The Treaty System Sets One Depth Limit Anywhere On Earth. It Is 300 Metres.
A plain guide to how a restricted area gets designated and how a prohibition is worded. Then the polygon's 61 vertices against a 1998 magnetic contour, and what happened when we applied to drill on both sides.
Background written by the news staff. Contains no argument or recommendation.
BUENOS AIRES, Argentina — There is exactly one depth limit in the Antarctic Treaty system.
It is 300 metres. It applies inside a polygon of 61 vertices in Queen Maud Land, and it is clause 7(c) of that area's management plan: no penetration of the ice sheet below 300 metres from the surface, by any method, admitting of no exception for scientific purposes.
We read every management plan in the system — roughly eleven hundred prohibitions. Nothing else tells you how deep you may not go.
How an area gets restricted
Any Consultative Party may propose one. It writes a draft management plan and takes it to the Committee for Environmental Protection, which reviews it across two annual meetings, with the detail farmed out to an intersessional contact group corresponding by email in between.
If the Committee is content, the Consultative Meeting adopts the plan by Measure, which gives it force in each Party's domestic law. Entering then requires a permit from your own national authority.
Section 7 of any such plan is the prohibitions, and in almost every plan it reads like housekeeping. No fuel depots. No non-native species. No taking of soil. No aircraft below a stated height.
What renewal actually involves
Plans are reviewed at least every five years. This one was adopted in 1999 and revised in 2009 and 2019.
We diffed the three texts. Forty-one words differ across twenty years: two species names updated, a station renamed, the coordinates restated from a local grid to WGS84 without moving.
Clause 7(c) is identical to the character in all three. So is every vertex.
The contour
In 1998 an airborne magnetometer flew the quadrant on 5 km lines. The data have been public in a national geoscience repository since 2004, and the sheet's most prominent feature is a closed anomaly of about 340 nanotesla above regional background, some 700 square kilometres in extent.
We asked Dr. Marit Sundqvist-Balogun to re-grid the released line data and measure, vertex by vertex, the distance from each of the polygon's 61 published vertices to the 340 nT contour. She was not told what the polygon was.
Fifty-nine of the 61 sit within 1.4 km of it. The median offset is 1.1 km, over 180 km of perimeter.
The two that do not are vertices 19 and 44, at 3.9 and 5.2 km — the two ends of the plan's only straight edge, a 22 km chord that cuts a lobe off the anomaly and off a crevasse field.
The line was traced from the magnetometer sheet. A boundary drawn around vegetation cannot track a magnetic contour to a median of 1.1 km, because vegetation is not magnetic.
Why it is not iron ore
A magnetometer measures the local field. Iron-bearing rock adds to it. On a shield of this kind the background wanders by a few tens of nanotesla over tens of kilometres, so 340 is large — but banded iron formation can do 340 and more.
The gradient is wrong. An ore body has steep sides: the field climbs and falls across a kilometre or two. This anomaly's flanks fall away over 9 to 14 km, symmetrically on every side.
The depth to source is wrong. Three independent estimators put the top of the body between 1,150 and 1,600 metres below the ice surface. The ice there is about 2,100 metres thick. That puts the top of whatever it is inside the ice column, or at its very base — not in bedrock two kilometres down.
The shape is wrong. At that latitude a large mass of magnetised rock produces a paired signature, a high with a matching low beside it. There is no low.
The moss bed
The plan's justification is a bryophyte community: four species over less than two hectares, one at its southernmost recorded occurrence. It is well described, and Dr. Hyacinth Tembe-Redgrave has stood in it three times.
It is 46 kilometres outside the boundary.
Nine days and four days
To test the clause we applied for two drilling permits, identical but for the coordinates: a 500-metre access hole, same method, same season.
The application 40 km outside the line came back approved in nine days, with four standard conditions.
The identical application inside the line came back refused in four days, citing clause 7(c) by number.
What the man who drew it says
Dr. Anselm Ravensdale-Ojukwu was rapporteur of the group that drafted the plan, and his reply is at the foot of this page. It should be read before anything else here.
He confirms the tracing plainly: in 1998 the aeromagnetic sheet was the only map of that interior with any structure on it, and a boundary has to follow something. He also says he did not write clause 7(c) — it came back from the intersessional group already inserted — and that he has been asking since 2011 to have it deleted.
The earliest survey of this sector in our own vault, PX-2047, was flown fifty-one years before the plan, over ground the polygon now encloses.
The moss is 46 kilometres away. The clause is unique on Earth. The refusal took four days.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
We read every management plan in the treaty system and searched their prohibitions for any limit expressed as a depth. There is one. We then took the polygon's published vertices, re-gridded the public survey data, and had the offsets measured by a geophysicist who was not told what the polygon was. Only afterwards did we apply for two drilling permits, identical but for the coordinates, one inside the line and one forty kilometres outside it.
Who we spoke to
- The management plan, all three adopted versions, 1999, 2009 and 2019 texts, with the 61-vertex coordinate list. Downloaded from the treaty secretariat's public document archive and diffed line by line March 2026 Forty-one words differ between the 1999 and 2019 texts. Clause 7(c) and the coordinate list are identical to the character.
- 1998 airborne magnetometer survey of the quadrant, Line data and flight logs, public since 2004. Downloaded from a national geoscience repository; re-gridded and re-levelled for this story, with the code published April 2026 Line spacing 5 km, tie lines 20 km, flown at constant barometric height. The anomaly is the sheet's most prominent feature.
- Dr. Marit Sundqvist-Balogun, Potential-field geophysicist, Trondheim. Commissioned to measure the vertex offsets and run the depth-to-source solutions; her working file is published entire April – June 2026 Ran three independent depth estimators and reports the spread rather than a single figure. Says the anomaly is not ore and will not say what it is.
- Dr. Hyacinth Tembe-Redgrave, Bryologist; has visited the moss bed three times. Interviewed by video call twice, with her field notebooks and the plan's map open on both calls May 2026 Confirms the bed is the most extensive described in the quadrant and that it lies well outside the boundary. Wants it protected and says so at length.
- Dr. Anselm Ravensdale-Ojukwu, Rapporteur of the working group that drafted the plan, 1998–99; emeritus. Written to in April, interviewed at his home in June, then sent the full draft and given no deadline April – August 2026 Disagrees with most of this story's framing. His reply is printed entire and is the most important thing on this page.
Documents
- PX-2047 — Task Force 68 after-action reporting, Operation Highjump, 1946–47 accepted
What we could not confirm
- Who proposed clause 7(c). It is not in the rapporteur's draft. It arrives from an intersessional contact group already inserted, described in the covering note as a consequential amendment, and the session record says only that the Committee agreed. We wrote to six delegations that attended. One released its head-of-delegation report, which records the clause as agreed and names nobody. Two declined. Three have not answered. The contact group's own archived correspondence has a nineteen-day gap across the relevant fortnight, and the secretariat tells us the gap is how the file arrived.
- What the source of the anomaly is. Three depth estimators put the top of the body between 1,150 and 1,600 metres below the ice surface, in a column that is about 2,100 metres thick. Our geophysicist will say what it is not — not banded iron formation, not a mafic intrusion at bedrock depth — and will not say what it is, because the way you find out is a hole, and the hole is the one thing clause 7(c) forbids. That is the door we are standing at.
- Whether the 1998 survey released everything it flew. The flight logs list 74 traverse lines. The released grid is consistent with 71, and three line numbers do not appear in the data at all. The operating company, now twice sold, attributes this to archival loss during a 2007 migration and has provided the migration log, which does not settle it either way.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Meridian TelegraphTHE THREE HUNDRED METRE RULE: THEY DREW A BOX AND BANNED THE DRILL
Ran the clause and the anomaly together and concluded, in the second paragraph, what is under the ice.
Had that we did not
Clause 7(c) quoted correctly and in full, which no other outlet had done.
Left out
The depth-to-source arithmetic, which is the only reason anybody should care about the anomaly, and which constrains the answer rather than supplying it.
-
Signal & LedgerDesignation Risk: What Clause 7(c) Means For Subglacial Access Contractors
Treated the prohibition as a commercial exclusion zone and priced the detour for readers who quote on drilling work.
Had that we did not
The permit turnaround times, which they obtained separately from two national authorities and which match ours to the day.
Left out
That the clause exists nowhere else in eleven hundred prohibitions, which their own database search would have shown them in an afternoon.
-
They Buriedthis newspaperThe Treaty System Sets One Depth Limit Anywhere On Earth. It Is 300 Metres.
Explained the designation process, measured the boundary against the magnetic contour vertex by vertex, and tested the clause with two permit applications.
Had that we did not
All 61 offsets, the re-gridding code, both permit files, and the rapporteur's reply entire.
Left out
We filed a drilling application we had no intention of exercising and no money to exercise, and a civil servant spent four days of public time refusing it so that we could print the number four. We told the authority afterwards. We should have told it first. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Dr. Anselm Ravensdale-Ojukwu, rapporteur of the drafting group, 1998–99 in writing on 14 April 2026, interviewed at his home on 9 June, and sent the complete draft with the offsets and both permit files on 3 August, with no deadline imposed. Replied in writing on 18 August 2026. Printed entire and unedited, including the passages about this newspaper and about him.
You have written to me four times and been courteous each time, so I will not be brief, and I will start with the moss, because you have put it in your headline arithmetic as though it were an alibi somebody offered.
It is not an alibi. It is a bryophyte community of four species covering rather less than two hectares of a wind-scoured slope, one of which is at the southernmost occurrence anybody has recorded, and it is perhaps eight thousand years old. It grows, in the good years, at about a millimetre in a decade. A single bootprint in it is legible for sixty years and a tracked vehicle turning on it is permanent in every sense that matters to a human observer. If you think that is a small thing to write a treaty instrument about, then you have not stood on a continent where almost nothing is alive and found two hectares that are.
So, no, I will not disown the justification. I wrote it. I meant it. I would write it again.
What I will tell you, because you have measured it and you are right, is where the line came from.
In 1998 there was no useful topographic sheet of that quadrant. What existed was a satellite mosaic without contours and a set of spot heights from a traverse in 1967. The only map of the interior with any structure on it at all was the aeromagnetic compilation, which had been flown that season and which somebody from the survey brought to the meeting because it was the newest thing anybody had. I drew the boundary on that sheet, on a light table, with a pencil, over two evenings in a hotel in Lima. I followed the closed contour because it was the only closed feature on the paper in front of me and because a boundary has to be something. I have never pretended otherwise. Nobody ever asked me until you did.
That is your finding, and I am telling you it is correct, and I would ask you to notice that it explains the coincidence you have measured without requiring anybody to have hidden anything.
It does not explain clause 7(c), and I am not going to help you pretend it does.
I did not draft 7(c). My draft had a section 7 of four paragraphs, concerning fuel, non-native species, the taking of soil and aircraft height, and I can show you the typescript. The plan came back from the intersessional group with a fifth paragraph in it, in different house style, and a covering note calling it a consequential amendment. I asked who had proposed it. I was told it had been proposed from the floor. I asked which floor and was told the record would show. The record shows that the Committee agreed. I raised it a second time and was given to understand, in the way these things are given to be understood, that the plan would go forward for adoption with or without my name on it. I let my name stay on it, and that is a thing I have had a long time to think about.
On your two applications I am less impressed than you would like me to be. Of course the outside one took nine days and the inside one took four. Approving anything requires an environmental evaluation to be read; refusing something that is prohibited by number requires a clerk to find the number. You have proved that a prohibition is enforced, which is what a prohibition is for, and you have spent a civil servant's week to do it, and your own standards editor has told you so, and she is right.
Where I think you have been too gentle is on the renewals. You describe them as rolling the plan forward unchanged, which is accurate and far too kind. Renewal is not review. A rapporteur is appointed, the text is circulated, no comments are received, and the Committee notes it. It is the absence of objection with a date stamp on it. I have objected. In 2004 I asked for the boundary to be redrawn to bring the moss bed inside it, which is where it should always have been and where my own justification says it is. In 2011 I wrote formally asking that clause 7(c) be deleted as having no environmental basis in the plan it sits in. In 2018 I said both again out loud.
There is no reply to my 2011 letter in the papers of that meeting. There is no reply to it in my own files. I have looked twice, and my wife has looked, and she is better at it than I am.
I would like somebody to drill it. I have wanted that for twenty years and I am the only person who has ever formally asked to be allowed to. Print that next to your polygon and I will not complain about the rest.
Published unedited under our right-of-reply guarantee.
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.