The indoor/outdoor ratio: one number that tells you what the envelope is doing

The math is simple. For any pollutant the device reads indoors and the outdoor feed reports for your location, divide indoor by outdoor on the same time window. A 24-hour average is best for envelope-baseline questions; a 1-hour average is better for active events like a wildfire-smoke push or an ozone afternoon. The ratio is unitless, and that is the point: it removes the absolute scale of the event and isolates what your building is doing about it.

Typical bands for PM_2.5, the most diagnostic species. Below 0.3 means a tight envelope with active filtration: HEPA running, MERV-13 in the HVAC, doors and windows closed, no significant indoor sources. 0.3 to 0.6 is moderate: either decent envelope without portable HEPA, or HVAC running on its stock filter with normal infiltration. 0.6 to 0.9 is cracked windows or a leaky older home with no filtration assist. Near 1.0 means windows are open, or the envelope is so leaky it might as well be. Above 1.0 is the alarming case: indoor concentration exceeds outdoor, which means there is a source inside. Investigate.

Per-pollutant subtleties matter. PM ratios are cleanest because particles do not chemically transform between outside and inside. Ozone is different: typical indoor/outdoor ozone ratios sit at 0.2 to 0.4 even with windows closed, because ozone reacts with indoor surfaces (skin, fabrics, paint) within minutes. CO₂ inverts the whole framework: outdoor is near 420 ppm and indoor is almost always higher because people exhale; the diagnostic question for CO₂ is "how much higher than outdoor," not "what ratio." NO₂ ratios are dominated by indoor combustion (gas stoves) when present; without an indoor source, the ratio runs at roughly 0.5 to 0.8 in homes near traffic.

Limits to keep in mind. The ratio assumes both readings are valid; a drifting sensor, a humidity-confused PM count, or a stale outdoor feed will all break it. The nearest outdoor monitor may be 5 to 30 km away, so during a hyperlocal event the ratio can mislead. And short averaging windows are noisy: a single passing truck or a one-minute candle burst can swing the 5-minute ratio dramatically. Use the ratio as a slow signal, average over hours, and pair it with co-movement patterns to confirm what is driving the number.