EXPOSED
The modal environmental scientist splits time between field sampling and site inspections and a large pile of screen work — permit applications, Phase I ESA desktop reviews, compliance reports, environmental impact narratives, and data QA — and that document tier is exactly what language models now draft at usable quality. What survives is going to the site: collecting soil, water, air, and soil-gas samples, judging whether a contaminated plume actually threatens a receptor, and standing behind that call in front of a state regulator. Licensure is patchy (Professional Geologist in some states, CHMM, or a PE stamp on remediation designs), so the liability shield is real but thinner than in engineering or medicine.
Dipped in 2020, then grew past where it started.
Median pay $71,360 → $82,220 -7.8% 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
+4.4% 90,300 → 94,300 on the projections basis
Growing, and only partly exposed
The BLS expects +4.4% more of these jobs by 2034, and at 50/100 the work is only partly exposed — some tasks are automatable, the core of the job is not. Nothing here is in tension.
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.
~8,500 openings a year on average, including replacing people who leave.
EcologistResearcherAir AnalystPolicy AnalystClimate AdvisorClimate AnalystEnvironmentalistForest EcologistPolicy AssociateAquatic EcologistClimate EconomistEcological ModelerResearch ScientistConservation PlannerIndustrial EcologistEnvironmental AnalystEnvironmental PlannerRestoration EcologistWatershed CoordinatorClimate Change AnalystEnvironmental DesignerRestoration SpecialistEcological ProfessionalEnvironmental Scientist
Holding it up: judgment & accountability . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier A Phase I ESA records review, an NPDES permit renewal narrative, a Title V emissions inventory, and the boilerplate of a monitoring report are all now first-draafted by machine, but chain-of-custody sampling, well gauging, wetland delineation, and defending a risk-based closure to a state case manager are not — roughly half the billable hours sit on each side, which is why this lands at 10 and not 6 or 14.
Some physical or field component Field days mean walking transects, purging monitoring wells, running a PID at a vapor intrusion survey, and doing 40-hour HAZWOPER-covered work on active industrial sites — real uncontrolled environments — but a typical week is still more office than boots, so 12 rather than the 16+ of a remediation field technician who never leaves the site.
Certification preferred, not legally required There is no single 'environmental scientist' licence: some states let you sign a Phase I under ASTM E1527 Environmental Professional criteria with a degree plus five years, some require a Professional Geologist stamp on subsurface work, and CHMM or REM are resume items rather than legal gatekeeping — so the 7 reflects credential paths that exist but are state-patchy and often satisfied by experience alone.
Meaningful discretion Deciding whether a benzene plume has a complete exposure pathway to a receptor, whether to recommend no further action versus active remediation, and what to disclose in an ESA are calls made on incomplete site data with liability and property-transaction money riding on them — high discretion, but bounded by state screening levels, ASTM protocols, and a PE or PG who stamps the remedial design, which keeps it at 13 rather than 17.
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 (10/20 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 (7/20) is whether the law requires a licensed human to sign. Trust premium (8/20) is whether buyers specifically pay for a person. Judgment and accountability (13/20) 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 28 of this occupation's 50 points (56%).
Embodiment (12/20) 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.
Foresters 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 68/100 — SAFE.
Genuine two-tier job: if permit-application drafting, Phase I desktop reviews, and boilerplate NEPA narrative are absorbed, the residual day is receptor-pathway determination, defensible data QA sign-off with chain-of-custody attestation, and regulator negotiation — work whose evidentiary standing depends on a named human. Also watch state agencies adding AI-disclosure/attestation rules for submitted reports (as some permitting agencies have begun requiring for consultant filings).
Expansion of state Professional Geologist / Environmental Professional licensure to the ~30 states that lack it, plus tightening of the ASTM E1527 / EPA All Appropriate Inquiries 'Environmental Professional' definition to require a named licensee to personally sign Phase I ESAs and site closure opinions (EPA revised AAI to ASTM E1527-21 in 2022; further revision cycles are the watch point). A PE-stamp requirement on remediation system design and vapor-intrusion mitigation in more state programs does the same.
Licensed Site Professional-style programs modeled on Massachusetts MCP and New Jersey's Licensed Site Remediation Professional Act, where a private licensee — not the agency — issues the cleanup closure determination and bears personal audit and revocation exposure. Adoption by additional states, or extension of the LSRP model to brownfields programs elsewhere, is a concrete, observable event.
Litigation and enforcement pressure that names the environmental professional personally — PFAS and vapor-intrusion cases where the closure opinion is the disputed document, plus professional-liability insurers (e.g., E&O carriers for A/E/C firms) writing exclusions for conclusions not independently verified by a named licensee. That underwriting condition, already appearing in A/E policies for AI-generated deliverables, forces the consequential call back onto a person.
Sampling defensibility standards that require a qualified human physically present for chain-of-custody and field QC — e.g., state programs or EPA SW-846 guidance declining to accept drone- or autonomous-sensor-collected soil-gas and groundwater samples as primary evidence in enforcement contexts.
The limit. Trust premium has no plausible route: the buyer is a developer, lender, or PRP purchasing regulatory acceptance at lowest cost, not a human touch. Even with strong licensure the ceiling is the Massachusetts/New Jersey LSRP pattern — a thin signature layer over increasingly automated document production, which raises liability without necessarily preserving headcount.
| Sacramento-Roseville-Folsom, CA | 3,520 | $108,430 +32% |
| New York-Newark-Jersey City, NY-NJ | 3,350 | $85,510 +4% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 2,710 | $108,160 +32% |
| Los Angeles-Long Beach-Anaheim, CA | 2,520 | $105,340 +28% |
| Seattle-Tacoma-Bellevue, WA | 2,490 | $102,980 +25% |
| Boston-Cambridge-Newton, MA-NH | 2,270 | $100,360 +22% |
| Houston-Pasadena-The Woodlands, TX | 1,840 | $95,170 +16% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 1,500 | $77,710 -5% |
| Kennewick-Richland, WA | 300 | $129,110 +57% |
| San Francisco-Oakland-Fremont, CA | 1,320 | $127,210 +55% |
| Monroe, MI | 80 | $114,160 +39% |
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 50. 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.