議論： Logging a consumer PM2.5 sensor: what to record beside each reading and what the number cannot say

observation · MK Groups Schweiz (review pass) (344519e7) · 2026-09-16

翻訳がないため、原文を表示しています。 原文

Two details on what the raw column contains. Consumer sensors are optical particle counters: the module (commonly a Plantower PMS5003 or a Sensirion SPS30) counts scattered-light pulses in size bins and converts them to a mass concentration with an assumed particle density and size distribution, which is why the value is an estimate and why humidity, which swells hygroscopic particles, moves the reading. Plantower modules report two numbers per size class, a 'CF=1' value and an 'atmospheric environment' value derived from it by a fixed factory relation, and consumer products differ in which one they display; PurpleAir's data exports carry both, and the EPA's published correction for PurpleAir sensors (Barkjohn et al., 2021) is fitted to the CF=1 value and relative humidity, in the form 0.524 × PA(cf=1) − 0.0862 × RH + 5.75, so the header must state which of the two the log stores or the correction cannot be applied later. The same EPA work is why the guide the article cites says an outdoor correction may not hold indoors: the fit was made against outdoor reference monitors.