# Field digest — the air you breathe, measured by your neighbors 2026-09-26 · ebungo field research · ten sources · twenty-three verbatim quotes The densest air-quality map on Earth is not built by governments. It is built out of small consumer laser particle counters hanging on fences, balconies, and school walls — owned by the people who live under the readings. This digest follows the chain the other way: how a Utah hobbyist's dust problem became a 35,000-sensor community network, how a nonprofit turned that noise into an open, harmonized data commons, how EPA scientists wrote the corrections that make consumer gadgetry comparable to lab instruments, and why the result now sits inside official wildfire response maps.[4][2][6][5] ## The stakes, in WHO's numbers Ambient (outdoor) air pollution was estimated to cause 4.2 million premature deaths worldwide per year in 2019 — mortality from fine particulate matter driving cardiovascular and respiratory disease, and cancers.[11] The broader figure that WHO repeats for air pollution in all settings is 7 million premature deaths every year.[12] Two figures, two scopes: the smaller counts outdoor air only; the larger counts the same burden with household air pollution added. WHO does not merge them, and neither does this digest.[11][12] The exposure behind those deaths is nearly universal. In 2019, more than 90% of the global population lived in areas where concentrations exceeded the 2005 WHO guideline for long-term exposure to PM2.5.[12] In September 2021 WHO updated the guidelines and adjusted almost all the levels downwards, cutting the annual PM2.5 guideline from 10 to 5 micrograms per cubic meter; its rapid scenario analysis estimated that almost 80% of deaths related to PM2.5 could be avoided worldwide if current pollution levels were reduced to the updated guideline.[12] Under a threat that touches nearly everyone, knowing the local reading becomes a public service — and that is where the citizens' grid enters.[12] ## The gadget: from a gravel pit to 35,000 sensors PurpleAir began in 2015, when a Utah engineer concerned about dust from a nearby gravel pit built his own air quality sensor; the closest government monitors were miles away and said nothing about his backyard.[4] He built a low-cost, Wi-Fi-enabled sensor that could measure particulate matter in real time and display the data online, with dual laser particle counters for accuracy and redundancy; by late 2015 he had built 80 sensors and given them away around the Salt Lake Valley.[4] What started as a personal project became, in the company's own words, "the world's largest network of community air quality monitors": the network now spans 100+ countries with 35,000+ sensors worldwide.[4] On the map, the density is the point: air quality data updates every few minutes, and with sensors on nearly every block in many areas the map shows variations within a single neighborhood that a handful of government stations could never resolve.[5] Municipalities, school districts, tribal nations, and health departments now deploy their own sub-networks of the same hardware.[4] ## The commons: OpenAQ makes the noise legible The raw flood of community sensors would stay siloed per vendor if not for OpenAQ — an environmental tech nonprofit that aggregates and harmonizes open air quality data from across the world onto an open-source, open-access data platform.[2] The platform mixes reference-grade government monitors and low-cost sensors from research institutions, community organizations, companies, and individuals, and shares it all for free.[3] The scale is a commons in action: since 2022 OpenAQ has added 43 new countries, reaching 141 countries and territories; it maintained a net increase of more than 7,100 active locations from 2021 to the end of 2025, with over 15,300 actively monitored locations; active and historical data combined exceeds 2 billion measurements.[2] Growth in 2025 skewed deliberately toward the global south: over 3,500 new or recovered locations added 86 million measurements, and approximately 42% of those locations span 50 low- and middle-income countries — 94 locations across 9 new countries where OpenAQ previously hosted no data, all of them World Bank-designated LMICs.[1] as-of note, 2026-09-26: probing OpenAQ's public API during this run — the legacy v1 and v2 endpoints now answer HTTP 410 "Gone", and version 3 requires an X-API-Key header. The open platform remains open by policy; the programmatic door now expects a key. ## The science: corrections are the real instrument A consumer particle counter is not, by itself, a measurement. As an Atmospheric Measurement Techniques evaluation put it: PurpleAir sensors are low-cost tools to measure fine particulate matter concentrations, but the raw data have significant biases, so the sensors must be corrected.[8] The correction pipeline is government science: EPA built a US-wide correction in 2019 — the widely quoted form being PM2.5 = 0.52 × PA_cf1 − 0.086 × RH + 5.75 — developed collocations against Federal Reference and Equivalent Method monitors across 16 states, with data cleaning steps based on the sensors' duplicate A and B channels.[6] The corrected data then crossed into officialdom. In August 2020, a new layer of corrected, publicly available PM2.5 data from PurpleAir sensors was added to EPA's AirNow Fire and Smoke Map — and that integration persists today: PurpleAir sensor data is integrated into the official map at fire.airnow.gov, supplementing government monitoring stations with community-powered data during wildfire events.[6][5] When heavy wildfire smoke is present, PurpleAir sensors tend to overestimate pollution levels, and the EPA conversion exists precisely to solve that.[13] The evaluation literature keeps the score honest. Under the new EPA equation, slopes against regulatory monitors were 0.95 for urban events and 0.88 for smoke events — a slight underestimate, especially in smoke.[8] The acknowledged weak spot is dust: in dust events, corrected PurpleAir data underestimate true PM2.5 by a factor of 5–6 regardless of which correction equation is used.[8] The paper's summary is the fair verdict: the sensors can give accurate PM2.5 data in urban pollution and smoke, but more work is needed to develop an improved correction for dust aerosols.[8] ## What the grid bought The combined system — consumer hardware, corrections, open aggregation — gives a city something it did not have: a hyperlocal, near-real-time picture of the air, maintained by the people who breathe it. Schools and districts use the map for outdoor recess and HVAC decisions; emergency responders track wildfire smoke plumes as they unfold; families decide when to open the windows.[5] The WHO guideline gives the denominator — the exposure that kills 4.2 million people a year outside — and the citizen grid gives the numerator a street address.[11][5] The chain runs deeper than it looks: a hobbyist's dust complaint, a nonprofit's data commons, a government calibration equation, and a smoke map that now leans on neighbors' sensors during the fires that the map exists to track. The atmosphere got a measuring instrument built from its own residents.[4][2][6][12] ## Sources [1] [OpenAQ 2025 Year in Data](https://openaq.medium.com/2025-year-in-data-59d62a4f31cd)"In 2025, OpenAQ introduced over 3,500 recovered/new locations, which in turn added over 86 million measurements, historical and active ingestion included." "With approximately 42% of new/recovered locations spanning 50 LMICs, 2025 was a great conclusion to a strategic period of remarkable progress in open air quality data in the world, particularly in the global south." [2] [Clarity: OpenAQ 2025 Year in Data](https://www.clarity.io/blog/openaqs-2025-year-in-data-report-rapid-growth-in-access-to-open-air-quality-data-globally)"OpenAQ is an environmental tech nonprofit that aggregates and harmonizes open air quality data from across the world onto an open-source, open-access data platform." "Since 2022, OpenAQ has added a total of 43 new countries, raising its current global coverage to 141 countries and territories across the globe." "OpenAQ has been able to maintain a net increase of more than 7,100 active locations from 2021 to the end of 2025, with over 15,300 actively monitored locations." "Active and historical data combined means that there are more than 2 billion measurements that users can now tap into using the OpenAQ platform." [3] [OpenAQ Data Platform](https://openaq.org/about/initiatives/openaq-data-platform/)"We actively seek data produced by reference-grade monitors and air sensors (aka, low-cost sensors) run by governments, research institutions, community organizations, companies and individuals across the globe." [4] [PurpleAir About](https://www.purpleair.com/about)"In 2015, a Utah engineer concerned about dust from a nearby gravel pit built his own air quality sensor. That personal project became PurpleAir - now the world's largest network of community air quality monitors." "35,000+ Sensors Worldwide" "The network spans 100+ countries." [5] [PurpleAir About the Map](https://www.purpleair.com/real-time-map)"With sensors on nearly every block in many areas, see air quality variations within your own neighborhood." "PurpleAir sensor data is integrated into the official EPA Fire and Smoke Map at fire.airnow.gov, supplementing government monitoring stations with community-powered data during wildfire events." [6] [EPA AirNow Fire and Smoke Map webinar](https://www.epa.gov/sites/default/files/2021-05/documents/toolsresourceswebinar_purpleairsmoke_210519b.pdf)"In August 2020, a new layer of corrected, publicly available PM 2.5 data from PurpleAir sensors was added to the map" "We built a US-wide correction in 2019" [8] [AMT 2023 evaluation of EPA correction](https://amt.copernicus.org/articles/16/1311/2023)"PurpleAir sensors (PASs) are low-cost tools to measure fine particulate matter (PM) concentrations. However, the raw PAS data have significant biases, so the sensors must be corrected." "Using the new EPA equation, we find slopes of 0.95 and 0.88 for urban and smoke events, respectively, indicating a slight underestimate in PM 2.5 using this equation, especially for smoke events." "For dust events, while the PAS and regulatory data still show significant correlations, the PAS data using either correction equation underestimate the true PM 2.5 by a factor of 5–6." [11] [WHO ambient (outdoor) air quality fact sheet](https://www.who.int/news-room/fact-sheets/detail/ambient-%28outdoor%29-air-quality-and-health)"Ambient (outdoor) air pollution in both cities and rural areas was estimated to cause 4.2 million premature deaths worldwide per year in 2019; this mortality is due to exposure to fine particulate matter, which causes cardiovascular and respiratory disease, and cancers." [12] [WHO 2021 Global Air Quality Guidelines announcement](https://www.who.int/news/item/22-09-2021-new-who-global-air-quality-guidelines-aim-to-save-millions-of-lives-from-air-pollution)"Every year, exposure to air pollution is estimated to cause 7 million premature deaths and result in the loss of millions more healthy years of life." "In 2019, more than 90% of the global population lived in areas where concentrations exceeded the 2005 WHO air quality guideline for long term exposure to PM₂.₅." "Almost 80% of deaths related to PM₂.₅ could be avoided in the world if the current air pollution levels were reduced to those proposed in the updated guideline" [13] [PurpleAir Community: New EPA correction](https://community.purpleair.com/t/new-epa-correction/10372)"PurpleAir sensors tend to overestimate pollution levels when heavy wildfire smoke is present––the EPA conversion solves this issue." "CF1 data (and thereby the EPA conversion for CF1 data) can be used outdoors." ten sources · twenty-three verbatim quotes · fetched live 2026-09-26 · ebungo field research --- Source: