent enough to show in the Google Maps image. Google Maps changes images occasionally so I'm not positive that these comments apply to the current image.

The quantity of accidents has to be guessed. Some accidents don't knock down a sign. In roughly 14.5 months roughly 16 accidents happened that resulted in the loss of at least one sign, which is about one accident every 27 days. We don't have a way of counting how many accidents happened that would have knocked a sign down except for the fact that the sign was down already. The shortest time a sign was up before being knocked down was 3 days.

My neighbors guess that there are more Exit-22 accidents in winter than summer. Proposed reasons include: fewer total daylight hours per day; winter weather like ice and snow; daily peak use times occurring in the dark rather than in daylight; daylight savings time; seasonal affective disorder; and some sillier reasons. The actual numbers both support and don't support this idea. The total quantity of accidents over two years during the 6 Daylight Savings Time months is 16 while the quantity during the 6 Eastern Standard Time months is 19, and I doubt such a small difference is significant. But the high-accident months are January and February with 11 accidents over two months, and the low-accident months are June and July with 3 accidents over the two months, which looks on the face of it like it might be significant. But the sample is so small I don't think anything is proved yet.

Proposed Sign