This experiment was - despite telling myself the previous year to "really start on time if I am going to do this again" - once again hastily put together on December 31st. I had prepared nothing; the Raspberry Pi 3 model A+ was collecting dust from my previous experiment which involved CO2 measurements in my room, of which I never published data, which I should probably do. Anyways, because during the previous time the whole setup fell on the ground and broke the SD card, I didn't have it ready. However, because I did program a similar program for CO2, I quickly patched it together, essentially in the same fashion as previously, with the same libraries. The hardware setup is completely the same as when I did the reference measurements, but differs slightly from the previous new year as it wasn't a lunchbox but a tube. Despite starting development at 21:40, there is even an improvement: sampling is now at 1/2 Hz and is accumulated and averaged to the minute, instead of just getting plain values every 30 seconds or 1/30 Hz. Due to an oversight there is now a data point every 60 seconds rather than 30 seconds, however.
This year, there is a difference however: Fireworks are banned for this New Year's Eve.rijksoverheid Yep that's right! but I thought it would still be nice to see the values for New Year's Eve.
Note of difference in the graph: this year I do not publish PM10. The spread was tiny, and there's new information about the sensor(Sensirion SPS30): the sensor seems pathethic at measuring PM10, but seems decent at PM2.5, and is relatively better at high humidity in the air compared to other optical sensors.samenmetenaanluchtkwaliteit, aqmd to note is that on overall it slightly overestimates concentrations, and a high RH further contributes to that.aqmd I can't complain given the price. Anyways, keep in mind that all these values are purely indicative, and likely not accurate at all. But, they're accurate enough to have an idea of the particulate matter situation.
|WHO guideline max values|
|Annual mean||20 μg/m3||10 μg/m3|
|24-hour mean||50 μg/m3||25 μg/m3|
Weather: some minor drizzle now and then, slightly misty maybe(sight at least 1 km), high humidity: around 95% RH according to a local weather station). Temperature at 00:00: around 3 °C. Barely any wind, fluctuating from north to west.
Despite there was a complete ban on all fireworks(except Cat. F1 fireworks, which 12-year-olds are allowed to set off), some people still set some off. At daytime, it was almost completely an ordinary day, only sometimes a distance bang or flute. At 21:00 there was gradually more and more, but unfortunately I do not have any data because I started so late, which I deeply regret now. A little after 00:00 the most was set off. It was considerably less than previous years however. I did not know what to expect. There was also more police around than previous years. I only saw one major thing being set off within 100 meter radius or so. I did see some smoke because of children setting of large amounts of F1 fireworks at the same time, this might where the major peak comes from. Overall I did not know what to expect, and it makes sense there was a group that kept setting it off anyways. For the first hour after 00:00, you see that the "base concentration" is actually not that high.
This year you also see what you saw the previous year: it takes some time before the values systematically rise. I do not have any explanation for this, it baffles me a little. Because of the low wind speed, and because I live relatively east of the country where there is no "fresh" air from the sea(example), it took long times before the high concentration became lower. This data is consistent with official sensors near my area.
All in all the data exceeded my expectations, I thought it would be a lot more "boring" and as usual. I wish I had started earlier. Improvements for next year: CO2, humidity and temperature readings, and gathering a full day instead of just a few hours before 00:00.
There seems considerably less PM in the air than previous New Year's Eve, however, values still seem above normal
link to previous year - reference values/control group whatever