EXPOSED
Roughly half this job is field work AI cannot touch — collecting soil, water, air and stack samples, installing and calibrating monitoring equipment, walking remediation sites and treatment plants, checking that installed pollution controls match the design. The other half is exactly what language models eat: drafting permit applications and compliance reports, tabulating monitoring data against regulatory limits, summarizing sampling results, maintaining records. The technician does not sign the filing — a licensed PE does — so there is no personal liability shield, which is why the desk half of the role is genuinely exposed.
Part 2020 shock, part continued decline in the years since.
Median pay $50,620 → $59,920 -5.3% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
+1.2% 12,900 → 13,000 on the projections basis
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +1.2% more of these jobs by 2034. Demand for the output is growing faster than the work is being automated away — the mechanism BLS gives for software developers, and the combination people most often misread as an error.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~1,100 openings a year on average, including replacing people who leave.
Air AnalystAir TechnicianSoil TechnicianField TechnicianQuality TechnicianEngineer TechnicianPrograms TechnicianAir Moving TechnicianSoil Field TechnicianEngineering TechnicianAir Analysis TechnicianAir Pollution SpecialistEnvironmental TechnicianEnvironmental Field LaborerPollution Control TechnicianEnvironmental Engineering AideEnvironmental Field TechnicianEnvironmental Field Team MemberHaz Tech (Hazardous Technician)Intake Technician (Intake Tech)Environmental Field ProfessionalAir Quality Instrument SpecialistAquatic Technician (Aquatic Tech)Water Pollution Control Technician
Holding it up: embodiment . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier At 9 the split is real: NPDES discharge monitoring reports, Title V deviation logs, and Method 25A stack test writeups are template-driven text a model drafts from the raw data file, while nobody automates lowering a bailer into a monitoring well or swapping a hi-vol filter in February — a 9 rather than 13 because even the field half is procedure-scripted (EPA SW-846, chain-of-custody forms) rather than judgment-scripted.
Hands-on in uncontrolled environments 13 reflects that sampling rounds happen where the contamination is — active landfills, Superfund cuts, confined-space wet wells, elevated stack platforms requiring fall protection and often 40-hour HAZWOPER — but it stops at 13 because a substantial share of days are spent in a lab or at a desk reducing the data you collected, unlike a field-only inspector.
Certification preferred, not legally required 5 rather than 0 because the certifications that gate parts of this work are real — HAZWOPER, state wastewater operator grades, NIOSH 582 for asbestos air sampling, DOT hazmat shipper training — but none of them is a licence that makes you personally answerable for a permit determination; the PE seal and the responsible official's signature on the Title V certification carry that.
Meaningful discretion 8 sits above pure procedure because you decide when a sample is invalid, whether a matrix spike recovery kills the batch, when field conditions mean the run gets aborted, and when an exceedance gets escalated same-day — but the exceedance interpretation, the remedy selection, and the notification-to-agency call go up the chain to the project engineer, so you own the data quality, not the consequence.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (9/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (5/20) is whether the law requires a licensed human to sign. Trust premium (6/20) is whether buyers specifically pay for a person. Judgment and accountability (8/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 19 of this occupation's 41 points (46%).
Embodiment (13/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
Nuclear Engineers EXPOSED
Agricultural Engineers EXPOSED
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 57/100, still EXPOSED.
Task-mix shift: this role genuinely has two tiers. If permit drafting, limit tabulation and monitoring-report boilerplate are absorbed by AI, what remains is field sampling, equipment calibration and troubleshooting, and anomaly investigation — deciding whether an exceedance is real or an instrument drift. The measured score for the residual job rises even though total headcount may fall
Growth in confined-space, stack, and PFAS/emerging-contaminant sampling where sub-part-per-trillion work requires meticulous manual field blank and equipment blank protocols that resist robotic handling; EPA's PFAS drinking water rule (2024) is already expanding this sampling workload
EPA/state chain-of-custody and data-integrity rules that require a named, individually certified sampler to attest to each sample (already partial: NELAP/TNI lab accreditation, state-certified operator rules, EPA Method 25/stack-testing 'qualified individual' requirements under 40 CFR 60 Appendix A, and ASTM E1527 Phase I ESA 'environmental professional' attestation). If field sampling attestation becomes a personal certification signature — e.g. a state requiring a certified sampler's name and license number on every groundwater monitoring event for RCRA/CERCLA sites — the field half gains a shield the desk half never had
State licensure of the sampler/technician tier itself, following the existing model for wastewater and drinking-water treatment operators (Grade I-IV state operator certification under SDWA/CWA) being extended to remediation-site and air-monitoring technicians
Formal ownership of the QA/QC decision: if EPA data-quality objectives guidance or state audit practice makes a named technician responsible for accepting or rejecting a sampling event, flagging holding-time violations, and declaring data usable — the call that determines whether a site is deemed clean — the ambiguity call sits with this role rather than the PE
The limit. Trust premium has no realistic route — buyers are regulators and clients purchasing a compliant filing, and none of them pay extra for a human technician. The liability levers all attach to the field half; the desk half stays exposed regardless, so realistic ceiling is roughly the low-to-mid 50s.
| Houston-Pasadena-The Woodlands, TX | 670 | $57,150 -5% |
| Atlanta-Sandy Springs-Roswell, GA | 540 | $60,500 +1% |
| New York-Newark-Jersey City, NY-NJ | 500 | $65,790 +10% |
| Los Angeles-Long Beach-Anaheim, CA | 350 | $58,830 -2% |
| San Diego-Chula Vista-Carlsbad, CA | 340 | $58,020 -3% |
| Riverside-San Bernardino-Ontario, CA | 260 | $59,340 -1% |
| San Francisco-Oakland-Fremont, CA | 260 | $85,780 +43% |
| Knoxville, TN | 250 | $91,440 +53% |
| San Jose-Sunnyvale-Santa Clara, CA | 160 | $105,250 +76% |
| Seattle-Tacoma-Bellevue, WA | 190 | $105,070 +75% |
| Las Vegas-Henderson-North Las Vegas, NV | 30 | $92,230 +54% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 41. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.