In Part One, I put up graphs of the last 11 years of API readings for Shanghai, China. The graphs were big, crowded, whacky and JPGs, so it is hard for readers to see anything really tangible. So with the power of statistics, let us see how we can make these more useful, shall we?
Reminder: I make no claims of any sort about the accuracy of the source data, which was gathered from the Shanghai Environmental Education Center. The site and the ministry do not go into detail about their methods, testing sites, or any other pertinent details, which ruins the fun. (click to biggify, as usual)
This is a year on year graph of the 30 day running average. That’s a mouthful, but it’s relatively simple, each data point is an average of the 29 previous days and itself. This helps smooth things out without decimating the real shifts in Air Quality, good or bad. Click through for SO2 and NO2, and more fun graphs!
If you remember the Daily year-on-year graph of SO2 from Part 1, this is much more fun to read. What do you see?
I have to remind everyone that early 2000 data for NO2 (that annoying Orange line) is highly suspect, and probably actually NOx levels (the ministry was transitioning from measuring NOx to NO2).
These next graphs simply extend the concept of Running averages to 90days, which really flattens out the seasonal spikes but I think still shows air pollutant trends.
90 day running averages
90 day running averages are when I started spotting the real effects of the Shanghai World Expo, when I did this exercise for the first time, so lets see how the updated graphs look: