It's 74 and Sunny in Durham

Let's get dressed up and have a garden party . . .

More...

Less...

Perfect in 10

I don't trust the Weather Channel very much, particularly since reading a funny article about their numerous foibles and failures in the New Yorker back in 2000. Things have improved a lot in eight years, but the 10-day forecast is still reliably unreliable, and my real favorite weather source doesn't bother with the crap shoot that is a forecast longer than five days.

Nonetheless, we are now within the "envelope" of professional weather prediction, and it is looking mighty fine!



From my direct observation, I have one very solid prediction: it's going to be chilly after dark. Assuming that the weather is more likely to get cooler than warmer over the next 10 days, I want to strongly reiterate to our female guests the particular importance of bringing a sweater, jacket or wrap.

More...

Less...

Sunrise/Sunset cont...

In an earlier post, I mentioned how perfect it was that sunset was going to be during the cocktail hour, right before dinner. We were very proud of ourselves for inadvertently planning around the rotation of the earth. However, we have recently made several observations, after discussions with photographers, that give us pause.

Our wedding is three weeks after the fall equinox. This weekend is three weeks before the spring equinox. Thus, physics and astronomy being what they are, this weekend should have been a very good indicator of what the sun will look like on our wedding day. And let me tell you, the lighting at 5pm is horrible! It gets dark a lot sooner than you might expect a day to get when sunset is officially at 6:45 pm.

We are therefore fairly certain that the ceremony will be at 4pm now, rather than at 5pm. Because as much as we reject many notions of traditional wedding photography, lighting remains one of the few things we cannot defy.

More...

Less...

Sunrise/Sunset...

It is clear that I am destined to obsess over every little detail of this impending wedding, including details that I cannot control. Such as: the weather and the sunset.

Now, this being the 21st century and astronomy and physics having progressed very far in the last 500 years, I can say with much certainty that the sun will set at precisely 6:45 pm. This fits in nicely with cocktail hour; shortly after the sun sets, the guests proceed from the veranda with their mint juleps to the main barn for dinner...



The weather, however, is less predictable, particularly with global warming relentless attacking us the way it does these days. I have done a careful analysis of the weather in Durham on October 11 over the past 62 years. Graph 1 shows a histogram of the average temperature (F) binned by 5 degrees. Fortunately it falls to an approximate Gaussian, meaning the sample size is large enough to make some guesses. The average temperature is 60 degrees with a standard deviation of 7, pretty much a nice fall day. But that is the mean of averages, which doesn't really tell us anything. We really want to know what are the chances of it being too hot or too cold.


Graph 2 is a histogram of the high temperature (F) with the same binning. Not as nice looking but we can still get some useful information. The average high for that day is 74 +/- 8. And since the wedding is most likely starting at 5:00, we should expect us to be just coming off such a high. The lows similarly fit more or less to a Gaussian, giving an average overnight low of 50 +/- 9.


Of course, global warming is still a concern. Fortunately the daily highs, lows and averages remain consistent if I only average them over the last 30 years. Plotting the temperatures in a scatter plot reveals no discernible pattern (Graph 3). This is good. It means that I am unable to predict the weather with high school level statistics.

So, we've covered sunset and temperature. All that's left is rain. That we will have to save for another day.

More...

Less...