Frequently Asked Questions

FEAST is designed for a wide range of users, including:

• Air district staff assessing PM_2.5 or ozone exceedances,
• Land managers and burn planners gathering data for post-fire analysis and assessment,
• State and local air districts developing Exceptional Event (EE) demonstrations,
• Researchers and community partners exploring relationships between fire activity and air quality, and
• The general public seeking to understand how prescribed fire and wildfire smoke affect local air quality.

All data and tools are publicly accessible through FEAST.

Prescribed fire reduces hazardous fuels, restores ecosystems, and can reduce the likelihood of catastrophic wildfires. FEAST helps quantify short-term smoke effects while supporting long-term forest health objectives.

FEAST provides interactive tools and visualizations that connect prescribed fire, wildfire, and air quality, supporting analyses from statewide patterns down to individual monitoring sites.

A set of tools to screen, evaluate, and document PM₂.₅ exceedances—linking concentration spikes to fire activity and assembling evidence for EE demonstrations in a structured, transparent way. The EE toolkit currently supports PM₂.₅ only; O₃ support is not yet available.

Annual and 24-hour PM₂.₅ design value maps (2012→present) to explore spatial patterns and long-term trends by region and year. The underlying data are obtained from the U.S. EPA.

A calendar view of daily PM₂.₅ for a selected site and time range. Each day is color‑coded by concentration tier to visualize exceedances and seasonality at a glance: values < 9 µg/m³ (green), 9–35 µg/m³ (yellow), and > 35 µg/m³ (red).

Highlights exceedance days and summarizes nearby fire activity using proximity rules (e.g., prescribed fires within 50 miles; wildfires within 100 miles) to assess potential influence on a site’s design value.

Creates a Word (.docx) draft for EE demonstrations, including figures, tables, design value calculations, and guidance on key content for each chapter. The document also includes a table of contents to support a clear workflow for generating reports, reducing manual formatting effort. The current template is based on the prescribed fire demonstration for Nevada County, CA, available at: https://www.epa.gov/air-quality-analysis/exceptional-events-documents-particulate-matter-nevada-county-ca

Daily exceedance reports are generated whenever either of the following conditions is met at a monitoring site:

• PM₂.₅ concentrations are ≥ 35 µg/m³, or
• Ozone concentrations are ≥ 70 ppb.

These reports are produced as preliminary, informational summaries and do not represent a regulatory determination.

The exceedance report consists of two main parts.

Part 1 – Exceedance Characterization

1. Exceedance Map
   • A map showing the locations of all monitoring sites where exceedances occurred.
   • Nearby prescribed fires within 50 miles on:
     • the day of the exceedance,
     • one day before, and
     • two days before.
   • Nearby wildfires within 100 miles.
   • Overlays of:
     • HMS-detected fire points, and
     • HMS smoke plume boundaries.
2. Site-Specific Details
   • Detailed information for each monitoring site with an exceedance, including key metadata as available (e.g., site ID, location, and date).
3. Time-Series Analysis
   • Hourly concentration time series to illustrate how the daily exceedance developed over the course of the day and the following day. The maximum hourly concentration of the day will be used as the starting hour for the backward trajectory analysis.
   • A five-year historical context, showing daily concentrations for the past five years, overlaid with the five-year average daily concentration, to help evaluate whether the exceedance appears anomalous or is consistent with typical seasonal patterns.
4. Preliminary Ranking and Interpretation
   • A preliminary analysis ranking the exceedance (e.g., relative to recent and historical values) based on the information above.
   • This analysis is intended for screening and context only.

---

Part 2 – Fire Influence and Transport Assessment

The second part of the report focuses on the potential influence of nearby fires on air quality:

1. Fire Proximity Analysis
   • Evaluation of the proximity of nearby prescribed fires and wildfires to the exceedance monitoring sites.
   • Preliminary assessment of the potential for fire-related impacts on the observed air quality.
2. Trajectory Analysis
   • Generation of air mass back-trajectory images to evaluate whether air parcels arriving at the monitoring sites may have passed over identified fires.
   • Trajectory context is provided for quick screening of possible source–receptor relationships.

---

Note: All information provided in these preliminary daily summaries is for informational and screening purposes only and does not represent a regulatory or official attainment/nonattainment determination.

Viewers and analyses combining PFIRS prescribed fires, FRAP wildfires, monitoring (AirNow/AQS/PurpleAir), satellite/HMS, observed meteorology, BlueSky, and CANSAC forecasts to understand smoke and air quality.

The Prescribed Fire View is an interactive, map-based tool that displays the geographic locations of prescribed fires recorded in PFIRS, based on user-selected date.

Each prescribed fire is represented by an icon on the map. When a user clicks on an icon, a pop-up window appears with a link that allows users to explore a comprehensive set of related datasets.

The available data include:

• PFIRS Fire Information
  • Fire name / ID
  • Location (coordinates, county, air district)
  • Burn type (e.g., pile burn, broadcast burn, understory, agricultural)
  • Responsible unit/agency
  • Planned and actual acreage
  • Planned burn dates (overall start/end dates and daily burn windows)
  • Burn status (planned, active, completed)
  • Linked smoke management permit / PFIRS ID
  • Special comments from the burner
• Satellite Imagery & Fire Detections
  • ~6-minute true-color satellite imagery animation
  • ~10-minute fire-temperature / thermal anomaly imagery animation
  • Overlays:
    • Prescribed fire location centered
    • HMS smoke plumes
    • Active fire detections
• AirNow
  • Hourly PM_2.5 and O₃
  • Within ~40 miles of the fire
  • ~±2 days around the burn
• PurpleAir
  • High-density PM_2.5 sensors
  • Within ~20 miles of the prescribed fire
  • ~±2 days around the burn
• Observed Meteorology
  • For fires > 1 year old (FAA, RAWS, CIMIS, AIRS, buoys):
    • Wind speed and direction
    • Relative humidity
    • Temperature
    • Precipitation
    • Wind roses
  • For fires ≤ 1 year old:
    • Meteorology observed from AirNow-related networks
• BlueSky Smoke Modeling
  • Forecast PM_2.5 at:
    • 100 m
    • 1,000 m
    • 4,000 m (or system-specific level)
  • Animated plume maps by height
  • Time series of PM_2.5 at each level
• CANSAC (WRF Modeling)
  • High-resolution winds, RH, PBL height, temperature, ventilation index
  • Zoomed in with prescribed fire at the center
  • Area-averaged time series for meteorological evaluation

This integrated view enables users to quickly assess prescribed fire activity, air quality impacts, meteorological conditions, and smoke behavior within a single, centralized interface.

A map-based interface displays the locations of wildfires based on the user-selected date range.

For wildfires older than one year (after CAL FIRE updates the annual FRAP fire perimeters), both the wildfire boundaries and centroid points are shown on the map. For wildfires less than one year old, only the latitude and longitude of the wildfire locations are available. A table of wildfires for the selected date are also shown below the map.

• Daily Exceedance Reports (preliminary screening)
• Historical Exceedance Summary (by day/month/year)
• Design Value Estimator (simulate removing high days to see DV impact)

A time-slider interface with ~6-minute GOES-18 true-color imagery plus HMS fire/smoke overlays to visualize smoke movement throughout the day.

CARB and DRI continually expand features and data coverage to strengthen prescribed fire planning, EE assessments, and statewide air quality management.

• PFIRS prescribed burn records — PFIRS
• CAL FIRE FRAP wildfire perimeters & points — FRAP
• Air quality:
  • AirNow (real-time PM_2.5, O₃) — AirNow Tech
  • AQS (EPA-certified; compliance & design values)
  • PurpleAir (dense PM_2.5 coverage; CARB-applied QA/QC including outlier screening, clustering-based methods, and established correction approaches)
• Observed meteorology (AirNow framework real-time; CARB historical: FAA, RAWS, CIMIS, AIRS, NOAA Buoys)
• Satellite imagery (GOES-18 true-color & fire-temp ~6-min)
• HMS fire detections & smoke plumes — NOAA HMS
• CANSAC WRF forecasts (~1.3 km; 00/12 UTC) — CANSAC
• BlueSky smoke modeling (PFIRS fires, driven by CANSAC)
• Backward trajectories (24-hr, 100/150/200 m AGL, CANSAC 1.3 km)

FEAST data are updated once per day in the evening. Because of source data availability, most datasets are available with a lag of up to three days relative to the current day. Below is the typical time resolution of the datasets included in FEAST:

• AirNow (PM_2.5, O₃): Hourly and daily
• PFIRS prescribed fire information: Daily
• FRAP wildfire information: Updated as new FRAP releases become available; FEAST checks for updates regularly.
• GOES‑18 imagery: True color ~every 6 minutes; fire temperature ~every 10 minutes
• HMS detected fire locations and smoke boundaries: Daily
• CANSAC WRF meteorological data: Daily (12 UTC runs)
• BlueSky: Daily; data may come from the 00 UTC or 12 UTC run, depending on availability (12 UTC is preferred)
• Backward trajectories: Generated for each exceedance

Near-real-time feeds are considered provisional until validated.

Satellite imagery (GOES-18 true-color & fire-temperature) provides continuous visual context (~6-min) for clouds, smoke, and hot spots.

HMS is a NOAA analyst-curated product that delineates smoke plumes and active fire detections using multiple satellite sources.

AirNow provides near-real-time observations for situational awareness (provisional).

AQS provides certified, quality-assured datasets used for compliance and design values.

CANSAC, operated by the Desert Research Institute (DRI), provides high-resolution meteorological modeling to support air quality analysis, smoke transport assessment, and atmospheric research across California and Nevada.

CANSAC runs the Weather Research and Forecasting (WRF) model using nested grids, with the finest horizontal resolution of approximately 1.3 km. This high spatial resolution enables detailed representation of complex terrain, land–sea interactions, and mesoscale circulations that are critical for accurately characterizing wildfire and prescribed-fire smoke transport.

The WRF model is initialized twice daily at 00 UTC and 12 UTC using large-scale meteorological analyses as initial and boundary conditions. Within FEAST, the 12 UTC model output is primarily used to characterize prescribed-fire conditions, as it typically captures daytime boundary-layer development and smoke transport processes.

For each prescribed fire, FEAST extracts and summarizes key meteorological variables at the fire location and surrounding area. These data are presented as spatial field animations and area-averaged time series, enabling rapid screening as well as more detailed post-event analysis. The resulting products also support long-term climatological analyses of prescribed-fire impacts on air quality and their relationships with meteorological conditions.

Key meteorological variables derived from CANSAC include:

• Wind speed and wind direction (surface and aloft)
• Temperature
• Relative humidity (RH)
• Planetary boundary layer (PBL) height
• 500-hPa geopotential height
• Ventilation index (wind speed × mixing height)
• Stability-related diagnostics (e.g., lapse rates, mixing conditions)

These CANSAC meteorological fields support BlueSky smoke modeling, providing key inputs for plume rise, transport, and dispersion calculations.

Within FEAST, CANSAC meteorological data are used to drive back-trajectory analyses, enabling evaluation of air parcel transport pathways leading to observed PM_2.5 and O₃ exceedances.

The high spatial resolution and frequent model updates improve FEAST’s ability to assess upwind source regions, plume transport consistency, and the potential influence of wildfire or prescribed-fire smoke.

CANSAC-based trajectory analyses in FEAST are intended for screening and contextual assessment only and are used in combination with satellite imagery, fire activity data, and HMS smoke products to support Exceptional Event evaluations.

BlueSky predicts smoke dispersion and PM_2.5 for PFIRS-reported burns using fire emissions and CANSAC meteorology (00/12 UTC). Outputs support pre-burn planning and post-event evaluation.

For each PM₂.₅ or O₃ exceedance, FEAST computes 24-hour back trajectories from the monitoring site at the hour of peak concentration, using CANSAC 1.3-km meteorological fields, with trajectory releases at 100 m, 150 m, and 200 m AGL to help identify upwind source regions.

Modeled plumes and HMS smoke extents indicate likely areas of transport, not exact concentrations. Use them with observations (AirNow/PurpleAir), trajectories, and meteorology to triangulate evidence.

Backward trajectories or trajectory images may be missing for several reasons:

1. The monitoring site is outside the WRF meteorological domain used by FEAST (approximately latitude 32.22–42.30 and longitude −124.71 to −113.03). For example, if the monitoring site is in Oregon or near Mexico, the backward trajectories may not be computed.
2. The WRF meteorological run for that period may be missing or incomplete. WRF is run automatically each day by DRI with 00 UTC and 12 UTC initial times. If there is a system or hardware issue and a run fails, the necessary meteorological fields may not be available, and trajectories cannot be generated.

An Exceptional Event (EE) is an unusual or naturally occurring event (such as wildfire smoke or high-wind dust) that affects air quality and is not reasonably controllable or preventable. Exceptional Event demonstrations document how these events caused or contributed to elevated pollutant levels and support the exclusion of affected data from NAAQS compliance determinations, consistent with U.S. EPA’s Exceptional Events Rule.

FEAST is designed as a practical toolkit to support Exceptional Event (EE) screening, analysis, and documentation, with a particular focus on wildfire and prescribed-fire smoke impacts on PM_2.5 and ozone.

• 1. Exceedance Toolkits
  • a. Daily exceedance screening and preliminary reports

    Each day, FEAST automatically screens monitoring data to identify stations that exceed the daily PM_2.5 and 8-hour ozone standards.

    For each exceedance, FEAST generates a preliminary exceedance report that includes:

    • Time series of hourly and climatological PM_2.5 and O₃ concentrations at the monitoring site
    • Nearby fire activity (e.g., PFIRS, FRAP) and relevant spatial context
    • A preliminary assessment of the pollutant exceedance and its proximity to fire activity
    • Computed 24-hour back trajectories from the monitoring site at the hour of peak concentration using CANSAC 1.3-km meteorological fields
      • Trajectories are initialized at multiple heights above ground level (e.g., 100 m, 150 m, and 200 m AGL) to help identify upwind source regions and potential plume transport pathways.
      • Trajectory plots are displayed alongside fire perimeters, satellite imagery, and HMS smoke products to support a coherent, time- and space-consistent narrative.

    These preliminary reports serve as a starting point for staff to quickly assess whether an exceedance may be fire-influenced and potentially relevant for an Exceptional Event demonstration.

  • b. Exceedance design value (DV) impact calculator

    A screening-level tool, based on AirNow data, that examines how removing selected high-concentration days (e.g., candidate EE days) could affect site design values.

    AirNow data are typically available with a 2–3 day lag. Results are for reference only and do not replace EPA’s official design value calculations using certified AQS data.

• 2. Exceptional Events Toolkit

  Most regulatory monitoring data used in FEAST are sourced from AQS. As a result, some datasets may have a lag of approximately six months to one year, depending on AQS update schedules.

  FEAST includes a dedicated Exceptional Events (EE) toolkit to support screening and documentation activities:

  • a. Design value tools

    Annual PM_2.5 and 8-hour ozone design value maps (2012–present) to explore spatial patterns, long-term trends, and EE-relevant sites and regions.

  • b. PM_2.5 tile (calendar) plots

    Daily PM_2.5 calendar views for a selected site and time range, with each day color-coded by concentration tier (e.g., < 9 µg/m³ = green, 9–35 µg/m³ = yellow, > 35 µg/m³ = red) to highlight exceedances, seasonality, and smoke episodes at a glance.

  • c. Fire–exceedance timeline analyzer

    Tools to visualize the temporal relationship between monitored exceedances and nearby fire activity, helping distinguish fire-influenced days from other high-pollution episodes.

  • d. EE Word template generator

    Automated generation of a draft Exceptional Event demonstration document in Word (.docx) format, including figures, tables, design value summaries, and chapter-by-chapter content prompts. The template structure is based on an EPA prescribed-fire EE demonstration and includes a table of contents to streamline report assembly and reduce manual formatting effort.

All FEAST design values and screening outputs are intended for screening and visualization purposes only and may differ from official EPA regulatory design values and determinations.

Prescribed fires come from PFIRS (planned/intentional under permits), while wildfires come from FRAP (unplanned/uncontrolled). FEAST uses distinct symbols and legends for each.

PM_2.5 Annual DV: 3-year average of annual means.

Ozone DV: 3-year average of each year’s 4th-highest daily max 8-hour average.

FEAST’s Design Value Estimator helps explore how excluding specific days may affect attainment status.

FEAST applies basic U.S. EPA design value completeness concepts so that calculated design values are broadly consistent with federal guidance. However, values shown in FEAST are intended for screening, visualization, and analytical support only, not for regulatory determinations.

What design values does FEAST show?

FEAST displays the following design values:

• Annual PM_2.5 design values (µg/m³), and
• 8-hour ozone (O₃) design values (ppb), defined as the 3-year average of the 4th-highest daily maximum 8-hour O₃ concentrations.

Data completeness and gaps (high-level)

Minimum data requirements
EPA design value calculations require a minimum amount of valid data in each year. In general:

• PM_2.5: A “complete” year typically requires sufficient valid daily values across the calendar year (e.g., at least 75% of scheduled sampling days in each quarter). A valid design value generally requires three complete years of data.
• 8-hour O₃: A “complete” year typically requires approximately 75% data capture for daily maximum 8-hour averages during the ozone season. Design values are based on three consecutive complete years.

Exact completeness criteria are defined in EPA design value guidance and in 40 CFR Part 50 appendices for PM_2.5 and ozone.

Handling missing data
Missing or invalid observations are excluded from design value calculations. If a year or analysis period does not meet basic completeness thresholds, FEAST may:

• Omit the design value if data gaps are too large.

Screening vs. regulatory values
Design values shown in FEAST are intended for screening, visualization, and Exceptional Event (EE) analysis support. They may not fully implement all U.S. EPA regulatory completeness tests, substitution procedures, or calculation rules and are not official regulatory design values.

Final regulatory design values must be calculated using certified AQS data and EPA’s official procedures, as described in EPA guidance and 40 CFR Part 50.

For detailed completeness and design value requirements, please refer to:

• EPA Design Value Guidance: https://www.epa.gov/air-trends/air-quality-design-values
• Pollutant-specific appendices to the NAAQS (40 CFR Part 50)

Yes. FEAST is publicly accessible and does not require a login.

Many datasets offer CSV downloads. Maps can often be exported as images using built‑in controls or your browser’s print‑to‑PDF function. For bulk data requests, please contact CARB/DRI at feast@arb.ca.gov.

Not yet. FEAST currently focuses on integrated visualization and analysis. If you need automated access to FEAST‑derived products, please contact us with your specific use case at feast@arb.ca.gov.

FEAST does not collect personally identifiable information. Browsers may store small preferences (e.g., last viewed layer) to improve usability.

For most datasets, FEAST does not apply additional QA/QC. Instead, it preserves the quality flags and processing choices of the original data providers so that the information remains faithful to the source agencies.

The primary exceptions are PurpleAir and historical observed meteorological data (typically one to two years behind real time):

• PurpleAir hourly data
  CARB applies a dedicated QA/QC and processing pipeline to PurpleAir data before they are made available to FEAST and related applications. This includes outlier detection (e.g., clustering‑based methods) and related screening steps to improve the reliability of the sensor network.
• Historical observed meteorological data
  Historical meteorological data originate from CARB’s internal met database and are then ingested into FEAST. These data are assigned quality flags using:
  • Provider‑supplied QA/QC where available (e.g., FAA, CIMIS), and
  • Simple, documented range‑check rules for AIRS and RAWS variables such as temperature, relative humidity, dew point, wind speed/direction, and precipitation.
  This approach ensures a consistent and transparent level of screening for long‑term meteorological records used in FEAST analyses, while still respecting the original provider QA/QC where it exists.

FEAST supports screening and documentation. Provisional data, satellite smoke extents, model outputs, and community sensors have uncertainties. Use certified data and documented methods for formal regulatory determinations.

FEAST is a screening and analysis tool. Its outputs—maps, design values, trajectories, and exceedance reports—are intended for planning, context, and Exceptional Event support only and do not constitute official regulatory determinations. Regulatory decisions must be based on certified data (e.g., AQS) and EPA-approved methods.

Community Sensor Terminology

For tools such as FEAST, we recommend referring to PurpleAir and similar devices as community sensors (or, where greater precision is needed, low-cost community sensors) rather than “low-cost sensors” alone. The term community sensors emphasizes how these devices are used—within dense, community-based networks—without implying that the data are inherently inferior.

Limitations of Community Sensors in Smoke

Most community sensors (e.g., PurpleAir) use optical particle counters that infer particulate mass from light scattering. As a result, these measurements have known limitations, particularly during smoke events:

• Optical measurements can over-respond in heavy smoke, especially when particle concentrations are very high or when particle properties differ from those assumed during calibration.
• High humidity can cause hygroscopic particles to grow, increasing light scattering and leading to overestimation of PM_2.5.
• Sensor performance may vary by model, firmware version, and installation conditions (e.g., proximity to buildings or local emission sources).
• Community sensors are not regulatory monitors and may not meet all regulatory siting, calibration, and quality assurance requirements.

How FEAST Handles Community Sensor Data

FEAST applies QA/QC checks and established correction approaches when using community sensor data. PurpleAir data, in particular, undergo several processing steps within CARB’s internal pipeline to ensure data quality and consistency before use in FEAST:

• Ingestion – Raw PurpleAir sensor data are retrieved from the source.
• Cleaning and outlier screening – Data are cleaned and screened for anomalous values using automated QA/QC procedures, including k-means clustering and related statistical methods to identify and remove outlier observations.
• Metric derivation – Final hourly PurpleAir PM_2.5 concentrations are derived. These include NowCast-based estimates, computed using a weighted average of PM_2.5 concentrations over approximately the previous 12 hours.

This process ensures that PurpleAir data displayed in FEAST represent a consistent, quality-controlled view of measurements from the community sensor network.

However, users should still:

• Compare community sensor data with nearby regulatory monitors where available.
• Interpret values with caution during exceptional smoke events or periods of high humidity.
• Consider the remaining uncertainty when using community sensor data to assess PM_2.5 levels.

FEAST supports the current versions of Chrome, Edge, Firefox, and Safari. When viewing PurpleAir data for prescribed fires, you may experience longer rendering times due to the large amount of data. For best performance, please close other resource‑intensive browser tabs and use a stable internet connection. The website is not yet mobile‑friendly; improved mobile support is planned for a future update. Because FEAST provides many features and detailed visualizations, we strongly recommend using a larger screen (such as a desktop or laptop) rather than a mobile phone.

All timestamps shown in FEAST are in Pacific Time (PT) and are not automatically adjusted to other time zones. Coordinates use WGS84 unless otherwise noted.

When you use FEAST in reports, maps, or analyses:

1. Name FEAST and the organizations behind it
   • State that you used FEAST, developed by the California Air Resources Board (CARB) and the Desert Research Institute (DRI).
2. Include the date you accessed FEAST
   • Online data can change over time. Always include the date you used FEAST (for example: “Accessed January 30, 2026.”).
3. Acknowledge the original data providers
   FEAST integrates data from several systems. List them together, such as:
   • PFIRS (prescribed fire information)
   • CAL FIRE FRAP (fire perimeters)
   • AirNow/AQS (air quality measurements)
   • NOAA HMS (smoke and fire detections)
   • DRI/CANSAC (WRF meteorological modeling)
   • BlueSky (smoke modeling)
4. Indicate if the data are provisional or preliminary
   • If the FEAST interface or documentation indicates that some data are “provisional” or “preliminary,” include a note such as: “Some datasets are provisional and subject to revision.”
5. Make clear there is no endorsement
   • Ensure your wording does not suggest that CARB, DRI, or any data provider endorses your analysis or product.
   • You may add a simple statement such as: “Use of these data does not imply endorsement by CARB, DRI, or any original data provider.”

---

Example text you can reuse

Analyses were performed using the Fire & Exceptional-Event Assessment, Screening, and Toolkit (FEAST; California Air Resources Board and Desert Research Institute), accessed January 30, 2026. FEAST integrates data from PFIRS, CAL FIRE FRAP, AirNow/AQS, NOAA HMS, DRI/CANSAC WRF, and BlueSky. Some datasets are provisional and subject to revision. Use of these data does not imply endorsement by CARB, DRI, or any original data provider.

Example: “FEAST (Fire & Exceptional-Event Assessment, Screening, and Toolkit), California Air Resources Board (CARB) and Desert Research Institute (DRI), accessed <date>. Includes data from PFIRS, CAL FIRE FRAP, AirNow/AQS, NOAA HMS, DRI/CANSAC, and BlueSky.”

Email feast@arb.ca.gov with a screenshot, URL, and steps to reproduce. For feature requests, include your use case, data source, expected frequency, and any constraints.