Averages - what do they mean?

November 2018 was an interesting month. We had the early peak in pollution with Guy Fawkes' night the likely culprit. You can see that on the graph in the region before Week 45.

PM2.5 graph for November 2018

The line on the monthly graph does not tell the whole story though - PM2.5 readings hit 85 μg/m³ on the evening of the 5th. This maximum is lost in the averaged monthly line so that the monthly line records a maximum which is less than half the actual. Monthly figures will need to be averaged over time, but it would be an interesting exercise to try to determine how to represent the maximums and minimums, along with a reasonable representation of what actual exposure would be like if you were breathing the outside "fresh" air in 24/7.

The sensor had a loss of internet connection the week after that, which is explains the lack of data at the start of Week 45.

The wind and the rain

The weather was bad during that week, which explains the drop in pollution - rain clearing the air and winds blowing what remained away resulting in a less polluted period during the second week.

PM2.5 picked up again in the second half of the month though with sustained periods of relatively high pollution. PM2.5 of 50 μg/m³ corresponds to an AQI of 137, "Unhealthy for Sensitive Groups". Bearing in mind the point in the first paragraph above about the averaging on the monthly line, peak figures for pollution will have been much higher indeed. Sadly I do not have those figures available.

Weather conditions for November 2018

Weather-wise there were lower winds through the middle fortnight of the month, and lower rainfall too, however, the peak in PM2.5 near the "Week 46" label on the pollution graph corresponds to a period of higher winds on the weather graph for the same period. I find this to be unexpected; normally higher winds would tend to result in lower pollution. The monthly weather graph does not currently record wind direction though, and clearly that could influence readings.


The other factor here is the temperature. The bulge in pollution in the second half matches very closely with a dip in temperature in the same period. Is this due to boilers running for much longer to keep houses and offices warm, and resulting in such a measurable increase in pollution? I know our domestic gas use increased considerably at that point. It's not just gas boilers, wood burners have only been growing in popularity over the past fifteen years. I need to read up on how much PM2.5 pollution these sources produce.

Wind and rain at the end of the month and into the start of December will have contributed to the reducing levels of pollution, even as there was a dip in temperature in the early days of that month.