## Daylight Saving Time

I did some mathematical jiggery-pokery and came up with this pretty little graph.

You will need to click on that to get it big enough to read.

The graph shows sunrise, sunset, waking hours, and business hours through the course of the year. Solar noon is at the center, sunrise is above, and sunset below. It makes some pretty wild assumptions, too. It is based on an average observer at 40°N latitude. It assumes that the average observer wakes at 6am and goes to bed at 10pm. It assumes business hours of 9am to 5pm. It assumes that the observer is longitudinally situated in the center of an ideal time zone, so that local dynamical noon (of the clock) coincides with local apparent solar noon (when the sun is at its zenith). It also assumes that the average observer desires never to wake earlier before sunrise than he does on the winter solstice. According to my calculations, the average observer wakes approx. 75 minutes before sunrise on the winter solstice. The black curve above represents this time carried throughout the year. Notice that at the end of DST, as it is currently practiced in the United States, the observer wakes earlier than 75 minutes before sunrise for about three weeks - the last three weeks of Daylight Saving Time - before standard time resumes and he again wakes less than 75 minutes before sunrise. Also notice that a similar situation does not occur at the beginning of DST. In fact, DST could start up to one week earlier than it currently does and the observer would still wake no more than 75 minutes before sunrise - no earlier than he does on the winter solstice.

Therefore, there is a morning-light saving inefficiency at the end of DST, but not at the beginning. Assuming that it would be inefficient to have the observer spend more time awake in the dark in the morning than he does on the longest night of the year.

In 1966, Daylight Saving Time started six weeks later than it does now, and ended one week earlier. There was still an inefficiency of two weeks at the end, but none at the beginning. In 1987, DST started three weeks later than it does now, and ended one week earlier. Same inefficiency. The 2006 amendments increased the inefficiency at the end of DST to three weeks, but still did not create an inefficiency at the beginning. In fact, DST could start up to one week earlier than it currently does and still not result in morning-light inefficiency.

The problem is at the end. Daylight Saving Time runs three weeks too long, resulting in a wake time of nearly 2 hours before sunrise on the second to last day of DST. (The last day of DST is a Saturday - the observer will in all likelihood sleep in.)

Of course, the other problem with DST is that it is forced on us by Congress. Aside from that, though, the most efficient practice of DST would have the clocks set ahead one hour from the first Sunday in March to the second Sunday in October. Apologies to trick-or-treaters.