more info on beijing's 6/18 air quality

Yesterday, beginning at around 10am, there was a sudden and dramatic spike of air pollution here in Beijing. I blogged about it here, and it was covered in the Guardian and Time's blog, with surely more to come. The pollution spike lasted until close to midnight yesterday. I presume yesterday evening's rain is what ended the event, although it should be noted that, as I write this, the pollution seems to be creeping up again.

Yesterday's pollution spike may be seen very clearly in the BeijingAir tweeted hourly data over the past couple of days. Shown here are PM2.5 concentration and AQI:


Note the missing data points in the afternoon of 6/18 and the maxing out of AQI at 500 during the same period.

Despite yesterday afternoon's stifling pollution, MEP's officially reported Air Pollution Index (API) for 6/18 was just 104 - indicating "slightly polluted" air quality. The reason, as noted yesterday, is that MEP's API does not report real-time air quality; it is an average air quality indicator covering noon to noon beginning from the previous day. Therefore, we wouldn't expect the afternoon pollution spike of 6/18 to show up until the 6/19 reported data point.

However, the API for 6/19, which was released a few minutes ago, is just 159 ("lightly polluted"), which is significantly lower than I would have expected.

--
Edit: An API of 159 - corresponding to a PM10 concentration of 266 ug/m^3 - still represents awful air quality, despite my use of the word "just." China's daily/yearly goals for PM10 are 150/100 ug/m^3, while the WHO's recommended targets are 50/20 ug/m^3.
--

The following graph shows MEP PM10 and API data, as well as BeijingAir PM2.5 and AQI data, for the last few days. Note that the absolute magnitudes of the BeijingAir and MEP data are not directly comparable due to slightly different measurements and scales. But the trending should be the same:


Although the MEP data increases beginning noon on 6/18, as one would expect, the increase just doesn't seem commensurate with the seemingly atrocious pollution yesterday afternoon and evening.

What's going on here? Well, there are a few options, but I'm not sure which one is correct:

First - it's theoretically possible that, because MEP averages over 24 hours over a number of different monitoring stations, the overnight reduction combined with lower pollution outside the city center brought the overall average down. Here is the daily individual monitor data from the Beijing EPB:

There are certainly some high readings, e.g. Dongsi, but there are also some that report only half as bad (Pinggu, Miyun).

Second - it is possible that the BeijingAir monitor is either not calibrated correctly or suffered some unusual activity (e.g. a car idling for an extended period outside the monitor). This seems unlikely.

Third - I'm not sure if this is correct or not, but it may be possible that the pollution event was largely PM2.5 - particles of diameter smaller than 2.5 microns - that do not register in the EPB's monitoring stations designed to measure PM10. Can anyone comment on this?

Fourth - I don't think I need to write this option explicitly.

More info as I learn it.