Skip to content

Hourly forecast reading guide

How to Read an Hourly Air-Quality Forecast

An hourly air-quality forecast is easiest to read when place, valid hour, freshness context, provider scale, pollutant fields, and data limits stay in a fixed order. Read an hourly air-quality forecast in this order: confirm the place and local valid hour; identify the documented point time and any stale label shown by AirAgora; read the official OpenWeather air-pollution scale from 1 to 5; inspect pollutant concentrations and the main-pollutant explanation; then compare every hour that covers the activity. AirAgora shows modeled grid estimates, not a street-level observation or a promise that one period is safe for everyone.

Person walking on a sunlit path above a city beneath translucent atmospheric arcs
An editorial illustration of reading a forecast as a sequence of sourced hours, not as a single magic number or a measurement display.

Start with the place and the hour the value represents

An hourly row is a modeled statement about a selected coordinate at one valid time. Before reading the color or category, confirm that the city or chosen location is the one you intended, that an ambiguous place name was resolved correctly, and that the displayed hour uses the location’s local timezone. A correct pollutant value attached to the wrong place or day cannot support the intended decision.

Keep the calendar date attached to the clock time, especially near midnight, while traveling, or during a daylight-saving transition. AirAgora evaluates absolute timestamps and then formats them in the location’s IANA timezone. That prevents a repeated local hour from being treated as the same instant and avoids inventing a clock hour that did not occur.

Sources[1][4][5]

Separate valid time from fetch, cache, and view time

The valid time answers “which hour does this forecast value describe?” OpenWeather documents that point time in the dt field. AirAgora records when it fetched or reused the response in technical provenance, while page-view time says only when the browser displayed it. The planner exposes the provider point time and an explicit stale label when applicable. The Air Pollution API documentation does not publish a separate model-run timestamp, so a generic “updated” label must not be assumed to mean forecast issuance.

When comparing two hours, use their valid times. When judging freshness in the planner, use the displayed provider point time and any stale label without rewriting the original timestamps. A refresh attempt does not turn an older response into a new observation or prove that the provider published a new run. The dedicated timestamp guide explains the internal fetch, cache, and view clocks in more detail.

Sources[1][4][5]

Read the provider category with its full name

The interface shows OpenWeather’s official air-pollution value from 1 to 5 with the matching category: good, fair, moderate, poor, or very poor. It may translate the category, but it does not rename that provider value as U.S. AQI, European AQI, a national regulatory index, or an AirAgora AQI. A bare numeral without the named scheme is incomplete information.

Within the same provider scale, a lower category can help describe a relatively lower modeled pollution level. It still does not establish personal exposure, indoor conditions, or universal safety. AirNow’s U.S. AQI is useful educational context for how an index communicates concern, while the WHO guidance supports cautious language rather than a claim that any category clears every person and activity.

Inspect concentrations without ranking raw numerals

OpenWeather returns modeled concentrations for carbon monoxide, nitrogen monoxide, nitrogen dioxide, ozone, sulphur dioxide, PM2.5, PM10, and ammonia, normally in micrograms per cubic metre. Those fields describe different pollutants. A concentration of 80 for one pollutant is not automatically more important than a concentration of 40 for another because the physical quantities require pollutant-specific context.

AirAgora derives continuous severity separately from OpenWeather’s documented concentration bands for six supported pollutants: PM2.5, PM10, nitrogen dioxide, ozone, sulphur dioxide, and carbon monoxide. The main-pollutant label comes from those compatible severities, not the largest raw concentration numeral. Nitrogen monoxide and ammonia may remain in the provider-backed response, but AirAgora neither displays nor ranks them because it does not invent bands for unsupported components.

Sources[1][4][5]

Read a sequence that covers the activity, not one attractive point

A real outing spans time. One low-looking hour can be followed by a deterioration, so AirAgora evaluates complete consecutive two-hour windows rather than choosing a single isolated point. Both hours need valid component data. The deterministic score gives 70 percent weight to mean severity and 30 percent to the worst hour, keeping a short increase visible without letting it erase the rest of the period.

If two windows are materially tied, the earlier one remains first for stability. If the leading window improves too little over the median candidate, AirAgora says conditions are broadly similar instead of naming a falsely precise winner. If every practical window is poor, the result uses “least polluted available,” not “good,” “healthy,” or “safe.”

Sources[3][4]

Check completeness, stale state, and modeled resolution

A complete-looking screen can still contain important status information. Check whether both hours are present, whether any pollutant field needed for severity is missing, and whether the result is current, stale, or unavailable. Missing data never become zero. A bounded stale response keeps the original timestamp and explicit label rather than being presented as newly fetched or newly observed.

The forecast is modeled for a provider grid and selected coordinate. It cannot resolve every road, building, park, indoor room, monitor site, or moving plume. That limitation changes the claim: the timeline can support a broad timing comparison, but it cannot certify the air at a specific face, route, window, or street. Official local information may answer a more urgent or localized question.

Sources[1][4][5]

Turn the timeline into a bounded decision

After reading the metadata, compare the full periods that are actually practical. Prefer a meaningfully lower complete window when one exists; use logistics as the tie-break when conditions are similar; and keep caution visible when the whole range is elevated. Record the place, valid hours, displayed provider point time, any stale label, provider scale, and main pollutant if you need to discuss the result later.

The forecast should remain one input, not the only authority. Follow official warnings during severe events, and use appropriate professional guidance for personal medical questions. AirAgora does not diagnose, prescribe, estimate individual dose, or provide personal clearance. Its narrower job is to make a modeled hourly comparison understandable, sourced, and honest about what the grid cannot know.

Direct answers

Frequently asked questions

What should I read first in an hourly air-quality forecast?

Start with the selected place, local date, and valid hour. Then check the displayed provider point time, any stale label, the named OpenWeather 1–5 category, pollutant concentrations, completeness, and the sequence covering the activity.

Is the newest-looking timestamp always the forecast time?

No. A response can involve valid time, internal application fetch or cache time, and browser view time. The planner exposes the provider point time and any stale label. OpenWeather documents dt for each point but no separate air-pollution model-run timestamp, and a page refresh alone does not create one.

Can I choose the hour with the lowest raw pollutant number?

Not responsibly across different pollutants. Their concentrations require pollutant-specific bands. AirAgora compares compatible derived severities and complete two-hour windows instead of ranking unlike raw numerals.

Does OpenWeather 1 mean U.S. AQI 1?

No. The interface names OpenWeather’s official air-pollution scale from 1 to 5. It is not relabeled or numerically equated with U.S., European, or local regulatory AQI.

Why does AirAgora compare two hours?

An activity usually lasts longer than one point. Requiring two complete consecutive hours keeps a deterioration in the second hour visible and prevents an isolated low value or missing hour from controlling the decision.

Does the best-looking window mean it is safe for everyone?

No. “Best” means lower relative to the available complete modeled windows. The grid can differ locally, another official warning may apply, and AirAgora does not make personal or universal-safety determinations.

Evidence register

Sources and further reading

The links below identify the primary references used on this page. Selection, summary, and presentation are AirAgora modifications.

  1. 1.OpenWeather Air Pollution API documentation
  2. 2.U.S. EPA AirNow — AQI basics
  3. 3.WHO global air quality guidelines (2021)
  4. 4.AirAgora methodology — rule airagora-1.1.0
  5. 5.AirAgora data sources, attribution, and modifications

Keep exploring

AirAgora provides informational planning guidance, not medical advice. Local conditions can differ from modeled forecasts; follow official guidance during severe events.