NSF Kept One Date in Astronomy and Killed the Renewal Path: MPS Astro Is $100 Million, 140 Awards, and November 16

Almost everything NSF released on August 17 abolished the deadline. Twelve foundational research solicitations, more than $1.5 billion, and solicitation after solicitation listing "Proposals Accepted Anytime" where the due date used to be.

Astronomy kept a date. NSF 26-522, MPS Astronomical Sciences Research Programs, sets a full proposal target date of November 16, 2026, and then the third Monday in November, annually thereafter. Proposals are technically accepted at any time, but the solicitation attaches a consequence to arriving late: "proposals received after the target submission date may be reviewed in the following fiscal year."

That is not a soft suggestion. In a discipline where a graduate student's funding, a postdoc's start date and an observing campaign all hang off the award date, slipping a review by a fiscal year is a year of somebody's career. Astronomers should treat November 16 as a hard deadline that NSF has politely declined to call one.

Three programs became one, and one of them was the renewal path

MPS Astro replaces NSF 22-624 (Astronomy and Astrophysics Research Grants, the AAG program that has been the backbone of individual-investigator astronomy funding), NSF 22-627 (Advanced Technologies and Instrumentation), and NSF 19-605 (the Mid-Scale Innovations Program). Three separate competitions with three separate cultures, folded into one document carrying $100 million and roughly 140 awards.

The split is published, and it is lopsided in a way worth internalizing:

• Astronomical Sciences Core Research — 130 awards. Individual investigators and collaborative groups doing "observational, theoretical, laboratory and archival data studies," spanning cosmology, extragalactic and galactic astronomy, stellar astronomy, exoplanets, solar and solar system astronomy, time-domain astrophysics, laboratory astrophysics and computational methods.
• Astronomical Sciences Technology & Instrumentation — 10 awards. Advanced technology development, feasibility studies, and "instrumentation or software that enables new observations that are difficult or impossible to obtain with existing means."
• Astronomical Sciences Innovations — 1 to 3 awards. The former mid-scale lane, described below.

Ten instrumentation awards a year, nationally, for a community that builds the detectors, spectrographs, adaptive optics systems and control software the entire field depends on. ATI has always been small — a 2020 review of the program's thirty-year history documented an outsized scientific return on a modest budget — but seeing the number printed as a flat annual expectation clarifies the odds.

Then there is the deletion. "NSF will not accept proposals for accomplishment-based renewals via this funding opportunity" — and, in the solicitation's own language, "will return such proposals without review."

Accomplishment-based renewals let an established investigator with a productive, continuously funded program submit a renewal built around published results rather than a full new project narrative. For long-baseline observational programs — decades-long monitoring campaigns, survey follow-up, catalog maintenance — the ABR recognized that the honest case for continuing is the record, not a speculative five-year plan. That instrument is gone from astronomy funding. Investigators who have renewed on accomplishments for three cycles now write a full competitive proposal against 130 slots, and they write it by November 16.

The Innovations track is small, strange and worth reading closely

One to three awards a year sounds like a rounding error until you see what qualifies. Astronomical Sciences Innovations supports four categories:

1. Self-contained science projects with defined phases.
2. Construction or operation of a stand-alone, long-term facility or center — scoped, the solicitation says, "comparable to an NSF Mid-scale Research Infrastructure 1 project."
3. Design and development investments ahead of a larger build.
4. Open-access capabilities, defined as "new instruments for existing telescopes in return for broader access to the entire U.S. astronomical sciences community" and "provision of observing time for U.S. scientists on existing telescopes."

Category four is the one that deserves attention, because it is a funding line for buying the community something it cannot otherwise get. If your institution or consortium operates a telescope, NSF will fund an instrument for it in exchange for open access. If a private or foreign facility has nights to sell, NSF will fund the purchase of those nights for U.S. investigators. That is a fundamentally different proposal from "fund my research" — it is a negotiation, an access agreement and a governance plan wrapped in a science case, and the teams that win it are the ones that arrive with the partner already at the table.

Against that, the sharpest exclusion in the document: "NSF will not prioritize proposals for routine upgrades to existing instruments and facilities." Detector refreshes, control-system modernization, and the ordinary maintenance-by-another-name that observatories have historically routed through instrumentation programs are now explicitly disfavored. If you plan to submit an upgrade, the proposal has to establish that it enables observations that were previously impossible — not observations that were previously harder.

What the numbers do to your odds

$100 million across 140 awards averages roughly $715,000 per award, though the three tracks are not comparable: a facility-scale Innovations award will consume a large multiple of a core research grant, which compresses what remains for the 130 core awards below that average.

Historical context sets expectations honestly. AAG achieved success rates of 30 to 35 percent through fiscal year 2003, drifted below 30 percent in the mid-2000s, and fell to roughly 15 percent in the following decade. Nothing in NSF 26-522 changes the underlying arithmetic of a field that generates more good proposals than the budget can fund. What it changes is where they all land — one document, one date, one review structure — and folding the ATI and MSIP communities into the same solicitation means the proposals that used to compete in separate pools now sit in a single portfolio decision.

Two compliance details that will silently disqualify otherwise strong submissions:

Voluntary committed cost sharing is prohibited. If your institution's standard proposal template carries a cost-share line — and many do, as a reflex of showing institutional commitment — that line makes the proposal non-compliant. Check the budget form, not just the narrative.

Unaffiliated individuals are not eligible. Independent researchers and retired faculty without an institutional affiliation cannot submit, regardless of the science.

There is also a quiet requirement for anyone whose project leans on space-based data: proposals emphasizing "NASA data or resources" should address why NASA data are required. NSF is funding ground-based exploration — it describes itself in this solicitation as "the primary federal funder of America's ground-based exploration of space" — and the burden is on you to explain why the archival NASA component belongs in an NSF proposal rather than a NASA one.

The calendar decision to make this week

Astronomers have roughly twelve weeks. The practical sequence: decide by early September whether your project is Core Research, Technology & Instrumentation, or genuinely Innovations-scale, because those are three different documents and the wrong choice cannot be fixed in October. If you have historically renewed on accomplishments, start the full proposal now — writing a competitive narrative after years of ABRs takes longer than you remember. If you are contemplating an Innovations submission, the partner agreement and the access terms are the long pole, not the science case.

And if November 16 slips, understand precisely what you are choosing: not a late submission, but a possible one-year deferral of review. In a field where the next solicitation's target date is the third Monday of the following November, that is the whole cycle.

Keeping track of which NSF track a project belongs in, what each one prohibits, and when the real date falls is exactly the kind of administrative overhead that Granted is built to carry so the proposal writing gets the hours instead.