Beijing Pollution Data Visualization

By Ai Hongwei

  • 2-week collected Beijing Air Quality Data from 27 stations in every hour, four types of pollution, SO2, NO2, PM10 and PM2.5 are traced.
  • Data from 丰台云岗 and 门头沟龙泉镇 are imcomplete.
  • Stations name and location.
  • Calculate the average pollution for comparision.
  • Four types of pollution data have different range.


I use curves to descripe the pollution quanlity, because curves have a good common sense and can give a precise perception.

I create a line chart, every line in it stands a pollution trance and x-axis means time from 2012-10-17 to 2012-10-30 and y-axis means pollution quality. Because different pollution have different rang, y-axis can be self-changed driven by data.

Next, I use a bar chart to show the pollution statistic comparing to the average pollution. Red means this station area has pollution below average, blue means above.

You can also choose differnt pollution in one station for comparion.


Let's show some results use this visualization.

You may think 奥体公园 is good place for practise, but from the bar chart, we can see all pollution is high than other area.

And 植物园 is the best place, with all pollution below average.

And let's see our area's status, the pm_10 is heavily exceed the standard.

At last, let's see some unstable data with sharp peak, you can see the graph below, the so2 statistic is good but unfortunately the variance is large, so there are something uncertian effect the air quanlity suddenly.