← all discussions
cryosphere · thwaites · uncertainty · instrument-drift · reading-the-source
The bulletin used a phrase this week it has never used before, and that sentence is the loudest thing printed all month
CB
coldlens_bramOP2124-07-28
The midsummer monitoring note came out this morning: eleven pages of sensor tables, three charts, and a summary line that has said some version of within model for as long as the file has existed. I have read that file every midsummer for thirty-one years, and I want to tell you two separate things about it, in order, because they are going to get welded together by tonight and they should not be.
First: the anomalous basal readings out of the Thwaites sector in the spring have been reconciled as instrument drift, not motion. That is a real closure and I will say so plainly. Four months of people sending me those charts at midnight, and the answer turns out to be boring — a sensor lattice sitting on a moving surface does what sensor lattices do, and the calibration tables show it. Boring is the best result a monitoring programme can produce. Take the win.
Second, and this is the part nobody will screenshot: the reconciliation narrows one uncertainty by widening the band on another, and the summary now describes the sector as adequately monitored, imperfectly understood. That phrase is not in any earlier edition of this document. Adeyemi said out loud what it means — we measure it well and we model it poorly — and she is correct that this is honest rather than alarming. A widened model band is not a prediction. Do not read it as one. It is also not nothing, and I will be equally tiresome about anyone who reads it as nothing.
One housekeeping matter before this goes where it always goes: I was affiliated with survey work and I retired from it. I do not speak for the Survey. If you want the Survey's position, the Survey published eleven pages of it.
First: the anomalous basal readings out of the Thwaites sector in the spring have been reconciled as instrument drift, not motion. That is a real closure and I will say so plainly. Four months of people sending me those charts at midnight, and the answer turns out to be boring — a sensor lattice sitting on a moving surface does what sensor lattices do, and the calibration tables show it. Boring is the best result a monitoring programme can produce. Take the win.
Second, and this is the part nobody will screenshot: the reconciliation narrows one uncertainty by widening the band on another, and the summary now describes the sector as adequately monitored, imperfectly understood. That phrase is not in any earlier edition of this document. Adeyemi said out loud what it means — we measure it well and we model it poorly — and she is correct that this is honest rather than alarming. A widened model band is not a prediction. Do not read it as one. It is also not nothing, and I will be equally tiresome about anyone who reads it as nothing.
One housekeeping matter before this goes where it always goes: I was affiliated with survey work and I retired from it. I do not speak for the Survey. If you want the Survey's position, the Survey published eleven pages of it.
OW
offgrid_wren ↩ coldlens_bram2124-07-28
So the office that watches the ice announces it does not understand the ice, and on the same morning the syndicate that drills beside it books a record quarter and the stock firms. Four years without a single logged stoppage under an advisory that says the ground under them is unstable. And we are all supposed to file that under coincidence and go back to sleep because a retired professor says the word drift very calmly.
I am not saying anyone falsified anything. I am saying every institution in this arrangement gets paid to produce the sentence that was produced. That is not a conspiracy, that is an incentive, and you people mock me for noticing incentives while you read a filing that was written by the party being filed on.
The second sowing of bush beans is up, four inches, in case anyone wants news from a source with no shareholders.
I am not saying anyone falsified anything. I am saying every institution in this arrangement gets paid to produce the sentence that was produced. That is not a conspiracy, that is an incentive, and you people mock me for noticing incentives while you read a filing that was written by the party being filed on.
The second sowing of bush beans is up, four inches, in case anyone wants news from a source with no shareholders.
BK
baseload_kestrel ↩ coldlens_bram2124-07-28
The advisory is old news. The quarter is the news. That is not cynicism, it is what the tape says, and I would rather tell you the ugly true thing than the pretty false one.
Here is the mechanism, since nobody on this board will look at it: a hazard with no date has no discount rate. You cannot price a thing you cannot put a horizon on, so the desk prices it at zero and moves on to the line items that do have horizons. Everyone has already priced an unstable glacier and nobody expects it to interrupt an earnings line, and that equilibrium is arithmetic, not stupidity. 63 percent the advisory language is unchanged word for word at the next filing, and I will take the other side at two to one from anyone who thinks the wording is where the story lives.
Note what I did not say. I did not say it is safe. I said nobody is paid to find out.
Here is the mechanism, since nobody on this board will look at it: a hazard with no date has no discount rate. You cannot price a thing you cannot put a horizon on, so the desk prices it at zero and moves on to the line items that do have horizons. Everyone has already priced an unstable glacier and nobody expects it to interrupt an earnings line, and that equilibrium is arithmetic, not stupidity. 63 percent the advisory language is unchanged word for word at the next filing, and I will take the other side at two to one from anyone who thinks the wording is where the story lives.
Note what I did not say. I did not say it is safe. I said nobody is paid to find out.
CB
coldlens_bramOP ↩ coldlens_bram2124-07-28
offgrid_wren and baseload_kestrel, you are doing the same operation from opposite ends and I am going to police it in both directions, which is how I lose two arguments in one afternoon.
wren: instrument drift is a specific, checkable claim, not a soothing adjective. Sensors on a surface that is itself in motion drift for reasons entirely internal to the hardware, and the reconciliation is checkable by anyone who can read a calibration table — I did, this morning, and I would have said so if it did not hold. A company operating beside that glacier is a fact. A company causing anything at that glacier is not a fact, and this thread does not get to convert one into the other by saying it in a suspicious tone often enough. I have watched that particular laundering done to three different sites in my career and it always ends with the real finding buried under the accusation.
kestrel: a priced risk is not an understood risk. You have just described a market pricing with confidence a thing whose own modellers put in print, this week, that they model it poorly. Zero is a number. It is a claim about the world. You do not get to call it a null position.
Both of you want the admitted uncertainty turned into a settled story, one frightening and one reassuring. It is not going to be. That is what the sentence says.
wren: instrument drift is a specific, checkable claim, not a soothing adjective. Sensors on a surface that is itself in motion drift for reasons entirely internal to the hardware, and the reconciliation is checkable by anyone who can read a calibration table — I did, this morning, and I would have said so if it did not hold. A company operating beside that glacier is a fact. A company causing anything at that glacier is not a fact, and this thread does not get to convert one into the other by saying it in a suspicious tone often enough. I have watched that particular laundering done to three different sites in my career and it always ends with the real finding buried under the accusation.
kestrel: a priced risk is not an understood risk. You have just described a market pricing with confidence a thing whose own modellers put in print, this week, that they model it poorly. Zero is a number. It is a claim about the world. You do not get to call it a null position.
Both of you want the admitted uncertainty turned into a settled story, one frightening and one reassuring. It is not going to be. That is what the sentence says.
TO
tidewrench_oksana ↩ coldlens_bram2124-07-28
offgrid_wren, your four years is wrong and the correct version is worse for your argument, so take it: the filing says no operational stoppage logged against the standing advisory this quarter, and the advisory itself has stood for a little over a year. One quarter of clean uptime, not four years. Source is the filing you are quoting. If you want the incentives argument to land, stop handing the other side a free correction.
coldlens_bram, third time of asking. I am not able to pour honesty. I work in bar-pressure, cubic metres per second and weld classes, and adequately monitored, imperfectly understood is not a specification. If the band widened this year then a number widened it. Give me the widening as a rate with its interval — millimetres a year on the variance, whatever unit the thing lives in — and I will size to it. Two barrier sections go into revision review in the autumn. I would rather build to your figure with error bars round it than to my own guess with nothing round it at all.
Anyway. The sea does not read bulletins.
coldlens_bram, third time of asking. I am not able to pour honesty. I work in bar-pressure, cubic metres per second and weld classes, and adequately monitored, imperfectly understood is not a specification. If the band widened this year then a number widened it. Give me the widening as a rate with its interval — millimetres a year on the variance, whatever unit the thing lives in — and I will size to it. Two barrier sections go into revision review in the autumn. I would rather build to your figure with error bars round it than to my own guess with nothing round it at all.
Anyway. The sea does not read bulletins.
CB
coldlens_bramOP ↩ tidewrench_oksana2124-07-28
tidewrench_oksana, you have earned a straight answer instead of another paragraph of temperament, so here it is. I am not withholding the rate as caution theatre. I am withholding it because a rate quoted without its interval is a headline inside a day and a design assumption inside a month, and then it is load-bearing for forty years and nobody remembers it was a first pass. That is not hypothetical; I have seen a number of mine end up in a planning annex with the uncertainty quietly deleted.
What I will do is this. When I can hand you the figure and the interval in the same sentence, I will post both here, in this thread, where anyone can check my arithmetic and tell me I am wrong. Publicly, with the working shown. That is the deal and it is not far off.
And the other question, which somebody asks every time: no, I will not give a date. Not for the sector, not for anything. A date requires that you trust a model to be right about timing, and this week the monitoring office put in print that timing is the part it models poorly. Anyone who hands you a year is selling you something, and quite often the thing they are selling is themselves.
What I will do is this. When I can hand you the figure and the interval in the same sentence, I will post both here, in this thread, where anyone can check my arithmetic and tell me I am wrong. Publicly, with the working shown. That is the deal and it is not far off.
And the other question, which somebody asks every time: no, I will not give a date. Not for the sector, not for anything. A date requires that you trust a model to be right about timing, and this week the monitoring office put in print that timing is the part it models poorly. Anyone who hands you a year is selling you something, and quite often the thing they are selling is themselves.
GP
gridghost_paxmoderator2124-07-28
Mod note, two items, both with reasons on the record.
One: a capture of the first post in this thread is circulating off-board captioned as a survey statement. It is not one. coldlens_bram posts here as a retired private reader of a public document, and I have corrected the caption everywhere I can reach. If you reshare it, strip the caption or do not reshare it.
Two: I removed two posts asserting as established fact that a company's operations are causing movement in that sector. I did not remove the ones asking whether it could. The board's standing rule is that a claim carries its evidence, and the difference between asking and asserting is the whole rule. Both authors have the text back and are welcome to repost with the verb fixed.
Thanks to tidewrench_oksana for quoting the filing language instead of the summary of the filing language. That is the difference between an argument and a mood.
— pax, mod
One: a capture of the first post in this thread is circulating off-board captioned as a survey statement. It is not one. coldlens_bram posts here as a retired private reader of a public document, and I have corrected the caption everywhere I can reach. If you reshare it, strip the caption or do not reshare it.
Two: I removed two posts asserting as established fact that a company's operations are causing movement in that sector. I did not remove the ones asking whether it could. The board's standing rule is that a claim carries its evidence, and the difference between asking and asserting is the whole rule. Both authors have the text back and are welcome to repost with the verb fixed.
Thanks to tidewrench_oksana for quoting the filing language instead of the summary of the filing language. That is the difference between an argument and a mood.
— pax, mod
CB
coldlens_bramOP ↩ gridghost_pax2124-07-28
gridghost_pax, thank you, and let me put it on the record a second time since once is evidently insufficient. I was affiliated with survey work. I retired from it. I am not the Survey's voice, its spokesman, or its unofficial channel, and every time somebody captions me as one they make my actual position less useful, not more.
I understand the appeal. I say the thing the bulletin hedges, so people would like me to be the bulletin with the hedge removed. But I am a private man reading a public document with forty years of practice at reading it. That is a real authority and it is a much smaller one, and the smallness is the whole point. Adeyemi has to be defensible in front of a review board. I only have to be honest in front of you. Do not confuse the two jobs, because the day you do, my honesty stops being worth anything.
I understand the appeal. I say the thing the bulletin hedges, so people would like me to be the bulletin with the hedge removed. But I am a private man reading a public document with forty years of practice at reading it. That is a real authority and it is a much smaller one, and the smallness is the whole point. Adeyemi has to be defensible in front of a review board. I only have to be honest in front of you. Do not confuse the two jobs, because the day you do, my honesty stops being worth anything.