Air Quality Research Platform

PurpleAir data, analysed and documented

Upload a sensor export and get EPA-corrected PM2.5, dual-channel validation, and downloadable reports. Every number shows the method behind it — the correction applied, the sensor's health, the data completeness, and the uncertainty.

Data Import

Report details — optional

Optional — appears at the top of the PDF report.

PurpleAir exports timestamps in UTC by default — pick your local timezone to convert. If you exported the file with timestamps already in your local time, choose No conversion so times are used exactly as recorded (selecting a timezone would shift them twice) — you can then name the timezone above so charts and reports state it correctly.

Upload data

Drop a PurpleAir CSV or XLSX file here

or

Don't have a file?

No file yet? In the PurpleAir dashboard, open your sensor → Download Data → export a CSV or XLSX (0-, 2-, or 10-minute average — all work).

One file per analysis. Dual-sensor setups are detected automatically. To compare multiple houses, run one analysis first, then use the Compare Houses panel.

View
The headline results, charts and reports. Switch to Advanced for statistical tests, correction comparisons and methods.

Overview

Analysis Summary

Refine Analysis by Timeframe

Select a specific date range to regenerate reports and visualizations for that period.

Data Completeness & Inputs

What the uploaded file contains, and the impact of anything important that is missing.

Analysis Results

PM2.5 Daily Mean — Calendar (WHO 15 / EPA 35)

PM2.5 Temporal Trend

PM2.5 Concentration Distribution

Diurnal Cycle — Mean & 10th–90th Percentile

Rolling Medians (1h · 24h · 7d)

Dual Sensor Agreement

Sensor Drift Detection

PM2.5 Correction Formula Comparison

Compare how three correction equations translate raw sensor readings. The main analysis uses EPA Barkjohn.

EPA Barkjohn (2021) PM2.5 = 0.524 × raw − 0.0862 × RH + 5.75 Standard for PurpleAir sensors; humidity-corrected. Used for all main analysis.
LRAPA PM2.5 = 0.5 × raw − 0.66 Lane Regional Air Protection Agency. Simple linear, no RH correction.
AQ&U PM2.5 = 0.778 × raw + 2.65 University of Utah / AQ&U network correction.

Correction Formula Comparison

Statistics and events

Descriptive statistics

Exceedance summary

Highest Pollution Events (Top 10 by Peak PM2.5)

Events ranked by highest recorded PM2.5 concentration, with timestamp and duration.

Data anomalies

    Custom Notes for Report

    Add notes, observations, or commentary here. Each note will appear as a dedicated section in the generated PDF report.

    Reports and Data

    Research Report

    Complete analysis with full methodology, statistical analysis, and quality assessment. Designed to support research publications and regulatory submissions.

    Data Exports

    CSV and JSON formatted data products for further analysis, including quality flags, EPA corrections, and statistical summaries.

    Chart Downloads

    Export individual visualizations as high-resolution PNG files for presentations and publications.

    Methods & Standards

    Every metric on this platform is traceable to a published method or an official health standard. These are the primary sources — citable directly in research, regulatory, and community documentation.

    WHO 2021 — 24-hour PM2.5 15 µg/m³ Air Quality Guideline (health-based)
    WHO 2021 — annual PM2.5 5 µg/m³ Long-term exposure guideline
    US EPA — 24-hour NAAQS 35 µg/m³ Regulatory standard
    US EPA — annual NAAQS (2024) 9.0 µg/m³ Strengthened from 12.0 in 2024
    1. Barkjohn, K.K., Gantt, B., & Clements, A.L. (2021). Development and application of a United States-wide correction for PM2.5 data collected with the PurpleAir sensor. Atmospheric Measurement Techniques, 14(6), 4617–4637. doi:10.5194/amt-14-4617-2021 Used for the default humidity-corrected PM2.5 equation.
    2. World Health Organization (2021). WHO global air quality guidelines: particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide. Geneva: WHO. who.int Source of the WHO 15 µg/m³ health threshold.
    3. U.S. Environmental Protection Agency (2024). National Ambient Air Quality Standards (NAAQS) for Particulate Matter — Final Reconsideration. epa.gov Source of the EPA 35 µg/m³ (24-h) and 9.0 µg/m³ (annual) standards.
    4. Cleveland, R.B., Cleveland, W.S., McRae, J.E., & Terpenning, I. (1990). STL: A Seasonal-Trend Decomposition Procedure Based on Loess. Journal of Official Statistics, 6(1), 3–73. Basis of the trend / seasonality / residual decomposition.
    5. Lane Regional Air Protection Agency (LRAPA). PurpleAir correction factor.  ·  Sayahi, T., Butterfield, A., & Kelly, K.E. (AQ&U, Univ. of Utah). Low-cost sensor calibration. Alternative correction formulas offered in the comparison panel.
    6. U.S. EPA (2021). AirNow Fire and Smoke Map — PurpleAir sensor correction methodology. airnow.gov Reference implementation of sensor correction for public reporting.

    Air quality categories on this platform are expressed against health-based PM2.5 mass thresholds (WHO / EPA) rather than a single AQI index, to preserve the underlying concentration and its uncertainty. Where an "AQI" value is shown it is the PM2.5 sub-index only: the official U.S. AQI is the maximum of several pollutant sub-indices (PM2.5, PM10, O3, CO, SO2, NO2), so the true AQI could be higher if other pollutants are elevated. Real-time AQI also uses EPA's NowCast (a weighted ~12-hour average) rather than instantaneous readings.