Second-hand smoke monitoring report

What the pattern shows

Across 31 monitored days, the equivalent of 10 days were free of regional haze; in that time the sensor recorded 145 separate local smoke episodes, about 14.6 a day, each lasting a median of 13 minutes. Every one of the 9 days that were mostly free of regional haze had at least one episode. They follow a daily clock: the first episode typically began around 06:05 and the last around 22:57, with 27% of episodes starting between 6 and 10 AM and 37% between 5 and 10 PM. Weekdays averaged 15.8 episodes a day and weekends 9.8. 22 days were dominated by regional wildfire smoke and are shown separately; the local episodes are short bursts that rise and fall within minutes, unlike regional smoke, which lifts the whole day. Short, repeated bursts on a daily schedule at a fixed location are the signature of a nearby, deliberate source rather than weather or traffic; the reporter's own observation identifies it as tobacco smoke from an adjacent property.

Calendar: every half hour of the period

Time of day

Share of monitored days (regional-haze days excluded) with a local smoke episode in progress during each hour.

Minutes of local smoke per day

Method and limitations

• Instrument. A Plantower PMS5003 laser particle counter, fixed outdoors in the rear yard, sampling every second and averaged over 30 seconds; a Bosch BME688 gas sensor beside it provides an independent smell cross-check (it independently confirmed 15 of the 145 episodes).
• What counts as an episode. Fine-particle (PM2.5) readings jumping above the rolling 30-minute background and staying up for most of a 20-second window, through to the air returning to background for three continuous minutes. Brief dips inside an episode do not end it. Detections whose peak never reached 10 µg/m³ are treated as faint and are not counted as episodes (60 in this period).
• Regional smoke is separated. When the background itself stays elevated for 30 minutes or more (wildfire smoke, fog) the system marks the span as regional; those hours are hatched in the calendar and left out of the local totals. Where available, the official regional index is shown for comparison.
• Limitations. This is a consumer-grade sensor: it shows presence, timing and relative intensity, not regulatory concentrations. The 30-second peaks reported here are not directly comparable with 24-hour guideline values. The instrument does not identify the source of the smoke; that identification rests on the reporter's direct observation.
• Calibration. Detection thresholds were finalised on 18 July 2026; readings before then were recorded with an earlier rule and are noisier.

References

• Health Canada, Second-hand smoke: there is no safe level of exposure to second-hand smoke. canada.ca/en/health-canada/services/smoking-tobacco/second-hand-smoke.html
• World Health Organization, Global Air Quality Guidelines (2021): PM2.5 24-hour guideline 15 µg/m³. who.int/publications/i/item/9789240034228
• British Columbia Air Quality Objective for PM2.5: 25 µg/m³ over 24 hours. www2.gov.bc.ca (Air quality objectives and standards)

Appendix: episode log

Every local episode in the period. Peak is the highest 10-second reading during the episode; background is the typical reading in the ten minutes before it; smell marks episodes the gas sensor confirmed independently. A further 906 detections that began during regional wildfire smoke are not listed. 60 faint detections (peak under 10 µg/m³) are not listed.