Ice age to affect Britain within decades?

135

Comments

  • Reply 41 of 82
    xenuxenu Posts: 204member
    A very quick first pass gave me the following



    1. A study has uncovered a change "of remarkable amplitude" in the circulation of the waters of the North Atlantic.



    2. Similar events in pre-history are known to have caused sudden "flips" of the climate, bringing ice ages to northern Europe within a few decades.



    3. The development - described as "the largest and most dramatic oceanic change ever measured in the era of modern instruments", by the US Woods Hole Oceanographic Institute, which led the research - threatens to turn off the Gulf Stream, which keeps Europe's weather mild.



    4. A report by the International Geosphere-Biosphere Programme in Sweden - launched by Nobel prize-winner Professor Paul Crutzen and other top scientists ...



    5. The new research, by scientists at the Centre for Environment, Fisheries and Acquaculture Science at Lowestoft and Canada's Bedford Institute of Oceanography, as well as Woods Hole, indicates that this may already be beginning to happen.



    So, studies have been conducted, results have been found. They have been compared to past events. Correlations have been found. Warnings have been issued. Some people felt these conclusions were important enough to have a Noble prize winner report them.



    Anything else?
  • Reply 42 of 82
    scottscott Posts: 7,431member
    Quote:

    Originally posted by billybobsky

    I beg to differ. By far I think I recieved a better education as a BA in science than I am recieving as a graduate student in science. And I did not take a step down to grad school (nor one up really). Graduate school is the equivalent of a holding pattern, a really crappy holding pattern. Think of it this way Scott, I was introduced to more broader and possibly deeper laboratory experiences in college than I will have been in graduate school simply by the nature of the beast. At least as an undergrad I was able to persue interests that ranged far outside the standard (bio)chemistry regimen and i wasnt inclined to stay in one lab for an extended period of time and so I saw more/understood more that way...



    You do realize that every PhD has a bachelors? Well not every but ... they all have at least the same level of education you do plus additional course work and a dissertation that represents an original contribution to their field. So somehow that equal something less than your degree?
  • Reply 43 of 82
    fishdocfishdoc Posts: 189member
    Scott,



    this is a paper that is being published in Nature. Peer reviewed by world experts. Curry is NOT the sole author, simply the lead author.



    So - we have a respected scientist doing work with other respected scientists, reviewed by experts, but you find the results of the study flawed because a reporter mistakenly thought she was a PhD? Let me tell you from experience, if that is the only mistake a reporter makes in an article, they did a hell of a job (and I speak from experience here). I should note that this is at LEAST her second first-author publication in Nature, a career-making accomplishment for any scientist, PhD or not. The PhD argument here is a distraction and irrelevant.



    The deeper matter - is there global warming and is it anthropogenic, is a question that the vast majority of scientists in the field consider answered. The debate now rages among those politically inclined to not believe it, and an incredibly tiny subset of climatologists and oceanographers.





    Fish
  • Reply 44 of 82
    scottscott Posts: 7,431member
    Oh good it's in Nature. Finally we know where to go to read this. I will read it and report back.





    The PhD thing is relevant to the article. The article was wrong and had other misleading "facts" in it. A Nobel Prize winner said something? It's must be true!
  • Reply 45 of 82
    fishdocfishdoc Posts: 189member
    The article was wrong and had other misleading facts in it? Which are? You never told us.



    You doubt the article because a reporter put PhD in front of a scientist's name who is not a PhD? If that is the best you can do for finding fault with the article, I think this is a losing argument.



    F



    edited to add - "A Nobel prize winner said something"? I suggets you re-read the article. They say nothing I can see about what a Nobel prize-winner said.
  • Reply 46 of 82
    scottscott Posts: 7,431member
    Fishdoc could you please read the thread from the start? What issue of Nature? I just checked the current one and couldn't find it.
  • Reply 47 of 82
    fishdocfishdoc Posts: 189member
    From the WHOI website, you will see it was Dec 17.



    edited ot add - I did read the entire thread, and saw you make no other claim of factual innacuracy in the article, other than your vaguely-worded problem with a Nobel laureate being mentioned.
  • Reply 48 of 82
    scottscott Posts: 7,431member
    OKAY I read the paper. It's not my area so I wont speak authoritatively.



    They did an analysis of ocean water density and salinity of a thin section that stretches between the two poles. They compared data from 1955-1698 to data from 1985-1999. They averaged out the data from column of water at that location. I have no basis to know if this is a reasonable thing to do but the author doesn't provide any for me. It's a short article so maybe there was no room for that. Or maybe it's in a reference.



    There aint no error bars on nothin'. You can't do a paper with tons of measurements and not include error bars. I guess you can because they did. In my experience the reason people leave error bars off is because they either don't have the expertise to work them out of the data OR the bars are large and the authors conclusion would be washed out in the variability.





    They look at difference between the two time periods. One thing about subtracting one datum from another is that it compounds the error. Of which we know nothing about because they didn't report it. As far as I can tell in my quick read.



    favorite quote



    Quote:

    Although multiple factors have been implicated in these longterm changes, the available measurement record has not been sufficient to quantify their relative contributions to the observed trends, and that partitioning remains a high priority in ongoing research.



    Hummmm? Conclusive!





    There's other problems. The time series analysis is ... inconclusive.



    After reading it I can't see how this paper supports the statements in the news article.
  • Reply 49 of 82
    Quote:

    Originally posted by Scott

    You do realize that every PhD has a bachelors? Well not every but ... they all have at least the same level of education you do plus additional course work and a dissertation that represents an original contribution to their field. So somehow that equal something less than your degree?



    Yes, because course work varies strongly from one institution to another. THere is no McDegree. In addition, many perhaps not all or nearly all undergraduates participate in research that represents an original contribution to a field.



    Regardless, you have a better argument in bringing up issues associated with the failure of climate models to take into consideration all sorts of factors known to affect climate change. The models are incredibly complex and have been developed for years. They are still a gospel and i think most climatologist realize this.
  • Reply 50 of 82
    scottscott Posts: 7,431member
    Here's another problem I have. The authors don't reference any of their previous work. Let's say this is a short article loaded with results and short on methods ... where's the corresponding paper in other journals to provided the details?
  • Reply 51 of 82
    Quote:

    Originally posted by Scott

    There aint no error bars on nothin'. You can't do a paper with tons of measurements and not include error bars. I guess you can because they did. In my experience the reason people leave error bars off is because they either don't have the expertise to work them out of the data OR the bars are large and the authors conclusion would be washed out in the variability.



    Granted error bars are important in most things, but error is not necessarily meaningful in all contexts as you should know. I dont have access to nature at home but if this debate is still on tomorrow I will read the article.
  • Reply 52 of 82
    fishdocfishdoc Posts: 189member
    Wow...where to begin....



    "OKAY I read the paper. It's not my area so I wont speak authoritatively. "



    And yet you start your "critique" nontheless.



    "They averaged out the data from column of water at that location. I have no basis to know if this is a reasonable thing to do but the author doesn't provide any for me. It's a short article so maybe there was no room for that. Or maybe it's in a reference."



    Actually, both. First, if you were familliar with the science, you would know that this averaging through the water column is a commenly accepted way to do this type of analysis (read the references to see how common this is). Also, it is up to the reader to decide if this mehtodology is appropriate - that is why it is described.



    "There aint no error bars on nothin'. You can't do a paper with tons of measurements and not include error bars."



    This is silly - the vast majority of the graphs presented, first of all, are vertical sections showing salinity data - since the x and y axes are depth and latitude, where do you propose they put the error bars? Again, not meaning to be mean, but you are demonstrating more aboujt how unfamiliar you are with the field than actually presenting a criticism that is valid.



    " because they either don't have the expertise to work them out of the data OR the bars are large and the authors conclusion would be washed out in the variability."



    Again, this is a reflection on you, not the Nature paper.





    "They look at difference between the two time periods. One thing about subtracting one datum from another is that it compounds the error. Of which we know nothing about because they didn't report it. As far as I can tell in my quick read."



    Good lord - they were analyzing change over time, so it is the differential between the two time periods that is the parameter of interest. This is a flaw?



    "the authors dont reference their previous work".



    They explained all of the methods that they needed to - they presented references wherever they were appropriate. Truly, the only reason you are confused is because you do not understand ther field, or you did not read the article all the way through.



    For example, you criticize their lack of dealing with error, and yet in the methods section they describe in detail how error was calculated, and add "All of the upper-ocean signals reported here are an order of magnitude greater (0.1?0.4) than the worst-case error; and the deep-ocean anomalies exceed twice the measurement uncertainties"



    "After reading it I can't see how this paper supports the statements in the news article"



    Well, that doesn't surprise me.





    Fish
  • Reply 53 of 82
    scottscott Posts: 7,431member
    Quote:

    Originally posted by fishdoc

    Wow...where to begin....



    ...



    Actually, both. First, if you were familliar with the science, you would know that this averaging through the water column is a commenly accepted way to do this type of analysis (read the references to see how common this is). Also, it is up to the reader to decide if this mehtodology is appropriate - that is why it is described.




    Maybe so.







    Quote:

    Originally posted by fishdoc

    This is silly - the vast majority of the graphs presented, first of all, are vertical sections showing salinity data - since the x and y axes are depth and latitude, where do you propose they put the error bars? Again, not meaning to be mean, but you are demonstrating more aboujt how unfamiliar you are with the field than actually presenting a criticism that is valid.



    Mostly I was thinking about the text. If I say that I measured something to be X then X has an error associated with it. If one is doing a 2D color plot they could also include an additional plot of the error or mention what the error is.







    Quote:

    Originally posted by fishdoc

    Again, this is a reflection on you, not the Nature paper.



    Um? No. I might be wrong about why they didn't provide any error analysis but it not being their is a reflection on them and not me.





    Quote:

    Originally posted by fishdoc

    Good lord - they were analyzing change over time, so it is the differential between the two time periods that is the parameter of interest. This is a flaw?





    Not it's not a flaw but if you want to think about it scientifically you have a hypothesis to test. You might take the null hypothesis that there is no change over time. Then prove that there is significant evidence to show that's not the case to a certain p value. But that would require more statistical analysis.



    Quote:

    Originally posted by fishdoc

    They explained all of the methods that they needed to - they presented references wherever they were appropriate. Truly, the only reason you are confused is because you do not understand ther field, or you did not read the article all the way through.



    I don't agree.



    Quote:

    Originally posted by fishdoc

    For example, you criticize their lack of dealing with error, and yet in the methods section they describe in detail how error was calculated, and add "All of the upper-ocean signals reported here are an order of magnitude greater (0.1?0.4) than the worst-case error; and the deep-ocean anomalies exceed twice the measurement uncertainties"



    They don't state at all how they arrived at the error analysis. The just state it. Do you really thing the error in the estimated temperatures back in the 50s and 60 was a hundredth of a degree Celsius? Average over a column of water? It just can't be. Maybe it's in the other paper about the software package? Plus when reading it It's hard to tell which error goes with what and then I can go back my original gripe that when you mention a difference then error is then compounded. It's basic stats.





    Quote:

    Originally posted by fishdoc



    Well, that doesn't surprise me.





    Fish




    The paper had nothing to do with the weather in Europe. Or did I miss something?
  • Reply 54 of 82
    scottscott Posts: 7,431member
    Here's another problem with it. They showed a time series for one latitude. One Could be typical or it could be data mined. If you look at 20 "things" one will be significant (on average). Not because it is but because the statistics demand that it is. If you set your confidence level at 95% then 5% or your normals will come across as significant. With and ocean full of data to look at it's not hard to find "significant" results. It's the plague of "cancer hot spots".



    Sloppy stats? Just say no.
  • Reply 55 of 82
    Actually, you missed quite a bit. Let me try to sumamrize for you - I will not bother with a point-by-point, because your last post demonstrates that is a waste of time.



    1. Read the original article that started this thread - it does not state that the Curry study declares an ice age in europe. Read tha article, and see who is saying what (hint - there is more than one report mentioned in the original link). Again, it is your poor reading/undertsanding that is the problem.



    2. I am not sure how much clearer to state it - they described their methods. They even provided references to the databases they used.



    None of your criticisms are substantive. Perhaps there ARE substantive complaints that could be made about this study, but it is now clear that you are not competent to do so (again, not meant as an insult, but ou clearly are not a scientist who deals with these types of data, and you seem surprisingly puzzled by Nature's format as well).



    Again - read the original article, read the Curry article, and get over the reporter calling her "Dr", because it is truly irrelevant.



    Edited to add...



    A. It is longitude you are thinking of, not latitude.

    B. It was NOT a single longitude either, but a transect that spanned many degrees.

    C. They EXPLICITLY explained how the transect was chosen (your oh-so-careful reading must somehow have missed that).

    D. I find it offensive that someone who clearly doesn't even understand a paper calls the Nature paper's statistics "sloppy". And 95% confidence levels don't mean what you think they mean, but that is a more subtle problem you have.

    E. I truly don't think you understand this enough to discuss it, and I am wasting my time to argue with tese silly points you are making. If you sincerely want to learn more about this, I suggest just trying to read up more on oceanography (the Open University has some great books).







    F
  • Reply 56 of 82
    bungebunge Posts: 7,329member
    Quote:

    Originally posted by fishdoc

    Let me try to sumamrize for you - I will not bother with a point-by-point, because your last post demonstrates that is a waste of time....and I am wasting my time to argue with tese silly points you are making.



    It's definitely not a waste of time. Some of us are getting a good informative laugh out of it. Keep up the good work.
  • Reply 57 of 82
    scottscott Posts: 7,431member
    Quote:

    Originally posted by fishdoc

    ...

    A. It is longitude you are thinking of, not latitude.




    I stand corrected. I've had a few beers.



    Quote:

    Originally posted by fishdoc

    B. It was NOT a single longitude either, but a transect that spanned many degrees.





    A single section so sue me.



    Quote:

    Originally posted by fishdoc

    C. They EXPLICITLY explained how the transect was chosen (your oh-so-careful reading must somehow have missed that).



    It's the maximum? Why not the minimum or the average or the median?



    Quote:

    Originally posted by fishdoc

    D. I find it offensive that someone who clearly doesn't even understand a paper calls the Nature paper's statistics "sloppy".



    I didn't call is sloppy. I commanded that we say no to sloppy stats.



    Quote:

    Originally posted by fishdoc

    And 95% confidence levels don't mean what you think they mean, but that is a more subtle problem you have.



    Oh do tell.



    Quote:

    Originally posted by fishdoc

    E. I truly don't think you understand this enough to discuss it, and I am wasting my time to argue with tese silly points you are making. If you sincerely want to learn more about this, I suggest just trying to read up more on oceanography (the Open University has some great books).



    F




    I've read something written much better than this. But they only have four pages. Were they limited by the editor? I've never written for nature so I don't know the page limit on letters.
  • Reply 58 of 82
    haraldharald Posts: 2,152member
    Quote:

    Originally posted by Scott

    Dire predictions by people who must know better than us because some reporter stuck "Dr" in front of someone's name



    Yep.



    People who claim doctorates to add to their credibility when they are clearly hopelessly underendowded for the honour are scum. Idiots, liars and self-delusionists. I've always said that.



    Right on, Scott.
  • Reply 59 of 82
    scottscott Posts: 7,431member
    Quote:

    Originally posted by Harald

    Yep.



    People who claim doctorates to add to their credibility when they are clearly hopelessly underendowded for the honour are scum. Idiots, liars and self-delusionists. I've always said that.



    Right on, Scott.






    That's not what I said an you should know that because the quote is right there in your post. Why not post something more intelligent?
  • Reply 60 of 82
    OK ? for the sake of those who might mistakenly think you are making relevant points, I will continue to correct you?





    Originally posted by fishdoc

    ...

    A. It is longitude you are thinking of, not latitude.



    Scott:

    I stand corrected. I've had a few beers.



    Fish:

    Thank you for conceding the point. Maybe those beers explain the rest of your misconceptions.



    quote:

    Originally posted by fishdoc

    B. It was NOT a single longitude either, but a transect that spanned many degrees.





    Scott:

    A single section so sue me.





    Fish: ?section?? Transect.



    quote:

    Originally posted by fishdoc

    C. They EXPLICITLY explained how the transect was chosen (your oh-so-careful reading must somehow have missed that).



    Scott:

    It's the maximum? Why not the minimum or the average or the median?



    Fish: are you even reading this article? ?It?s the maximum?? It is not the maximum ? they say ?This line was chosen with some care. It spans the maxima and minima of E?P in both hemispheres?. And in the same paragraph they say :our transect crosses the regions of maximum salinity in the subtropics of both hemispheres as well as the surface salinity minima along the Equator and at both poleward ends of the line.?



    I am inclined to believe that you either have no understanding at all of the article or you are only skimming it. Now, I will leave it as an exercise for you ? why would they want a transect with these properties if they are looking at a change in the freshwater balance of the Atlantic Ocean? It should be reasonably clear.





    quote:

    Originally posted by fishdoc

    D. I find it offensive that someone who clearly doesn't even understand a paper calls the Nature paper's statistics "sloppy".



    Scott:

    I didn't call is sloppy. I commanded that we say no to sloppy stats.



    Fish: Ah, then I suppose that was just an unrelated comment. My bad. I assumed you were talking about this paper.



    quote:

    Originally posted by fishdoc

    And 95% confidence levels don't mean what you think they mean, but that is a more subtle problem you have.



    Scott:

    Oh do tell.



    Fish: Well, okay?..?confidence level? refers to the probability associated with a given confidence interval. And you are right that it is often 95% (1 ? alpha). But again, that refers to the confidence interval, so let?s think about what that means.



    Confidence intervals are used to state confidence in estimates of parameters - most commonly, some measure of central tendency like the mean or median. So you do not get ?5% of the normals coming across as significant?, because it measures our confidence in our estimate of the mean, not of individual values. The confidenc einterval should be interpreted (strictly) to mean ?if many samples are collected and the confidence interval computed, in the long run about 95% of these intervals would contain the true mean.? This is not (quite) the same thing as ?I am 95% certain this interval contains the true parameter?, and not even close to your ?95% of the data points will fall within this interval?.



    quote:

    Originally posted by fishdoc

    E. I truly don't think you understand this enough to discuss it, and I am wasting my time to argue with tese silly points you are making. If you sincerely want to learn more about this, I suggest just trying to read up more on oceanography (the Open University has some great books).





    Scott:

    I've read something written much better than this. But they only have four pages. Were they limited by the editor? I've never written for nature so I don't know the page limit on letters.



    Fish: I do not doubt that you have not written for Nature. However, on their web site they have ?instructuions for authors?, so you can see what the page limits are for each type of submission if you are interested.



    Finally, I note that you didn?t respond to my comment on your complaints about not seeing mention of weather in Europe in the Curry paper ? do you see where you were mistaken there?



    Really ? if this paper is so full of holes, I am certain other scientists, scrambling to get their names in Nature, published replies in the next issue, correcting Curry et al, right?



    Right?





    Oh?I guess not.



    Again, just as I would be reluctant to publicly criticize an ee paper, since I have little understanding, I suggest you learn a little humility when trying to critique something so far out of your depth.





    Fish
Sign In or Register to comment.