Trust and comparison
Why air-quality apps disagree — and how to compare them fairly
Two air-quality apps can show different numbers without either screen being fabricated. Before judging the difference, align the index scheme, pollutant, valid time, averaging period, location, evidence type, and update cycle.
Compare the metadata
45
European AQI · model · 14:00
90
U.S. AQI · station · 13:00
Numbers without context are not equivalents
Start with the scale name
European AQI and U.S. AQI use different category boundaries and labels. Other services may use national, provider, or proprietary scales, so a numeral without its scale is incomplete.
AirAgora always labels its live value as the OpenWeather air-pollution scale from 1 to 5. It never presents that provider scale as U.S., European, or local regulatory AQI.
Align the timestamp and averaging period
One value may be current, another forecast, and another the latest available station average. Their visible clock times can refer to issue time, valid time, refresh time, or local time. Pollutant indexes can also use different averaging conventions.
A fair comparison records the timezone and asks what the number represents. AirAgora shows provider update time and local forecast hours separately. It does not compare a screenshot from yesterday with a newly issued model and call the gap an error.
Modeled grids and point monitors answer different questions
OpenWeather provides modeled forecasts as a broad forecast grid. A monitoring station samples one site. A phone sensor, low-cost device, or interpolated map may add another method.
The model can provide future coverage; the monitor can describe its point more directly. A road, hill, coastline, indoor space, or neighborhood can differ from a grid cell. The provenance should be visible before deciding which source is more relevant.
Location selection can shift the answer
Apps may snap a postcode to different coordinates, choose another city centroid, interpolate from a station, or use the device location. Even a small coordinate change can cross grid or administrative boundaries. Timezone selection can also change displayed hours.
AirAgora’s city directory uses the published GeoNames coordinates, while broader searches open the live tool without creating a permanent city page. The result identifies the selected city so an ambiguous place name is not silently resolved.
Refresh cadence, caching, and stale data
A provider model updates on its own schedule. An application can cache responses to protect latency and rate limits. During an outage, one service may show a timestamped stale result while another shows unavailable and another may already have a new run.
AirAgora considers air-quality cache fresh for a bounded period and can retain a stale entry only temporarily after an error. The original timestamp and stale state remain visible. It never relabels cached data as newly observed.
- Scheme: European, U.S., national, or proprietary?
- Data: modeled forecast, station observation, interpolation, or sensor?
- Time: issue time, valid time, refresh time, and timezone?
- Place: exact station, grid cell, centroid, or device coordinates?
- Pollutant: aggregate index or one component?
When disagreement should change your decision
If official local warnings report a severe event, those instructions should take priority over a consumer forecast comparison. If the difference is modest, it may be more useful to focus on whether both services show a broadly stable or changing day.
AirAgora’s goal is not to declare every other screen wrong. It offers a consistent method, named source, visible limitations, and honest unavailable states. Those details let a user evaluate the claim instead of trusting a polished number without context.
Direct answers
Frequently asked questions
Why does one app say 45 and another 90?
They may use different OpenWeather scales, pollutants, averaging periods, timestamps, locations, or evidence types. Compare the metadata before comparing numerals.
Is a station always better than a model?
A station measures one site; a model covers an area and future hours. Which is more useful depends on the question, siting, quality control, and timing.
Can cached data cause disagreement?
Yes. Services may refresh at different times or temporarily show stale data. A responsible service labels the original timestamp and stale state.
Does language explain different scale values?
It should not. AirAgora locale changes translated labels while the live provider scale remains OpenWeather 1–5.
How can I compare two apps quickly?
Check scheme, source type, valid time, timezone, location, aggregate versus pollutant value, and update time. If those differ, the numbers are not like-for-like.
Which app should I follow during an official warning?
Follow the relevant public authority and official warning. Consumer planning tools should not override urgent local guidance.
Evidence register
Sources and further reading
The links below identify the primary references used on this page. Selection, summary, and presentation are AirAgora modifications.
Keep exploring
AirAgora provides informational planning guidance, not medical advice. Local conditions can differ from modeled forecasts; follow official guidance during severe events.