When wildfire smoke turns a major city orange, air quality leads the news for a week. What that coverage never mentions is who runs the monitoring network measuring the smoke — a workforce operating aging equipment on funding that has fallen in real terms for two decades, and now retiring faster than it’s replaced.
What the data says
The EPA reports roughly 40% of the U.S. population lives where air pollution exceeds health-protective limits — even after five decades of Clean Air Act progress — while recent wildfire seasons among the worst on record pushed unhealthy smoke over tens of millions.
Where the readings diverge
“Calibrated, defensible — and sparse.”
The official monitoring network produces the legally defensible data that determines which areas meet health standards. But it’s expensive, thinly distributed, aging, and — per GAO — unable to meet growing demand for local, real-time information on hot spots and smoke.
“Dense, cheap — and inconsistent.”
Low-cost sensors can go anywhere and enable hot-spot and community monitoring. But GAO’s own field demonstration found meaningful performance variability — dense data isn’t decision-grade without calibration and QA.
What this means for the professionals we serve
For calibration engineers, air-quality data scientists, and atmospheric modelers, the modernization of the monitoring network is a distributed, recurring mission market — a sensor-integration project here, a smoke-season surge there. Agencies need fractional, verified expertise to turn cheap dense data into decision-grade information, precisely the work that doesn’t justify a permanent hire at any single agency.
The smoke gets covered when cities turn orange. The people running the network never do.
Air-quality coverage spikes with each smoke emergency and vanishes with the haze. The systematically under-covered story is the network underneath: state and local staff on two decades of declining real funding, equipment past its design life, 1990s software whose remaining experts are retiring, and a flood of low-cost-sensor data nobody is staffed to quality-assure. The next smoke emergency will be measured by whoever is left — and that fractional, specialized, trust-dependent demand is exactly an assembly problem.