I've stopped maintaining this tool/website, as Google changed their API for Fusion Tables and I simply don't have the spare time to spend updating this project so it keeps running.
You can get detailed air quality information at http://aqicn.org/
You can also have a look at the raw data my tool gathered over the years at:
What follows is the old
Graphing air quality readings from several major Chinese centers: Beijing, Shanghai, Guangzhou, Chengdu & Shenyang.
This is the source code of the project which is viewable at http://kopf.github.com/chineseair/
These files may be of interest:
chineseair/updater.py, which scrapes the data from web feeds
chineseair/google.py, the interface for the Google Fusion Tables API
chineseair/process_historical_data.py, which processes full historical records (see here for an example) published on stateair.net
Why is this repo so huge? Why so many commits?
I put this project together in an evening. In its first incarnation, the
updater.py script would simply scrape twitter and save any new data to a json file in this repo, commit and push. This json file would then be loaded when the user visited the project webpage.
It was a very quick-and-dirty solution, and led to the tens of thousands of commits (and the huge repo size) today. I would like to clean up the git history, but I'm reluctant to, since people have already forked this repo.