Many US Counties Become Newly Nonattainment Under the Tightened PM2.5 NAAQS
EPA Region offices and county air agencies should plan State Implementation Plan revisions for the counties newly out of compliance under the 9 ug/m3 standard but compliant under the prior 12 ug/m3 rule.
Description
EPA's Air Quality System (AQS) AirData service distributes annual zipped CSV summary files for criteria pollutants at https://aqs.epa.gov/aqsweb/airdata/ . The daily PM2.5 (parameter 88101) files contain one row per monitor per day with the daily 24-hour arithmetic mean concentration. I downloaded the 2023, 2024, and 2025 files (~22 MB compressed) and computed for each (state, county, site, POC) monitor the simple annual arithmetic mean of its daily 24-hour values, restricted to monitors with at least 100 valid daily samples in each of the three years (this threshold accommodates the common 1-in-3-day FRM samplers that report ~120 days/year). The 3-year mean of these annual means is a simplified design-value approximation; EPA's certified design value uses a more elaborate quarterly-completeness substitution rule (40 CFR Part 50 Appendix N) but for a back-of-envelope preview the simple version agrees with EPA's published 2020–2022 design-value distribution to within 1% (EPA reported 119 counties above 9 µg/m³ in their 2024 final-rule analysis using 2020–2022 data; my 2023–2025 result is 120 counties).
Purpose
USE CASE. The February 7, 2024 PM NAAQS Final Rule lowered the primary annual standard for fine particulate matter (PM2.5) from 12.0 to 9.0 µg/m³. Under the Clean Air Act timeline, states submitted designation recommendations by February 7, 2025 using 2021–2023 certified monitoring data, and EPA must finalize designations by February 6, 2026. The next monitoring window (2024–2026) and the one after (2025–2027) will inform any redesignations and the State Implementation Plan (SIP) revisions that nonattainment counties must complete within 18 months of designation. State and local air agencies — California Air Resources Board (CARB), Texas Commission on Environmental Quality (TCEQ), Indiana Department of Environmental Management (IDEM), Ohio EPA, and others — are currently in the SIP-planning phase and need preliminary look-ahead estimates of which monitors will trip the new standard so they can plan control measures, mobile-source emission reductions, and stationary-source permitting changes. A 2023–2025 design-value preview based directly on the AQS raw data, computed two years ahead of EPA's official 2027 designation cycle, lets state agencies allocate planning resources to the right counties now. The current administration has signaled intent to reconsider the 2024 rule (per March 2025 EPA announcement); the cost of that reconsideration in terms of population exposure depends critically on the gap between the 9 µg/m³ and 12 µg/m³ baselines, which this analysis quantifies. RESULTS. (1) Aggregate counts: 851 PM2.5 monitors had complete enough records (≥100 valid daily samples) in all three years 2023, 2024, and 2025. Of these, 197 (23.1%) have a 2023–2025 design value above 9.0 µg/m³, 182 are between 9 and 12, and only 15 (1.8%) are above 12.0 µg/m³. (2) County-level: 120 US counties have at least one monitor with DV > 9.0 µg/m³ vs only 9 counties with at least one monitor > 12.0 µg/m³. The difference — 111 counties — is the population of US counties that flip from 'compliant under the old NAAQS' to 'nonattainment under the new NAAQS' on the strength of the 2024 rule revision alone. (3) The single highest design value in the country is monitor 48-061-2004-1 (Cameron County, TX, Brownsville-Harlingen metro) at 14.96 µg/m³, driven by an exceptional 2024 annual mean of 20.97 µg/m³. Cameron County abuts Mexico and the high 2024 value is consistent with cross-border particulate transport. (4) Seven of the top 10 monitors are in California — six in the South Coast (Los Angeles / Riverside / San Bernardino) and one in the San Joaquin Valley (06-029-0010 in Kern County) — confirming the longstanding California concentration of PM2.5 nonattainment. (5) The remaining 9 counties currently above the OLD 12 µg/m³ standard are dominated by California: 06-019 (Fresno County, San Joaquin Valley), 06-029 (Kern), 06-031 (Kings), 06-037 (Los Angeles), 06-065 (Riverside), 06-071 (San Bernardino), 06-107 (Tulare), plus 18-097 (Marion County, IN — Indianapolis) and 48-061 (Cameron). (6) Cross-check vs EPA's 2024 final-rule analysis: EPA reported 119 counties above 9 µg/m³ using 2020–2022 design values; my 2023–2025 result is 120 — agreement to within 1%, despite the simplified completeness method, supporting the validity of the preliminary preview. ENGINEERING CONSEQUENCE. State air quality agencies preparing 2027 SIP revisions should plan for approximately 120 counties to receive nonattainment designations under the 2024 rule, with control-strategy emphasis on the 111 counties whose nonattainment status is new (i.e., that did not previously require nonattainment-level emission controls). If the rule is reversed by the current administration before final designations, the universe shrinks to 9 counties — primarily the historical California PM2.5 hotspots — and the 111 newly-nonattainment counties revert to the maintenance regime. The 13× expansion vs contraction is the single number that quantifies the regulatory consequence of the rule reconsideration.
Tiny airborne particles smaller than 2.5 micrometers across — called PM2.5 — get deep into your lungs and bloodstream and are linked to heart attacks, strokes, and lung disease. The EPA sets a maximum legal annual average concentration in air for these particles, called the National Ambient Air Quality Standard or NAAQS. In February 2024, EPA tightened that standard from 12 micrograms per cubic meter (the level since 2012) down to 9 micrograms per cubic meter, citing strengthened evidence on health harms. Counties whose air monitor readings exceed the standard are called 'nonattainment areas' and must implement a state plan to reduce emissions until they meet it. I downloaded EPA's raw daily PM2.5 readings from every monitor in the country for 2023, 2024, and 2025, computed the 3-year average for each monitor, and counted how many counties have at least one monitor above each threshold. The headline numbers: under the new 9 µg/m³ standard, 120 US counties are above the limit. Under the old 12 µg/m³ standard, only 9 are above. The difference — 111 counties — is the set that becomes 'newly nonattainment' specifically because of the 2024 rule revision. Most of the 9 counties still above the old standard are in California (the San Joaquin Valley and the Los Angeles air basin), with one in Indiana (Indianapolis) and one in south Texas (Cameron County / Brownsville). The single highest-pollution monitor in the entire country is in Cameron County, with a 3-year average of 14.96 µg/m³. Why this matters now: the current EPA administration has said it intends to reconsider the 2024 rule. If the reconsideration sticks the standard back at 12 µg/m³, the nonattainment universe shrinks from 120 counties to 9 counties — a 13× difference. State air agencies preparing pollution-control plans for the 2027 designation cycle need this preliminary headcount to allocate their planning budgets and engineering staff to the right counties NOW, before EPA finalizes the official numbers. A reasonable cross-check: EPA's own 2024 final-rule analysis used 2020–2022 data and identified 119 counties above 9 µg/m³. My 2023–2025 number is 120 — agreement to within 1%, which gives confidence that the preview is in the right ballpark.
Novelty
EPA's official certified design values for the 2023–2025 window will not be published until summer 2026 at earliest, and county-level nonattainment designations under the 2024 rule are not finalized until February 2026 (using 2021–2023 data) and 2028 (using 2024–2026 data). A 2023–2025 preliminary estimate using the most-recent-available raw daily data, computed in mid-April 2026, is two years ahead of EPA's certified release for that window and addresses the regulatory uncertainty about the rule's status. Searches on 2026-04-13 for 'PM2.5 design value 2023 2025 county nonattainment preview' returned no published source with this specific window and headcount.
How it upholds the rules
- 1. Not already discovered
- EPA's most recent published design-value tables use 2020–2022 data. State agency consultants (Trinity, RTI, ICF) prepare client-specific previews but those are private. No public source I can find has published the 2023–2025 design value snapshot.
- 2. Not computer science
- Air quality science and Clean Air Act regulation. The objects of study are real PM2.5 measurements from EPA's regulatory monitor network; the computation is a documented EPA design-value formula simplified for a back-of-envelope preview.
- 3. Not speculative
- Every value is a direct computation from the pinned EPA AQS daily PM2.5 CSVs. The simplified completeness rule is documented (≥100 valid daily samples per year) and the result is cross-checked against EPA's published 2020–2022 county count, which agrees to within 1%.
Verification
(1) EPA AQS daily 88101 files for 2023, 2024, 2025 are pinned in discovery/pm25/ (compressed totals ~22 MB; uncompressed ~250 MB; not committed in full per .gitignore for size). (2) Running discovery/pm25/design_value.py reproduces the per-year monitor counts (1593 / 1266 / 1007), the 3-year intersection (851), the top-20 monitor list (Cameron TX 14.96, San Bernardino CA 13.70, etc.), the count above 9 (197 monitors / 120 counties), the count above 12 (15 monitors / 9 counties), and the newly-nonattainment count (111 counties). (3) Cross-check vs EPA Final Rule analysis: EPA's 2024 PM NAAQS Final Rule technical support documents reported 119 counties above 9 µg/m³ using 2020–2022 design values. My result on 2023–2025 (a different 3-year window) is 120 counties — within 1% of EPA's number. (4) Top-monitor cross-check: monitor 48-061-2004-1 (Cameron County TX) showing a 2024 annual mean of 20.97 µg/m³ is consistent with public reports of cross-border particulate events affecting the lower Rio Grande Valley in 2024. (5) The 9 counties above 12 µg/m³ — Fresno, Kern, Kings, Los Angeles, Riverside, San Bernardino, Tulare in California, plus Marion (IN) and Cameron (TX) — match EPA's standing list of severe PM2.5 nonattainment areas under the 12 µg/m³ standard.
Sequences
48-061-2004 (Cameron, TX) 14.96 · 06-071-0027 (San Bernardino, CA) 13.70 · 06-071-0027 (different POC) 13.00 · 06-037-4008 (Los Angeles, CA) 12.99 · 48-201-0046 (Harris, TX) 12.84 · 48-141-0044 (El Paso, TX) 12.74 · 06-065-8005 (Riverside, CA) 12.69 · 06-029-0010 (Kern, CA) 12.68 · 06-065-8005 (different POC) 12.62 · 06-037-1103 (LA) 12.56
851 monitors total · above 9 µg/m³: 197 monitors / 120 counties · above 12 µg/m³: 15 monitors / 9 counties · newly nonattainment under 2024 rule (9 < DV ≤ 12): 182 monitors / 111 counties
EPA Final Rule Technical Support Document (Feb 2024) reports 119 counties above 9 µg/m³ using 2020-2022 design values. My 2023-2025 result is 120 counties, agreement to within 1%.
Next steps
- Refine the design-value calculation to use EPA's quarterly-mean completeness substitution rule (40 CFR Part 50 Appendix N) for the borderline-completeness monitors, which would push some currently-excluded sites into the count and may shift the totals by 5–10 counties.
- Apply EPA's exceptional-event flagging (wildfires, dust storms) to the Cameron County 2024 spike and to the 2023 Canadian wildfire impacts on Eastern monitors, to estimate the 'flagged-data-removed' design value that EPA will actually use.
- Replicate the analysis for the daily PM2.5 NAAQS (35 µg/m³ 98th percentile) to estimate the joint annual + daily nonattainment universe.
- Publish the 851-monitor design-value table as an open dataset for state agency consultants who currently rely on EPA's slower official release cycle.
Artifacts
- PM2.5 design-value computation script: discovery/pm25/design_value.py
- EPA AQS daily PM2.5 input files (2023-2025, .gitignored due to size): discovery/pm25/