# About BreatheCheck

> Source: https://breathecheck.com/about/
> Three data sources (AirNow, Open-Meteo, PurpleAir), no named founder, no lab, no product tests — what this project does and refuses to do.

BreatheCheck pulls air-quality data from three sources: EPA AirNow, Open-Meteo and PurpleAir. It reconciles all three into one number per city. It stores its own history, starting the day each city page went live. Nobody runs these numbers through a lab. Nobody here operates a sensor.

There's no bio photo on this page, and no named founder. That's on purpose. A verifiable pipeline is a stronger trust signal in this vertical than an invented credential. A fabricated "Dr." headshot is worse than having none at all. What follows is the honest version of an about page: what the project does, what it refuses to do, and where it still falls short.

## What this project actually does

BreatheCheck ingests EPA AirNow, Open-Meteo and PurpleAir readings into its own database. It keeps its own city-level history — the reason a 30-day chart, or a "days above AQI 100" count, can exist on a city page at all. [Full methodology](/methodology/) covers how the four sources get reconciled, and the five specific ways a reading can disagree with what you're seeing outside.

Every threshold quoted on this site is attributed to whichever body set it. "EPA classifies 101–150 as Unhealthy for Sensitive Groups" is an EPA classification, cited as one. It's never rewritten into "AQI 120 is dangerous for you." A number can also reflect community-reported behavior instead of an agency standard: a runner's cutoff, a parent's, one asthmatic's own 156-run log. We label that kind of number as that, every time. It's a different kind of number, and it earns a different kind of trust.

## What this project doesn't do

We don't operate air-quality sensors. Every reading traces back to AirNow, Open-Meteo or PurpleAir, credited by name. Never to hardware we own.

We don't test air purifiers. The recommendation inside a city page's verdict block comes from what residents and independent HVAC guidance report, not from a lab we run. The [affiliate disclosure](/affiliate-disclosure/) explains exactly how that recommendation gets built, and how it's paid for.

We don't sell air purifiers, either. IQAir, the market leader whose city pages hold the top spot on most of these searches, also manufactures air purifiers. That's a real conflict between the number it shows you and the product it wants you to buy next.We don't manufacture or sell hardware, and nothing here is ranked to move units.

We don't write guidance for children specifically. A daycare or school threshold, where a state has published one, is attributed to that state or agency. It's never asserted as our own recommendation. Where no state rule exists, the relevant page says so instead of guessing at one.

## Editorial standards, stated plainly

Attribute every threshold. Assert none as a verdict on your own body. Never claim to have tested a product. Never publish pediatric guidance as our own. Correct errors in public, on a dated log, rather than editing them away quietly. And when the pipeline itself goes dark — a source times out, a reading turns too old to trust — the page says "we don't know right now." Never yesterday's number, dressed up as current.

## Scope

BreatheCheck covers 28 US metros, on US federal data (AirNow) plus international feeds (Open-Meteo, PurpleAir) that cover US ground too. PurpleAir's sensor network runs denser in some metros than others. A city page says so where it's thin, rather than implying even coverage everywhere.

## When we get it wrong

A station distance will be off at some point. A threshold will get mislabeled. Readers who check our numbers against their own gear will catch some of this before we do. That's the expected outcome of publishing to an audience that owns its own instruments, not a failure state. Every dated correction goes on the public [corrections log](/corrections/), permanently. A corrected page carries a link back to it.

Found a number that doesn't match what you're seeing, or a mistake in how something's described? The corrections page is the route for that. There's no separate inbox on this site. Read [what we measure and why it can disagree with your senses](/methodology/) for the mechanics behind most mismatches people report. See [the privacy policy](/privacy/) for what little data the site collects along the way.