Sunday, October 21, 2018

Sinusoidal trend and global warming UPDATED

In a previous post from six years ago, I fit a sinusoidal trend, with auto-regressive component, to daily temperature data. (Spoiler alert: It's still getting warmer.) Recently I've received inquiries about the script for that analysis. I disinterred the ancient script, updated it, and grabbed more recent temperature data. The script and data file are linked below.

The result of the new analysis:
As you can see from the plot (above), the slope of the linear spine of the sinusoidal variation is 0.068 degrees Fahrenheit per year. The 95% HDI on the estimate spans zero, just as it did with the smaller data set in the previous post from six years ago. But I'm pretty sure that if this city were put into a big hierarchical model with lots of other cities across the globe, the high-level estimate of slope on the linear spine would be clearly greater than zero.

But evidence for global warming is not the point of this post. The point is to link the full script and data file. Here they are: R script; data. Hope this is helpful.