NSF Merged the Atmosphere, the Solid Earth and the Ocean Into One $110 Million Pot. Geoscientists Have Run This Experiment Before.

Nine years ago, the Division of Earth Sciences ran a controlled experiment on its own community. In 2017 NSF issued a Dear Colleague Letter, NSF 17-067, removing the deadlines for Tectonics, Petrology and Geochemistry, Geophysics and EarthScope. A companion letter, NSF 17-086, did the same for Marine Geology and Geophysics in the Division of Ocean Sciences. Proposals could arrive any day of the year.

Geoscientists therefore already know the thing the rest of NSF's research community is about to find out, because NSF's own 2022 literature review on no-deadline review says it plainly: a no-deadline approach piloted in a GEO program "was associated with a steep decline in proposal volume." Removing the deadline did not merely redistribute submissions across the calendar. It reduced them.

That history is the most useful lens on NSF 26-516, Geosciences Research: Atmospheric, Geospace, Earth and Ocean Sciences (GEO Core), released August 17, 2026 as part of the twelve-solicitation foundational research package that also produced BIO Core and MPS Chemistry. GEO Core carries $110 million across an estimated 200 to 350 awards, merges the core research offerings of three divisions into one document, and lists its Full Proposal Deadline as "Proposals Accepted Anytime."

What $110 million buys once you divide it three ways

Run the arithmetic before you get excited about the headline. $110 million spread across 200 to 350 awards works out to roughly $314,000 to $550,000 per award. The solicitation encourages proposals "requesting up to five years of support." A five-year award at the midpoint of that range is on the order of $85,000 per year.

Set that beside the rest of the August package. BIO Core averages between $930,000 and $1.12 million per award. MPS Astronomy lands near $715,000. MPS Physics anticipates roughly 300 awards against $180 million. Geosciences drew one of the thinner per-award slices in the reset — and it did so for the directorate whose work most often requires vessels, aircraft, drill rigs, field seasons in remote places, and instrument networks.

The mitigating detail is that the most expensive item in ocean science is not usually charged to the grant. Ship time on the Academic Research Fleet is funded separately, which is why the solicitation instructs that "proposals that involve vessels from the Academic Research Fleet must include a University National Oceanographic Laboratory System Science & Marine Equipment (SME) request form as a supplementary document." That form is not a formality and it is not part of your budget justification. Omit it and you have submitted a cruise-dependent proposal with no cruise attached.

The three constituent programs remain visible inside the merged document: Atmospheric and Geospace Sciences, covering "foundational research that advances understanding of Earth's atmosphere and the near-Earth space environment"; Earth Sciences, covering "Earth's interior and surface processes"; and Ocean Sciences, covering "all fields of ocean studies." What disappeared is the separate solicitation, the separate deadline, and the separate front door.

The deadline was doing work you did not credit it for

Every geoscientist who has submitted to a no-deadline program has the same private admission: the proposal that was going to be ready in March was ready in September, and the one that was going to be ready in September is still a folder of figures.

A deadline is a coordination device. It synchronizes your co-investigators, gives your research office a date to staff against, gives your department chair a reason to protect your calendar, and gives your own procrastination an external opponent. Remove it and the proposal competes against teaching, service, fieldwork, manuscripts, and the next proposal — permanently, with no forcing function.

The upside is real and NSF states it directly in its own case for the change: investigators get "more time to prepare proposals and build strong collaborations," can "think more creatively without the pressure of a deadline," and can "propose more complex, interdisciplinary projects." Institutions benefit too, because submissions spread across twelve months instead of piling into two sprint weeks that overwhelm sponsored-programs staff.

But the practical consequence of a volume decline is worth stating without euphemism. If fewer people submit and the money stays constant, the funding rate for those who do submit goes up. Continuous submission is a quiet advantage for organized labs and a quiet penalty for everyone else. The correct response is to impose the deadline yourself: pick a date, put it on the department calendar, tell your co-PIs it is real, and treat it as if NSF had set it.

One more scheduling caution. GEO Core does not publish target dates the way MPS Astro does with its November target. Panels still need proposals in batches to convene, so submission timing relative to when review units assemble is now an unwritten variable. Ask the cognizant program officer for your subdiscipline when panels meet. That conversation is more valuable this cycle than it has ever been.

Three compliance items that decide proposals before review does

The merged solicitation carries requirements that were previously scattered across divisional documents, and each one is a clean administrative kill.

The off-site safety protocol. "Proposals that include off-site work require a one-page description of protocols that address potential hazards, including dangerous, remote or unpredictable conditions that may impact participant safety." Geoscience is a field discipline. If your project involves a research vessel, a remote camp, a volcano, an ice sheet, a cave, an aircraft or a boat of any size, this page is mandatory. It is one page — and one page is exactly the length of document that gets forgotten at 11 p.m.

Permits, in the Project Description itself. "Principal investigators are required to describe in the 'Project Description' how permitting for collecting, destroying (if applicable) and importing/exporting samples will be addressed." Not an appendix, not a letter — body text, competing for space with your science. Anyone working in foreign territorial waters, on tribal or federal land, in marine protected areas, or with samples subject to export control needs a paragraph that demonstrates a real permitting strategy rather than an assumption that it will work out.

The two-year archive clock. "NSF GEO requires that final data sets, metadata files, derived data products, relevant software/code, and physical samples be deposited in a long-lived and publicly accessible archive upon publication of results or within two years of collection, whichever comes first." Read the last clause twice. The obligation is not tied to your publication schedule. Data collected in year one of a five-year award must be archived by year three whether or not the paper exists. Cores, samples and code are named explicitly. Budget the curation cost and the personnel time in the proposal, because the requirement will arrive whether or not you funded it.

Who this opens the door to

Two eligibility facts are worth reading together. First, "unaffiliated individuals are ineligible" — this is an institutional competition. Second, and more interesting, there are no restrictions or limits on the number of proposals per organization, PI or co-PI.

That is a genuine divergence inside the same August package. MPS Chemistry caps senior personnel at two proposals per year and MPS Materials at one, a patchwork of quotas that forces chemists and materials scientists to ration submissions across their own portfolios. Geoscientists face no such quota. Combined with rolling submission, a GEO PI can submit, get a decline with useful panel feedback, revise, and resubmit inside a single fiscal year — a cycle that used to take eighteen months.

NSF also "encourages proposal submissions from institutions in the Established Program to Stimulate Competitive Research (EPSCoR) jurisdictions." For geoscience specifically that language has teeth, because so much of the discipline's subject matter — seismic hazard, aquifers, coastlines, severe weather, mineral systems — sits physically inside EPSCoR states whose institutions have historically submitted below their share.

The strategic read on GEO Core is straightforward. The money per award is modest, the door is open year-round, nobody is rationing your submissions, and the binding constraints are administrative rather than scientific: the ship-time form, the safety page, the permitting paragraph, and a data-archiving obligation that starts running the day you collect. Researchers who want a system that keeps track of solicitation requirements and deadlines they have to invent for themselves can use Granted to hold the details while they get back to the science.