The measurement gap that residents filled
Poland's state air monitoring network — operated by the Chief Inspectorate of Environmental Protection (GIOŚ) — maintains roughly 270 reference stations across the country. For a nation of 38 million people spread across 312,000 km², that works out to one station per 1,155 km². In practice, many of these stations sit in administrative centres, leaving residential neighbourhoods, suburban fringes, and smaller towns without reliable readings.
The shortfall became a civic issue around 2016, when smog episodes in Kraków and the Silesian agglomeration attracted national media attention. Rather than waiting for regulatory expansion, groups of residents began buying inexpensive optical particle counters — primarily SDS011 and Plantower PMS5003 sensors — and connecting them to microcontrollers with Wi-Fi modules. The data was pushed to shared servers and made publicly visible on web maps.
Sensor.Community: the backbone of Polish community air data
The most widely used data aggregation point in Poland is Sensor.Community (formerly Luftdaten.info), a German-origin open-hardware project that Poland adopted faster than almost any other country outside Germany. As of early 2026, Poland accounts for more than 4,000 active nodes — the largest national contribution to the global Sensor.Community map.
Each node uploads PM2.5, PM10, temperature, and humidity readings every 145 seconds. The data streams are publicly downloadable from the Sensor.Community archive under an open data licence, with records dating back to 2017 for many Polish cities. Researchers at the AGH University of Science and Technology in Kraków have published peer-reviewed studies cross-referencing Sensor.Community data against reference measurements from GIOŚ stations, finding mean absolute errors below 15 µg/m³ for PM2.5 in most urban conditions.
How a typical community sensor node is set up
A standard citizen-built node consists of three components: an optical particle counter housed in a corrugated pipe or commercially produced weatherproof casing, a NodeMCU or D1 Mini microcontroller running custom firmware, and a power supply — typically a USB adapter or, in some rural deployments, a small solar panel. Assembly takes two to three hours for someone with basic electronics experience. Total hardware costs range from 80 to 160 PLN depending on supplier and sensor model.
Nodes are typically mounted on balconies, exterior walls, or fence posts, 2–5 metres above ground level. Sensor.Community's firmware documentation recommends a north- or east-facing exposure away from direct sunlight to reduce thermal drift in humidity readings, which affect particle count accuracy in some sensor designs.
What the distributed data reveals
Aggregated maps produced from community sensor data consistently show spatial variation in PM2.5 that official monitoring cannot capture. In Warsaw's Praga district, for instance, a cluster of 40+ Sensor.Community nodes recorded seasonal PM2.5 averages 30–45% higher than the nearest GIOŚ reference station 2.8 km away in a park setting — a difference attributable to residential coal and wood heating in older tenement buildings.
Similar hyper-local patterns have been documented in Łódź, Wrocław, and the Trójmiasto conurbation. In each case, the community data identified street-level pollution gradients invisible in official monitoring reports, prompting local councils in some districts to accelerate thermomodernisation subsidy programmes.
Seasonal burning and heating season dynamics
Poland's pollution calendar shows a marked transition between October and March, when residential solid fuel heating drives PM2.5 concentrations to levels that regularly exceed WHO annual guideline values (5 µg/m³) and the EU's own 25 µg/m³ standard. Community sensor data visualised hour by hour makes this phenomenon tangible to non-specialist audiences: peaks typically appear between 6–9 AM and 18–22 PM, matching household heating cycles, and are strongly correlated with wind speed — a relationship clearly visible in the raw data streams.
Data quality considerations
Community-grade optical sensors are not reference instruments. They cannot distinguish particle composition, and their readings are sensitive to humidity — above 70% relative humidity, optical scattering from condensed water can inflate PM readings by 30–50% in some sensor models. Several Polish academic groups have developed humidity correction algorithms specifically calibrated against GIOŚ reference data, and these are increasingly applied as post-processing steps by data aggregators.
Users interpreting community air maps should treat readings as indicative of broad spatial and temporal patterns rather than regulatory-grade concentration values. For health decisions and policy, official GIOŚ data and GIOŚ-certified laboratories remain the authoritative source.
Related coverage
- Biodiversity Surveys in Polish Forests: What Volunteers Document
- Open Data Weather Networks in Poland: How They Work