Hacker News Reader

Hacker News

Comment by gd1 | Hacker News Reader

Full thread

gd1

Marshall Islands sea level, unchanged for 20 years:

http://sealevel.colorado.edu/content/interactive-sea-level-t...

No one denies the science more than alarmists.

View on HN

Really hard to not take you for a crackpot. You:

* initially call everyone science-denying alarmists

* ignore all the responses that point out that there is trend in the data you linked.

* respond only to the reply that gives you the benefit of the doubt ('what smoothing do you recommend so we can see it's unchanged.') by switching datasets and trying to make the same point again (unable to defend it in with first source?)

* make a claim of your own ('it's unchanged'), but switch to a different, easier-to-defend claim when contested ('it's not a foot, like the original article claims')

So do you actually believe it's unchanged, or are you irrationally dead-set on disagreeing with the article? Can't take you seriously like this.

'Crackpot' is a several thousand word article on how climate change is submerging the Marshall Islands when all the hard data shows that sea level has done fuck all except bounce around for 23 years.

But sure, yeah, I'm a crackpot.

By the way, you will note in my initial source that the final point in the dataset is level with the initial point (slightly below). So it is, in fact, unchanged. You can quibble about smoothing methods, how to fit a regression, etc., but that is the literal english definition of 'unchanged'.

Is that defended enough for you? Right now the sea level in the Marshall Islands is roughly where it was 23 years ago. That would be UNCHANGED.

You ignored my point entirely.

I agree with you that the current sea level is about where it was 23 years ago. I don't agree with you that the graph you linked isn't trending upwards. Obviously if you have a sinusoid that slowly trends up (where the yearly delta is << the (max-min) periodic variation), it's going to oscillate in such a way that a point-in-time view of it might see lower values than the average. That's what sinusoids do.

(If it's not obvious to you that the trend, rather than the current value, is the interesting data, then this whole discussion is a waste of time.)

So now that we've gotten that out of the way:

I'm arguing that you're doing a really bad job of convincing anyone because of your hostility to rational questions and discussion on the subject. I'm doing this a) to inform you, in case it wasn't obvious to you (and it seems like it wasn't) and b) because it's fun to point these things out; Hacker News seems to often enjoy meta-discussions about arguments.

>I agree with you that the current sea level is about where it was 23 years ago.

So you'd agree that it is, um, how should I phrase this... unchanged??

Why then did you explicitly call me out, and call me a crackpot, for not defending that claim? If you agree with it?

As it turns out, a single sentence out of a comment is actually not the entire content of the comment. It's very disingenuous to cherry pick a sentence and pretend as though it's the entire point.

If you want to be utterly pedantic:

If I or anyone else claims the sea level is 'unchanged', what I and anyone else means is that 'the average sea level over a regular oscillation' is unchanged. This is clearly false; the average has increased.

No one - not me, not you, not climate scientists, not hacker news posters - means that "The sea level as of December 02 2015 is the same as the sea level as of <date in 1993>". We're interested in averages of the sea level over a year/cycle/whatever unit you want.

So you are still probably incorrect, unless you intentionally twist words against their intention. If you do that, there is no point in arguing with you; you are not interested in debate and are not genuinely trying to find truth. Though without agreeing on a method and timespan of averaging things, I wouldn't swear you're incorrect. My whole point is that your argument is failing because of your hostility to rationality and being challenged on it. Not that you're necessarily wrong. (though eyeballing it definitely suggests that you're wrong.)

> As it turns out, a single sentence out of a comment is actually not the entire content of the comment

the key to gd1's entire world view

Congratulations on scoring points with the choir. Now would you care to say something substantive?

I will have one last go at explaining my position, as simply as I can.

I used the word 'unchanged' to indicate that the sea level in the Marshall Islands is unchanged. It is currently approximately where it was 23 years ago. This is a pretty common usage. If a stock trades up and down all day and ends where it started, you will hear people say 'it ended the day unchanged'. Or 'unch'. Sea levels in the Marshall Islands are currently 'unch'. Over a period of 23 years.

And make no mistake, I acknowledge the worldwide trend in sea level is upwards, the evidence of one tidal gauge doesn't trump the evidence of thousands of tidal gauges. And it has been upwards of course this entire interglacial.

Now, your example of a linear trend superimposed on a sinusoid is a good one. So let me say this, it's a pretty fucking pathetic linear trend if, 23 years later, the bottom of the sinusoid is still overlapping the starting point. The noise in the measurement, and the seasonality, is greater than the supposed trend. It's worth considering for example the coastline of Manhattan 100 years ago. This trend you speak of is utterly insignificant on a human timescale. We can literally grow land faster should we so desire. For example, consider the fact that many Pacific islands are growing due to sediment accumulation (http://www.bbc.co.uk/news/10222679), so weak is this trend. And yet, we're expected to sit here and swallow the premise of this article, the concept of 'climate refugees' and compensation...

So, I'll reply to your argument and then make a point about the meta argument.

So either:

you believe that comparing a point-in-time value today to a point-in-time value 23 years ago is a legitimate way to talk about sea levels being 'unchanged' (it's not)

you believe that the ~10cm change in the average (over various periods) is a legitimately 'unchanged' mean, in that it lies with, say, a couple standard deviations around the mean and is therefore not indicative of a trend.

The second is a legitimate stance, so I'll respond to it: I'm pretty sure your intuition is misleading you badly here. First, note that the existence of the islands and their habitability depends on the high point, not the average or the current value (okay, that's my intuition at least). The high point in 1992-95 was about 1cm; the high point in 2012-2015 was about 20cm. [I chose those bounds to try to avoid the dip in 1998 that we may be entering again - not sure how to interpret that without more historical data.] Second, they only have about 180cm of leeway, so that's ~1/10th of the way there. On a scale of 0-180, I think a shift of about 20 over 23 years is significant.

Meta point:

The whole reason I initial responded to you was that you were badly failing to argue your point. You wrote a couple posts above "Why then did you explicitly call me out, and call me a crackpot, for not defending that claim? If you agree with it?". Do you not see the difference? I called you a crackpot precisely for not defending it, and just insisting it was true several times despite challenges. It has nothing to do at all with whether I agree with it. (I don't, actually, but as above - that seems to be a difference of interpretation. I would need more data to be confident either way though.)

What are you basing your numbers on? From a quick look at the raw data [1], I see the high point in the 1992-95 timeframe as 13.74cm (early 1995). The high point in 2012-15 was 21.55cm (mid 2012). This is a difference of 7.81cm. The standard deviation among yearly high points is 5.62cm across all years, so this is not even close to being a significant change. If you look at the top 5 yearly high points starting with the largest, they occurred in 2007 (26.76cm), 2011 (21.79cm), 2012 (21.55cm), 2010 (18.56cm), and 2014 (18.24cm). Note the 2014 value is over 8cm less than the one in 2007.

Look, I get that living right at sea level isn't exactly ideal, and these people might need help. But let's not make more of this than there is.

EDIT: Also note the big difference in granularity across years. Most complete years had 36-37 measurements. 1992 only had 2 and 1995 had only 31. Thus it's quite possible (even likely) that timeframe's real high point was considerably higher than what is recorded.

[1] http://sealevel.colorado.edu/cgi-bin/table.cgi?q=content%2Fi...

Ah, I didn't see that there was a table. I was eyeballing it and did a pretty bad job. I meant to go up to that peak a bit after the 94 marker. Not the point really, though. The point is that there's a visibly larger difference in the first few years compared to the last one, if we ignore one giant dip in each, which seems as though they're an external effect. But I really don't have a strong stance about the interpretation except to say that 'you can read it in multiple ways depending on how you do it'.

This seems to be quite a change from calling gd1 a crackpot simply because his or her interpretation differs from yours.

I never did that. I called him a crackpot because his defense of his opinion was insisting that it was true, switching datasets, and avoiding challenges. Those are crackpot behaviors. I invited him to do a better job defending his point. Haven't I explained that enough times by now?

> (If it's not obvious to you that the trend, rather than the current value, is the interesting data, then this whole discussion is a waste of time.)

Both are interesting. The only thing that isn't is your needless snark. Yes, a linear regression fit shows a slight upward trend, but the current value is also worth noting because the whole point of the article is that the Marshall Islands are being submerged. If the sea level is at the same point it was 23 years ago, then it seems there are three possibilities to consider:

1. The islands were similarly submerged 23 years ago.

2. The sea level data is somehow inaccurate.

3. Sea levels are at the same point, but geological facters may be causing the islands to sink.

What's wrong with snark?

You left out a possibility:

1. The islands were similarly submerged 23 years ago. [close to submerging, but sure]

2. The sea level data is somehow inaccurate. [always must be included]

3. Sea levels are at the same point, but geological factors may be causing the islands to sink.

4. The sea levels are generally higher, but are currently in a lower point in their regular cycle - similar to what happened in 1998.

Now, I won't make any claim either way about this without doing more research. But it should be abundantly clear that the value oscillates (with several effects that have different periods) and therefore you must include those in your list of possible explanations. And also that the instantaneous value, taken in a vacuum (ignoring all predictions we can make from the periodic effects), is utterly useless for predicting what will happen in the next 1-50 years.

I think everyone agrees that the sea level is oscillating around a positive trend. The disagreement seems to be about: a) how steep it is, and b) whether the rise poses an immediate or serious problem for humanity.

gd1 posted the actual empirical sea level data which shows clearly that sea level is not rising as significantly in this region as is claimed. You then called him a crackpot, an ad hominem response. Your denialism in the face of actual scientific factual data is more likely to be the crackpot science-denialist position.

a) you don't win points for using the buzzphrase "ad hominem". fyi, the capital-L-capital-F as-taught-in-high-school Logical Fallacy 'ad hominem' refers to insulting a person in lieu of an actual argument. Hopefully we can agree that I've in fact argued my point in detail.

b) I argued that:

1. he is behaving like a crackpot, for a commonly-accepted understanding of the word 'crackpot'

2. that he has not adequately defended his position against challenges

3. that the data does not show what he claims ('refuting' and 'denying' are very different things)

Now, I agree that the data he posted " shows clearly that sea level is not rising as significantly in this region as is claimed". You'll hear no objection from me. Fortunately, that is not the claim gd1 made, nor the claim I attempted to refute.

And I stand by my (admittedly easier-to-defend) position that: my post was pointing out that he's getting harangued for defending his point abysmally. I don't really want to take a side in the actual debate, except to say that the position that the sea level HAS changed is very clearly legitimate, and if we're going to quibble over changed vs unchanged we have to agree on definitions and methodology.

Calling someone a crackpot is clearly an ad hominem argument. Claiming it is not in a followup, and then furiously downvoting is neither a valid nor legitimate response.

Hm. I think you're confused about what 'ad hominem' means. That or you're struggling to read more than one sentence at a time as a coherent thought. Do you realize that every sentence past the one with the word 'crackpot' in it was a justification of that statement?

https://en.wikipedia.org/wiki/Ad_hominem says "... is an attack on an argument made by attacking the character, motive, or other attribute of the person making the argument, rather than attacking the argument directly".

So can we agree, at least, that I have attacked the argument directly, at length, and therefore the "rather than..." clause is not satisfied?

Moreover, even if you don't agree with that: I have no obligation not to attack a person directly if it gets the point across faster. I am, after all, writing to the readers as much as gd1 - and if gd1 is truly irrational then I'm not particularly interested in trying to convince them of anything, nor avoiding offending them. And my attack (if you can call it that - I did say "Really hard to not take you for a crackpot", which intentionally gives him a wide benefit of the doubt) was, I think, well-justified:

The definition of the word "crackpot" that I am reaching for is that of "a person who believes absurd notions". That would be as opposed to "a person who believes rationally justifiable notions". I called gd1 a crackpot because as a way of capturing the notion that they were not defending their argument via rationality, but by insistence that it was correct, and were (as I pointed out in my original post) intentionally evading challenges to the argument - but responding to things to that did not challenge it. These to me suggest that they had no ability to defend the argument (and I invited a proper defense), and therefore that they would be a person who believes an argument without rational defense, aka a crackpot.

So I posted the argument that "gd1 is behaving like a crackpot for reasons A, B, C, D", and I listed the reasons. We can substitute the phrase "one who believes absurd notions" for "crackpot". Both the word and its substitution are derogatory terms - we can't get around that. It's a negative quality to believe absurd things.

"a person who believes absurd notions"

A person directly citing unbiased factual sources containing empirically measured data does not meet this criteria.

The data he cites refutes the claim he made so ...

You have a lot of aliases. Consider consolidating them.

He noted that the specific claims regarding amount of rise over time were not validatable from evidence, and he posted specific reliable credible third party sources that validated his claim.

Yet here we are several hours later, long enough for every last person to review the data and still many aliases claim it is not so. A notable exclusion is these person's lack of data, but rather reposted claims supposing to refute empirical data, while offering no sources whatsoever. What may we then conclude about those posters other than that they are unreliable sources of data who do not respect the scientific method and actual verifiable data?

What do you mean by 'aliases'? Do you think those posters are me? Should I be wondering if you, kevinskii, and gd1 are the same person?

S/he didn't post anything that validated his claim. S/he posted additional data that continued to leave it contentious, and then pretended like it validated the claim. How is that not clear? The whole point of this is that not defending a claim is the opposite of defending it.

No. You called gd1 a crackpot because that's an easy way to gain karma when people express views contrary to the HN consensus.

No, I didn't do it for that reason. Does karma do anything? I wrote a Chrome plugin that hides my karma score anyway so I'm not sure what it is..

I enjoy meta-arguments and there's a certain satisfaction to trying to set things straight. It devolved into a nightmare though, but it was still decently good practice at arguing.

> and then furiously downvoting

Isn't the downvote karma threshold 750?

I believe it's 500.

Hm, to me it looks as if there was a slight upwards trend in the data. Some fitting would have to go on to see if that's true; however, most of the time, fitting by approximating by hand is not so bad. Look at the smoothed data too, it's more obvious there.

Also, mind that the scale is in cm and numbers given for sea level rise up to now is in the mm range. We can argue about the clear drop in 2014. And I wonder what happened in 1998. Good source, thank you!

edit: Stopped wondering about 1998, it's clearly a strong El Niño signal, and the theory fits as well. Thus, it will probably go back up after next year.

I applied a polynomial trend line to the data: seems there is a 7.5cm rise over 23 years, but most of that was in the first half and has tapered off and may actually be about to head down. I'd like to see longer-term data.

Recurring issue: climate changes. Sea levels change. Land altitudes change. Overlapping cycles interact to great effect covering hours & millennia. And when you're talking about all of these happening in regards to someplace that is no more than 2 meters above "sea level", with nearly a meter of tidal variation daily, it's not unreasonable to conclude "living on the edge of survivability isn't a good choice".

Sadly, it isn't a choice in many cases.

I'd like to see a polite HN discussion about this notion. To wit: in light of a looming existential crisis where circumstantial (mostly financial) difficulties are pitted against prospects of death/destruction, should one be resigned to one's "fate" vs "do anything necessary to move".

The data might be noisy, but what is decidedly not noisy is the intrusion of sea water into the groundwater on those islands. These people used to keep vegetable gardens. Now they have to raise them and rely on rainwater tow after them. On these atoms you can see lots of palm tree stumps where the rising salinity killed them off one by one.

And the Maldives, supposedly in a similar situation, have been "disappearing" since the early-mid 1800s (and probably earlier).

https://pbs.twimg.com/media/CVGSWCvUEAAF4pR?format=png&name=...

I don't know how you're reading that chart, but it's pretty clearly trending up, unless you're only looking at the past year or so of data points.

What smoothing are you recommending, so that we can see that it's unchanged? Hard to see anything in the raw chart.

Edit: as you can see below, what I'm contesting is the assertion that it's unchanged.

The article claims that it has increased by a foot, which is 30 centimeters:

" Changing global trade winds have raised sea levels in the South Pacific about a foot over the past 30 years"

What smoothing do you recommend I apply, so that I can see that?

That graph pretty clearly shows 10 or 15 cm over the last 20 years. I wish I could see the data for 30 years but it doesn't seem impossible that there was a 15 cm rise in those 10 years.

http://sealevel.colorado.edu/content/interactive-sea-level-t... shows about a 9 cm increase over 20 years.

http://sealevel.colorado.edu/content/interactive-sea-level-t... does less drastic smoothing, to show a bit more detail.

Maybe that refers to the maxima, which are in the same order of magnitude? Temporary flooding is also described in the article, with the main issue of destroying freshwater supplies and trees. But you're right, that phrasing is a bit weird.

While its fun to make fun of the non-scientific article, I'm actually probing at your claim that it's unchanged. You cited evidence, but it doesn't support your claim.

Most GW people are fooled by randomness. It's just this generations version of how people are going to destroy the world. Every generation needs something to freak out about.

Here's all you have to know about sea level rise...

When land in Ohio is as expensive than land in NYC, we have a big fucking sea level problem. Until then, all the smart and wealthy people in the world aren't concerned and neither should you be.

We'll try the BOM then, that goes back to '93:

http://www.bom.gov.au/ntc/IDO70052/IDO70052SLI.shtml

Try any source you like. Show me something hard to back up "The Marshall Islands are Disappearing"

The chart you linked shows sea levels are rising, but that doesn't seem to match the argument you're making?

Wow. I see - at best - a random walk. A random walk where the noise is about 5 times the amplitude of any signal you might argue is there. I also see the mean tailing off the last three years to be lower than it was in '93. What do you see? I certainly don't see the "The Marshall Islands Are Disappearing". Is this some sort of Rorschach test that I'm failing?

Look at your data. The last time it trended down significantly there was a major weather event (hint: 1997-1998)... The same major weather event is upon us right now.

Anyhow, fit a regression line, and the trend is upward for most parts of the world. Maybe someday we'll return to an ice age, I mean, we were there in some parts of the world 10000 years ago. Right now the data shows we're in a warming cycle. It may be human-caused, it may be forces beyond our understanding (after all, the planet has gone through several heating-cooling cycles), but either way it's happening.

Except NASA says El Nino warming can cause up to an inch of sea level rise due to thermal expansion, so if you see a downtrend in 1097-8 and this year, it is in spite of weather events, not because of it.

Complex systems can have strange behaviour. I'm just looking at data.

El Nino has changing wind patterns. Wind pushes the ocean water around:

It leads to the Eastern Pacific (America) being lower than the West (Asia) normally (surface wind goes west), normally about ~50cm above mean before Asia and about 20cm below mean before America.

This changes during El Nino events, when the surface wind changes direction and the water is pushed the other way (~20cm below mean before Asia and 20cm above mean before America). Those values include thermal expansion and stuff.

The Marshall islands are more to the Asia side of the Pacific. There is a ~20cm dip in 1998. [1] It fits.

[1] http://sealevel.colorado.edu/content/interactive-sea-level-t...

There's some noise, but the trend seems clear to me. The median starts out oscillating around the '1' line, but by 2008 - 2014 it barely touches the line at all.

There's a drop from 2014-2016, but it's too early to say if it's temporary like the drop between 1997-1998 (El Niño?)

We can say a priori that the sea level is not a random walk. A random walk is boundless - the sea level is not going to suddenly rise 100 meters.

This comment gives a good smoothing in the interactive data set, it might help:

https://news.ycombinator.com/reply?id=10664069&goto=item%3Fi...

from the chart "Maximum recorded level of 2.405 metres at 0430 hours 18/02/2011"

I'm really unclear on how this chart shows rising given the max was in 2011?

A good question, if all you looked at is just one chart. Next, read the scientific literature about how people interpret the chart. In this case, you appear to be confusing climate and weather? If you're looking at an effect on the average of something over decades, the fact that the maximum was 5 years ago is not that significant.

"In this case, you appear to be confusing climate and weather?"

No, I'm just looking at a suggested chart and seeing nothing

Is there a chart that goes farther back with the raw values like this one?

Ah. If you're only looking at the chart, and not the huge body of science that goes with it, then pointing out another chart with a longer baseline isn't going to help you much.

http://sealevel.colorado.edu/content/map-sea-level-trends

Reply on news.ycombinator.com