Prev  1  2  3  Next

Comments 26 to 50 out of 57:

26. daniel at 22:26 PM on 28 June, 2010
    Doug why don't you explain the detail required? I have provided all the detail nessecary. A short term trend of ~150 years, measured using direct measurement techniques at high resolution is compared to a general linear trend over ~550 years approximated using 10 paleo data points with clearly alot of uncertainty in each. There is no statistics to "undo" can you explain what you mean by that? I doubt only the conclusions section of the paper. 10 noisy data points do not adequately measure sea level trends at the 150 year time scale between 1300-1850 AD. Therefore it is invalid to conclude there has been a significant recent increase in seal level rise. As I mentioned in my previously deleted comment both you and peter have put words in my mouth by claiming I doubt the methods or error analysis of the Donnely paper or reffered to a "downturn" in recent sea level trends (good one pete). You're nit even listening to what I am saying. Which is a reflection on your credibility not mine.
27. Doug Bostrom at 01:02 AM on 29 June, 2010
    Let me explain why you need to do the work or must instead pipe down, Daniel. Any bachelor degree graduate can ... The Donnely paper...is simply an utter joke. Scientists who after having found the time and inclination to review the data of climate scientists are utterly apalled at the conclusions drawn. Those words are what is known as the "Badge of Hubris." You have said you believe the work you're criticizing to be defective, you have said that "any bachelor degree graduate" can show why. You have spoken for a number of scientists you claim are appalled. The problem is, you've not yet earned your Badge of Hubris because you have not shown exactly what is wrong with the authors' conclusions. Where are their ranges of uncertainty incorrect, how did they calculate that incorrectly, for instance? "I doubt it" is not an argument that can earn you the Badge of Hubris. Your demand that I supply the detail required to make your argument earns a Badge of Comedy. You're the person flinging assertions, support 'em or pipe down. Simple enough, right?
28. Doug Bostrom at 01:45 AM on 29 June, 2010
    Amending my earlier remarks to Daniel or that is to say supplying necessary detail. Daniel, you're dismissing research performed by workers practicing in several disciplines entailing a myriad of details and knowledge. I'm going to take a flying leap of speculation and say that I don't think you yourself practice in those disciplines. My speculation is informed by my observation that beyond not addressing even such simple matters as the calculation of uncertainties of the conclusions formed in Donnelly, you have not attacked in detail the disciplinary practices employed in selecting and analyzing the samples you sneeringly dismiss as inadequate. If I'm correct and you indeed do not practice in at least the core discipline producing the research you are dismissing, you have an enormous amount of work to to perform before you are capable of usefully critiquing research output in that discipline. That is, unless you are very lucky and find a blatant blunder in the work, and you've not yet even bothered to test your luck. You want me to help you with your critique, but I'm not so arrogant as to imagine I can suddenly take on the attributes of a person who has earned an advanced degree in geology with a strong bent to a narrow part of that discipline involving a wealth of arcane information. Not as an "appeal to authority" but as an indicator of how much work you need to do before you're suitable for producing useful critique of work produced in a specialty discipline, let's take a brief look at a bit of Jeffrey Donnelly's CV. Ask yourself, do you -really- believe you're as well informed on his discipline as is he? B.S. Earth Science, University of Massachusetts M.E.S. Coastal and Watershed Systems, Yale University Ph.D. Geological Sciences, Brown University Mann, M.E., J.D. Woodruff , J.P. Donnelly, and Z. Zhang, submitted, El Nino, Tropical Atlantic Warmth, and Hurricanes Over the Past 1500 Years, Nature. *Boldt, K.V., P. Lane, J.D. Woodruff, and J.P. Donnelly, submitted, Sedimentary evidence of hurricane-induced coastal flooding in southeastern New England over the last two millennia: Geophysical Research Letters. *Newby, P., J.P. Donnelly, and B.N. Shuman, 2009, Evidence of centennial-scale drought from southeastern Massachusetts during the Pleistocene/Holocene transition: Quaternary Science Reviews (in press). Shuman, B.N., P. Newby, and J.P. Donnelly, 2009, Abrupt Climate Change as a Catalyst of Ecological Change in the Northeast U.S. throughout the Past 15,000 Years: Quaternary Science Reviews (in press). Madsen A.T., G.A.T. Duller, J.P. Donnelly, H.M. Roberts, and A.G. Wintle, 2009, A chronology of hurricane landfalls at Little Sippewissett Marsh, Massachusetts, USA, using optical dating: Geomorphology (in press). *Woodruff, J.D., J.P. Donnelly, and A. Okusu, 2009, Exploring typhoon variability over the mid-to-late Holocene: Evidence of extreme coastal flooding from Kamikoshiki, Japan: Quaternary Science Reviews (in press). *Woodruff, J.D., J.P. Donnelly, K. Emanuel, and P. Lane, 2008, Assessing sedimentary records of paleo-hurricane activity using modeled hurricane climatology: Geochemistry, Geophysics, Geosystems v. 9, Q09V10, doi:10.1029/2008GC002043. Donnelly, J.P., and L. Giosan, 2008, Tempestuous Highs and Lows in the Gulf of Mexico: Geology, v. 36, p. 751-752. *Woodruff, J.D., J.P. Donnelly, D. Mohrig, and W. R. Geyer, 2008, Reconstructing relative flooding intensities responsible for hurricane-induced deposits from Laguna Playa Grande, Vieques, Puerto Rico: Geology, v. 36, p. 391-394. *Carlson, A., J. Stoner, J.P. Donnelly, and C. Hillaire-Marcel, 2008, Response of the southern Greenland Ice Sheet during the last two deglaciations: Geology, v. 3 6, p. 359-362. *Ashton, A., J.P. Donnelly, and R. Evans, 2008, A discussion of the potential impacts of climate change on the shorelines of the Northeastern USA: Mitigation and Adaptation Strategies for Global Change, doi: 10.1007/S11027-007-9124-3. Cheung, K. F., L. Tang, J.P. Donnelly, E. Scileppi, K. Liu, X. Mao, S.H. Houston, and R.J. Murnane, 2007, Coastal Overwash Modeling in Paleotempestology: Journal of Geophysical Research, v. 112, F03024, doi:10.1029/2006JF000612. Donnelly, J.P., and J.D. Woodruff, 2007, Intense hurricane activity over the past 5,000 years controlled by El Nino and the West African monsoon: Nature, v. 447, p. 465-468. Hill, J.C., N.W. Driscoll, J. Brigham-Grette, J.P. Donnelly, P.T. Gayes, and L. Keigwin, 2007, New evidence of very high discharge to the Chukchi shelf since the Last Glacial Maximum: Quaternary Research, v. 68, p. 271-279. *Scileppi, E., and J.P. Donnelly, 2007, Sedimentary Evidence of Hurricane Strikes in Western Long Island, New York: Geochemistry, Geophysics, Geosystems, v. 8, Q06011, doi:10.1029/2006GC001463. Thieler, E.R., B. Butman, W.C. Schwab, M.A. Allison, N.W. Driscoll, J.P. Donnelly, and E. Uchupi, 2007, A catastrophic meltwater flood event and the formation of the Hudson Shelf Valley: Palaeogeography, Palaeoclimatology and Palaeoecology, v. 246, p. 120-136. Keigwin, L., J.P. Donnelly, M.S. Cook, N. Driscoll, J. Brigham-Grette, 2006, Rapid Sea-Level Rise and Holocene Climate in the Chukchi Sea: Geology v. 34, p. 861-864. Donnelly, J.P., 2006, A Revised Late Holocene Sea-Level Record for Northern Massachusetts, USA: Journal of Coastal Research, v. 22, p. 1051-1061. Giosan, L., J.P. Donnelly, S. Constantinescu, F. Filip, I. Ovejanu, A. Vespremeanu-Stroe, E.Vespremeanu, G.A.T. Duller, 2006, Young Danube delta documents stable Black Sea level since the middle Holocene: Morphodynamic, paleogeographic, and archaeological implications: Geology, v. 34, p. 757-760. Shuman, B., and J.P. Donnelly, 2006, The Influence of Seasonal Precipitation and Temperature Regimes on Lake Levels in the Northeastern United States during the Holocene: Quaternary Research, v. 65, p. 44-56. Donnelly, J.P., 2005, Evidence of Past Intense Tropical Cyclones from Backbarrier Salt Pond Sediments: A Case Study from Isla de Culebrita, Puerto Rico, USA: Journal of Coastal Research, SI42, p. 201-210. Giosan, L., E. Vespremeanu, J.P. Donnelly, J. Bhattacharya, and F. Buonaiuto, 2005, Morphodynamics and evolution of Danube delta: Journal of Sedimentary Research, SP 83, p. 391-410. Shuman, B., P. Newby, J.P. Donnelly, A. Tarbox, and T. Webb III, 2005, A Record of Late-Quaternary Moisture-Balance Change and Vegetation Response from the White Mountains, New Hampshire: Annals of American Association of Geographers, v. 95, p. 237-248. Donnelly, J.P., Driscoll, N., Uchupi, E., Keigwin, L., Schwab, W., Thieler, E.R., Swift, S., 2005, Catastrophic Meltwater Discharge down the Hudson River Valley: A Potential Trigger for the Intra-AllerØd Cold Period: Geology v. 33, p. 89-92. (Research Highlight: Nature 17 February 2005 433: 702) Donnelly, J.P., and Webb III, T., 2004, Backbarrier sedimentary records of intense hurricane landfalls in the northeastern United States. In: Murnane, R. and Liu, K. (eds.), Hurricanes and Typhoons: Past Present and Potential, New York: Columbia Press, pp. 58-96. Donnelly, J.P., J. Butler, S. Roll, Micah Wengren, and T. Webb III, 2004, A backbarrier overwash record of intense storms from Brigantine, New Jersey: Marine Geology, v. 210, p. 107-121. Donnelly, J.P., Cleary, P., Newby, P., and Ettinger, R., 2004, Coupling Instrumental and Geological Records of Sea-Level Change: Evidence from southern New England of an increase in the rate of sea-level rise in the late 19th century: Geophysical Research Letters, v. 31 L05203 doi:10.1029/2003GL018933. Donnelly, J.P., and M.D. Bertness, 2001, Rapid shoreward encroachment of salt marsh cordgrass in response to accelerated sea-level rise: Proc. Nat. Acad. Sci., v. 98, p. 14218-14223. Donnelly, J.P., S. Roll, M. Wengren, J. Butler, R. Lederer, and T. Webb III, 2001, Sedimentary evidence of intense hurricane strikes from New Jersey: Geology, v. 29, p. 615-618. Donnelly, J.P., S. S. Bryant, J. Butler, J. Dowling, L. Fan, N. Hausmann, P. Newby, B. Shuman, J. Stern, K. Westover, and T. Webb III, 2001, A 700-Year sedimentary record of intense hurricane landfalls in southern New England: Geological Society of America Bulletin, v. 113, p. 714-727. (editors’ choice: Science 8 June 2001 292: 1801) Donnelly, J.P., T. Webb III, and W.L. Prell, 1999, The influence of accelerated sea-level rise, human modification and storms on a New England salt marsh: Current Topics in Wetland Biogeochemistry v. 3, p.152-160. Donnelly, J.P., 1998, Evidence of late Holocene post-glacial isostatic adjustment in coastal wetland deposits of eastern North America: Georesearch Forum, v. 3-4, p. 393-400.
29. Doug Bostrom at 11:49 AM on 29 June, 2010
    I've been inspired by Daniel to scrutinize Donnelly et al more closely. One thing to note right away: I'm struck by Daniel's powerful rhetoric (Donnelly's paper is an "utter joke") compare to Donnelly's measured language in his conclusions: The likely increase in the rate of SLR in the late 19th century A.D. is roughly coincident in time with climate warming observed in both instrumental and proxy records [e.g., Mann et al., 1998; Pollack et al., 1998]. The results indicate that this recent increase in the rate of SLR may be associated with recent warming of the global climate system. Daniel might have a point about uncertainties, if Donnelly's stratigraphic conclusions were taken in isolation and were naively dated, and if Donnelly's main objective was to derive a reasonable absolute measure of sea level for each stratigraphic sample. However, Donnelly's dating interpretation of the sequence is consistent with independent markers and in any case Donnelly's objective is not to obtain a series of accurate historical sea levels attributed to particular years over a 700 year period but rather to form an estimate of rate of rise over that entire span. That means that if the samples can be boxed into periods of a few years they're suitable for Donnelly's requirements. It turns out that markers constrain the uncertainties of individual measurements sufficient to eliminate extended large excursions of the type Daniel hypothesizes making it quite unlikely that the bulk of the rise indicated by the sequence was concentrated in short bursts. The various tools deployed by Donnelly to constrain dates are exactly what I'm talking about when I say the neither Daniel nor I are equipped to offer criticism of this work with an eye to disproving it. Here's a typical example: To further refine our C-14 chronology, we used fossil pollen evidence of European clearance/agriculture and industrial revolution-related heavy metal pollution horizons (Figure 3). Peat samples for pollen and metals analysis were also taken from just above the contact with the erratic. The initial rise in Rumex spp. pollen (a native weed) (-46.5 to -50.5 cm) coincides with land clearance for agriculture between 1650 and 1700 A.D. [Clark and Patterson, 1985; Donnelly et al., 2001]. An uncertainty box has been plotted (light gray) based on the presence of Jg and Sp remains at this interval (indicative meaning of 6.7 ± 10.4 cm above MHW) and the time interval of initial land clearance (box with diagonal line fill; Figure 2). The combination of the indicative meaning of the sample (including 2s uncertainty) with its accepted age range yields boxes representing the most likely elevation of MHW in the past (Figure 2). The appearance of Plantago lanceolata (an introduced species) between -32.5 and -35.5 cm (Figure 3) suggests deposition in the early 19th century [Clark and Patterson, 1985]. Based on the presence of Jg and Sp remains at this interval we plotted a box representing the indicative meaning of this interval (vertical line fill) and the associated uncertainty box (light gray) (Figure 2). Other methods were used to constrain other samples, methods of which Daniel and I know nothing. This is what I mean when I refer to unearned hubris; dismissing a paper as "junk" from a position of ignorance of the specialized tools used to produce it is foolish. Daniel complains There is more than enough slack in this data to periodically reproduce the apparently rapid sea level rise of 2.8mm/year in the NYC tide gauge data of the last ~150 years but that's speculation. The conservative way to interpret the data is to take it for what can say. Donnelly: A linear rate of rise of 1.0 ± 0.2 mm/year intersects all the 2 [sigma] uncertainty boxes of the record from the 14th to the mid-19th century. That's it, and in any case we've already seen that wild slews in rate changes don't seem to fit the C14-independent constraints of the samples. Donnelly himself is carefully circumspect about his conclusions: Coupling the Barn Island record and regional tide-gauge data indicates that the rate of SLR increased to modern levels in the 19th century (Figure 2). However, given that the center of each uncertainty box has the highest probability the most conservative interpretation of the data is that the SLR increase to modern values occurred in the late 19thcentury (Figure 2). Daniel needs to do better than Donnelly at performing this same work in order to dismiss Donnelly's paper but he can't because he's not trained in Donnelly's area of specialization. Because Daniel cannot address the paper at this level of detail but can only use general purpose adjectives to support his case, I'm with Donnelly on this matter. What choice do I have? Donnelly makes a reasonable case using tools he describes adequately but which I'm unqualified to judge, as is Daniel. I suggested an attack via uncertainty to Daniel because that's the only technique I could use in this case, lacking the disciplinary tools to address Donnelly's methods as I do, and in all probability that's true of Daniel as well.
30. Peter Hogarth at 06:11 AM on 30 June, 2010
    daniel at 22:03 PM on 26 June, 2010 I did make an assumption and I apologise, I'm not sure where that came from, but it was late! Perhaps "The article claims that skeptics are guilty of interpreting small recent trends from noisy data as significant" figure 1 etc. I assumed this was an oft used reference to the Jason 1 satellite altimeter data showing a decrease in trend a couple of years back (not a decrease in level, it appears we both got the wrong end of the stick), that's what I meant anyway. More relevant to your debate, see Grinsted 2009 which is pertinent to your points on sea level reconstructions. There's a few more I can dig out if of interest and I get time. I'm not sure "Therefore it is invalid to conclude there has been a significant recent increase in sea level rise" is really supportable. Doubt whilst you accumulate more evidence would be ok.
31. Doug Bostrom at 08:16 AM on 30 June, 2010
    Upon further scrutiny of my last post, I think it communicates poorly the conservative nature of Donnelly's interpretation of his samples versus Daniel's assertions concerning hidden slewing. Daniel posits There is more than enough slack in this data to periodically reproduce the apparently rapid sea level rise of 2.8mm/year in the NYC tide gauge data of the last ~150 years (cited and compared to by the authors). Donnelly on the other hand sticks to the available data. Looking at Donnelly's figure 2 where he marries together the various data I think shows how Daniel might be right that while short episodes of discontinuous rise and fall may indeed be invisible, a linear interpretation not only avoids speculating signal features where none can be derived from the data but in fact more likely yields a result that is plausible on its face. Supposing for a moment that we are free to make up data however we please, how exactly would discontinuities of the kind Daniel imagines may have happened actually fit within the constraints of connections between the samples while still connecting to the more recent instrumental record? If I felt free to draw lines wherever I pleased between the samples it might be possible to squeeze in some excursions, but then I'd not only be substituting fiction for reality, I'd almost have to end up with an implausible looking graph, and again I'd have to be creating data to do so. So my conclusion is that Donnelly's more conservative than Daniel. Here's the figure from Donnelly by way of illustration:
32. daniel at 10:49 AM on 30 June, 2010
    Here's a quote from the above comment from doug # 31 "If I felt free to draw lines wherever I pleased between the samples it might be possible to squeeze in some excursions, but then I'd not only be substituting fiction for reality, I'd almost have to end up with an implausible looking graph, and again I'd have to be creating data to do so. So my conclusion is that Donnelly's more conservative than Daniel." You wouldn't be substituting fiction for reality doug :) The error estimates allow you to draw those lines. The fiction comes from believing that given the uncertainty in the data points we can conclude that short term deviations from the proposed trend are non-existent. Look at the uncertainty in time for samples 7 and 8, it's approximately 150 years. That means the authors are saying that the assigned height (which has it's own level of uncertainty) lies somewhere in the range of 150 years (between about 1500-1650). That is the time span of the current instrumental record. That should give you an idea of the vast difference in certainty between the two sets of data. Can you see that sample 11 has two date ranges assigned to it? Does that sound like a high level of certainty to you doug? We can see also that sample 9, which by the authors own admission should be younger than sample 10, has a date range generally older than sample 10. How much of sample 10's 95% confidence interval can actually be so confidently assigned when sample 9's 95% confidence interval is not even as young as that? It's true statistical methods lead to these confidence intervals but then logic needs to be applied before we write our conclusions section. That portion of the graph, 1300-1500 AD, has a lot of potential for a significant deviation from the proposed trend. As does 1600-1750 AD, if we could more confidently assign samples 7 and 8 toward the younger end of their current 95% confidence intervals then a short term trend of much greater than 1mm/year SLR through sample 5-8 could potentially exist. If such deviations from the trend were visible then the recent sea level rise would not be as alarming as is made out to be. These short term rates of paleo sea level rise do not even have to match the 2.8mm/year observed in recent times it only has to be closer to it than the average 1mm/year in order for the recent rise to be less alarming. The low resolution data really undoes the conclusions of Donnely et. al. but we find that although the Gehrels paper tries to address this issue the uncertainties are still too high to obtain a meaningful result. These attempts to measure paleo sea level rise are certainly commendable for the level of effort put in but the conclusions drawn are unsound.
33. Doug Bostrom at 11:41 AM on 30 June, 2010
    Hi Daniel, glad to see you back, I was afraid I was going to have to argue with myself. I thank you for forcing me to take a closer look at Donnelly and strain my eyesight squinting at his graphs. I see your point about samples 7 & 8, I'm sure Donnelly would have been happier if they'd resolved better but because they're embedded in the middle of the series their effect is not very drastic; interpretation of those is constrained by the surrounding boxes. As to your problems with multiple date ranges for samples, if you read the text carefully you'll see how Donnelly eliminated date ranges by using methods beyond C14: In some cases we can use the Principle of Superposition to determine which range most likely represents the age of the sample. For example sample 9 should be younger than sample 10 (since sample 9 was recovered 3.5 cm above sample 10), so we eliminate the two older ranges (1306–1356 and 1357–1365 A.D., gray on Figure 2); the youngest range from sample 9, 1386–1440 A.D., best represents the age of that sample. Other sample ambiguities were treated with different methods appropriate to the individual cases, with the result that multiple date ranges appear to have been eliminated in all cases if I'm reading Table 1 correctly. With regard to drawing a line through the whole collection, if I get you right and correct me if I'm misunderstanding you, you're suggesting that it's equally reasonable to pick and draw a series of lines perhaps pointing up and perhaps pointing down between any chronologically linear pair of samples. That's not as conservative as doing what Donnelly did. As well, doing such a series of arbitrary choices leaves the issue that the entire series must begin somewhere within the region circumscribed by the sample 4 and 11 confidence boxes, meaning that the overall conclusion of the series of choices made to connect individual samples ends up being nearly the same, confined by the beginning and ending samples. Meanwhile, it appears that the slope described by the direct recent tidal measurements is inevitably going to be steeper than the sum linear product of whatever combination of ups and downs you might choose to impose on the paleo series, and as well covers a disproportionate vertical range compared compared to the paleo series. This suggests to me that attempting to create and insert arbitrary additional information into the series is pointless. So again my take is that you're suggesting a liberal interpretation of the data, Donnelly is picking a conservative approach. And I do think neither of us are equipped with the specific skills we need to cast technical judgment on this article, certainly not to fling the term "utter junk" in describing it. The suite of dating refinements employed by Donnelly I refer to are an example our ignorance, as I mentioned before.
34. daniel at 17:19 PM on 30 June, 2010
    Doug you do a lot of commenting on this website so I don't really know how cosy you are with the authors. I have been absent because my comments have been deemed inappropriate by a rather draconian comments policy. I will assume you know little about that but I am suspicious since you continued to argue with me and seemed to address some of the issues I was raising in those comments. I will address your recent comments soon. I have only skimmed over them now. I would like to say that this comments policy is not endearing to the authors of the website. If you deem your opponents comments as uncivilised, off topic, whacky or inappropriate then you can just casually reply saying as much and allow them to discredit themselves as they rant some more. Only truly foul language should be deleted. Explain to me doug why it is that you can use ad hominem arguments against my credibility by saying that I'm not a climate expert and therefore have nothing to say in regards to the quality of work coming from those fields? It really doesn't look good for you guys.
35. Doug Bostrom at 17:40 PM on 30 June, 2010
    Ad hominem, Daniel? Yes, some things concern attributes of individual persons, specifically in this case what they know and don't know. Are you a geologist with an advanced degree specialized in paleochronology, Daniel? Have you spent a significant portion of your life learning how to tease dating information out of stratigraphic sequences? If you can honestly answer "yes" then my comparison of your abilities with regard to paleochronologies with those of Donnelly is less relevant. If you can only answer "no" then your assertion that Donnnelly's paper is "pure junk" is notably arrogant and makes your lack of qualifications a matter of complete relevance. If you answer "no" you are an amateur without a professional record casting rather nasty aspersions on the work of a professional with an extensive research track record in the subject you purport to be able to judge. There's entirely too much of this sort of thing going around, it's debasing to everybody concerned. You seem upset that you're not free to make whatever remarks you please here. I suggest that you've developed some poor habits by frequenting places where debased discussion is tolerated. Your choice of the term "pure junk" effectively made you part of the subject we're discussing because naturally anybody reading that remark is going to immediately wonder, "who says that and why should I believe him?" As you can tell, your language certainly got my attention. By your language you chose yourself as a topic, Daniel. Please don't complain to me about your choice.
36. daniel at 01:04 AM on 1 July, 2010
    No I don't have a degree in advanced Geology or a shining track record in paleochronology, but people like yourself just don't seem to understand how irrelevant that is. The debate on these issues is debased by this kind of ad hominem garbage. The details of what the researchers did is all laid out in the paper. Scientists from multiple fields need only read, perhaps suppliment that reading with some supporting material and they can have a thorough and detailed understanding of what has been done. It is a fantasy for you to think that you yourself cannot research the methods of the scientists and critique their papers just because you have not studied it or worked in the field. It is patronising nonsense. You put these people on a pedestal of heavenly heights and praise them as infallible heroes who shall not be questioned..... but that is just not realistic in the slightest. I did not say "pure junk" but "utter joke" and you felt you needed to leap to the aid of researchers who may actually be embarassed by your amateur attempts to dress them up as god like figures. How do you know that they might not agree with me in retrospect? How many times do I need to explain that I was never disputing the quality of the data but referring to the validity of the final comparison of two very different data sets? You have done it again in comment #33 Lets go through that one with some quotes. You said: "I see your point about samples 7 & 8, I'm sure Donnelly would have been happier if they'd resolved better but because they're embedded in the middle of the series their effect is not very drastic; interpretation of those is constrained by the surrounding boxes." Does Donnely et. al. actually say that? At what level of confidence can we say that the true paleo sea level is in the middle of or at the extremes of the boxes? If you want to say that it is closer to the middle you will lose statistical confidence to less than 95% At 95% you can speculate that the paleo sea level may have been at one or the other extremes. This reality is part and parcel of scientific data doug you can't wish it away regurgitating the word "conservative" from the paper. Here's where again you (after having plenty of time to just read what I say and not imagine it) seem to be putting words in my mouth: "As to your problems with multiple date ranges for samples, if you read the text carefully you'll see how Donnelly eliminated date ranges by using methods beyond C14" I'm not disputing the eliminations for samples 1 - 10. But..... look carefully at the graph like I asked you to and you will see that sample 11 has two date ranges assigned. Two boxes, not greyed out lines, but boxes at the same height labelled 11. This is actually in conflict with table 1 which seems to suggest the younger range is rejected. But the researchers can't use the principle of superposition adequately here to dismiss the younger age range for sample 11 because it is still slightly younger than or equal in age to samples 9 and or 10. Either the box assigned to the younger age range is a printing error or the authors neglected to discuss this inconvenient data point in detail. How did such an error occur in the indestructable field of climate science?
37. KR at 01:53 AM on 1 July, 2010
    daniel - while the tone on this thread has become quite heated, you have made some extremely strong statements (utter joke) regarding the Donnelly paper. Looking at your initial comments, are you indeed saying that the current rise in sea levels could drop in between the samples Donnelley collected? And that therefore their data is not strong enough? Keep in mind that while there _may_ be space between samples for a steep rise, it would have to be accompanied by an equally steep decline or halting trend to match later samples. And that the samples are independently dated except for elimination of carbon date repeats by physical position ordering - an excellent technique for disambiguation, I would add. Between the multiple species examined, carbon dating, choice of uncompacted site, etc., this is an excellent paper. And hence, it shouldn't be a surprise that some people have reacted strongly to your harsh dismissal of it. As to sample 11 (representing ~8% of the data) - you may have a point there, it looks like they left the younger (eliminated) date box for #11 on the chart. But their fitting appears to use the information from Table 1, and while this looks to be an editing issue, that doesn't seem IMO to affect their calculations or their conclusions. They certainly seem to have used the #11 older date for the curve fitting. And as to how such an error might occur? While editors appear to be inhuman in nature (grrr) they are actually fallible in reality. I'd suggest dropping a note to Donnelly et al and asking whether this is the correct chart. It might be interesting to ask if these samples could be used as date tags, and examine intermediate samples (in some number) to see if there were fossil variances indicating different levels of sea rise (short term variations) - but as it stands, with the data they extracted, the linear mapping with a curve at the beginning of the industrial era is perfectly justified by the data used to generate the trend lines.
38. daniel at 02:03 AM on 1 July, 2010
    Further quotes from #33 "if I get you right and correct me if I'm misunderstanding you, you're suggesting that it's equally reasonable to pick and draw a series of lines perhaps pointing up and perhaps pointing down between any chronologically linear pair of samples." Well almost, what I'm really saying is that the paper hasn't got a hope of determining the short term sea level trends during 1300-1850 AD. This ties in with what you say next: "That's not as conservative as doing what Donnelly did." You like to regurgitate that word but I wonder if you know why Donnely et.al. used it? It's because they are trying to claim that the centres of their 95% confidence boxes are of a higher likelihood of being where the true paleo record lies. I cautiously agree with them on that, but I can't see how it legitimises their final comparison. If the centres of the boxes were the true paleo sea levels then in an attempt to obtain short term trends from the data (which is tje only valid comparison to make when trying to detect unusual recent uptrends in the last 150 years) you could in principle draw lines between each pair or in other words connect the dots. But then you would have to explain the jump back in time between samples 10 and 9 or explain why the rapid rise in sea level between samples 7 and 8 is not as, or even more alarming, then the recent rise over 150 years. No I won't let them have it both ways. If thier long term trend line doesn't need to cut right through the centre of the boxes then neither do my short term trends. You continue: "As well, doing such a series of arbitrary choices leaves the issue that the entire series must begin somewhere within the region circumscribed by the sample 4 and 11 confidence boxes, meaning that the overall conclusion of the series of choices made to connect individual samples ends up being nearly the same, confined by the beginning and ending samples." Sigh. When will you understand that the important point here is that a lack of certainty in short term trends invalidates any claim that recent rises are alarming. "Meanwhile, it appears that the slope described by the direct recent tidal measurements is inevitably going to be steeper than the sum linear product of whatever combination of ups and downs you might choose to impose on the paleo series..." Same as above "sum linear product" is irrelevant, short term uptrends of similar rate and range are. They cannot be detected by a study of this type. "...and as well covers a disproportionate vertical range compared compared to the paleo series. This suggests to me that attempting to create and insert arbitrary additional information into the series is pointless." No the paleo data covers something like 60-70cm and the recent data covers 30-40cm. Plenty of slack for a similar uptrend and a plateau. " The suite of dating refinements employed by Donnelly I refer to are an example our ignorance, as I mentioned before." You make it sound complicated but as I've outlined before, people who are educated in distant fields to climate science can easily understand a climate science paper. The principle of superposition is simply applying the logic of higher stratum are younger than lower stratum. Using known historical markers from the introduction of different plant species as added refining tools for dating is not complicated. Just becausr you don't understand it doug doesn't mean that I don't.
39. Peter Hogarth at 02:25 AM on 1 July, 2010
    daniel at 10:49 AM on 30 June, 2010 Much of this debate is focusing heavily on one Donnelly paper based on data from one area (Southern New England). Let me take a different tack. If we accept that the temporal sampling in Donnelly 2004 on historical sea level is sparse, we have two options. 1) We look for more data to fill in the gaps from other sources, and build up a higher resolution composite. Though there are difficulties with different rates of rise in different regions this process is ongoing and 6 years is a long time in climate research. From what I have read and am aware of in a professional capacity the evidence suggests relatively small changes in sea level over this period consistent with Donnelly (allowing for occasional dramatic localized crustal movements). 2) We also look at the physical effects which cause sea level rise and see if these have changed over the period in question. This might be viewed as “modeling”. Any dramatic sea level variability between or over the temporal range of uncertainty of the samples (as you suggest could hypothetically be present) would be driven by dramatic variability in temperature, land ice volume or hydrological cycle, or some combination. The evidence on past variability in temperature is far denser temporally and spatially and better researched, and is consistent with the published estimates of past changes in sea level (for example see Grinsted on Medieval Warm period). The ice core data which can give not only regional (North and South) temperature proxies but estimates of deposition/loss rates is consistent with this also. Then we must apply this same logic to what is happening now, and look at recent research in other areas of climate related science. Temperature is rising, global ice mass is diminishing and sea levels are rising with both thermal and ice melt contributions. For some more recent overviews and a few more clues on extra data see Church 2008, and Milne 2009 as well as Grinsted linked previously.
40. Doug Bostrom at 03:01 AM on 1 July, 2010
    I don't find your remarks persuasive, Daniel. Your entire thesis depends on deriving -more- interpretive detail from a data series you yourself claim has insufficient power to describe -less- detail. That's nonsensical. By the way, Donnelly does not own the term "conservative." It's commonly used in science the same way we might use "circumspection" or other words suggested suitable humility in the face of ignorance.
41. daniel at 17:15 PM on 1 July, 2010
    Yes pete I know that we are focused too much on the one paper here. But I am not going to take the blame for that one (nor am I suggesting that you're blaming me). When I used the infamously inflammatory "utter joke" comment (and I apologise for the severity) I made it clear that I was saying as much considering the paper "on it's own" see for yourself #19. Others decided to take offence and make it part of the focus of the discussion. I just wanted to examine the papers cited by the article and show the poor quality of conclusions that can come from what is supposed to be "peer reviewed" climate science and how this kind of data is then used to support AGW to the public. I feel that there is far too much public trust in the quality of work, not just in climate science, but in the entire body of technical literature out there across all fields. The kind of ad hominem rhetoric above is being used as a debate strategy by proponents of AGW and it is simply not science. Can we please just discuss the data and the quality of conclusions drawn. It is extremely unconvincing (or should be) to anyone who works in a technical field to simply quote your qualifications and report a list of published articles. All this proves is that you are active in the field and fairly knowledgeable. But your papers can still be scrutinised by others who have been educated in distant fields but still have an understanding of (and can easily read up on) how the basic underlying principles (Math, Stat, Phys, Chem, sampling techniques etc.) are used to perform the work in your field and whether you have gleaned logically valid conclusions from your study. To adress your points from #39 in order 1) I believe that Gehrels et. al. 2006 cited in the original article tried to address the low resolution issue and even mention it in the paper. But I will go on to argue that this paper is another insight into the nature of the methods used in these studies which appear to be creating large height uncertainties, coupled with time uncertainties, that undermine any detection of short term recent trends (even with high resolution data). I'm not saying that any significant errors in the methodologies have occurred, just that the methods employed have too much uncertainty to detected the trends described. Please cite the papers and I'll try and get a hold of them. If these studies suffer from the same problems I see in Gehrels 2006 then I will most likely not be convinced of recent rapid SLR. 2) Yes this is a good point there needs to be a driver of SLR. But given that my confidence in the quality climate science is currently very weak due to issues like those in my complaints above then I doubt that my appraisal of the Grinsted and other SLR (or other) articles (that also investigate other factors like paleo-climate etc.) will be similar to yours. I fear I will find the same skewed conclusions I have thus far read in the two papers cited. I will need time to read the Grinsted paper and the reviews/articles you have linked to. Thankyou for those please let me know of any others you think are relevant. Doug, you may almost get my point on the Donnely paper. You said in #40: "Your entire thesis depends on deriving -more- interpretive detail from a data series you yourself claim has insufficient power to describe -less- detail." Donnely's thesis is actually guilty of claiming -more- detail through 1300-1850 AD than is actually detected. I am saying that it is largely unclear as to what short term trends may or may not have existed in that time. Therefore the final conclusion by Donnely et. al. based on the data they have provided (not counting sea level data from any other papers) was invalid and may be just the tip of the iceberg when it comes to the true state of the quality of climate literature. If this, seemingly popular, website uses these papers to back it's message of AGW then I fear what else may be going on
42. Doug Bostrom at 17:47 PM on 1 July, 2010
    Daniel, Donnelly interprets his data as offering -no- detail between 1300 to 1850, instead chooses as much as possible to smooth his interpolation by taking a linear approach, you propose that there may be detail therein. Who is more liberal in interpretation? We disagree and I don't think either of us is going to change our minds.
43. Peter Hogarth at 05:19 AM on 2 July, 2010
    daniel at 17:15 PM on 1 July, 2010 I accept the data points on the chart are sparse, and have temporal uncertainty and height uncertainty. The time series trend developed from these samples is a simple low order curve fitted through them, which is common practice when trying to extract trends. This assumes low long term variation of the variable in question. An alternative fit involving high short term variation which still explains the limited data points would involve undersampling and aliasing. Is this likely? You do not suggest it is, but you propose it is possible. I have spent some hard earned cash looking at over a dozen recent site specific salt marsh studies related to sea level. Without exception they display the poor resolution which you would rightly criticize if any study was the single source of our evidence. One of the best that is free (I honestly did not select by author!) is Donnelly 2006. It is possibly close enough geographically to the 2004 data set so that gaps in the record in each could be reduced, which makes short term variations far less probable. Likewise, archaeological and historical evidence on sea level changes, from around the world, taken in isolation, means little - and could easily be written off using local crustal depression etc. It is the integrated wider evidence based picture that emerges when researchers try to put all this stuff together that proves persuasive to many, and to me. A further point is that researchers like Donnelly cannot help but acquire a great deal of background knowledge or expertise, and will be aware of large amounts of evidence that might not even be in publication, but nevertheless adds to the overall common sense probability based conclusion. Is this conclusion overwhelmingly robust? I couldn’t say without analysis, but it is consistent with the majority of recent and emerging published data. daniel at 17:15 PM on 1 July, 2010 I accept the data points on the chart are sparse, and have temporal uncertainty and height uncertainty. The time series trend developed from these samples is a simple low order curve fitted through them, which is common practice when trying to extract trends. This assumes low long term variation of the variable in question. An alternative fit involving high short term variation which still explains the limited data points would involve undersampling and aliasing. Is this likely? You do not suggest it is, but you propose it is possible. I have spent some hard earned cash looking at over a dozen recent site specific salt marsh studies related to sea level. Without exception they display the poor resolution which you would rightly criticize if any study was the single source of our evidence. One of the best that is free (I honestly did not select by author!) is Donnelly 2006. It is possibly close enough geographically to the 2004 data set so that gaps in the record in each could be reduced, which makes short term variations far less probable. Likewise, archaeological and historical evidence on sea level changes, from around the world, taken in isolation, means little - and could easily be written off using local crustal depression etc. It is the integrated wider evidence based picture that emerges when researchers try to put all this stuff together that proves persuasive to many, and to me. A further point is that researchers like Donnelly cannot help but acquire a great deal of background knowledge or expertise, and will be aware of large amounts of evidence that might not even be in publication, but nevertheless adds to the overall common sense probability based conclusion. Is this conclusion overwhelmingly robust? I couldn’t say without analysis, but it is consistent with the majority of recent and emerging published data.
44. daniel at 18:21 PM on 2 July, 2010
    Peter Hogarth at 05:19 AM on 2 July, 2010 "An alternative fit involving high short term variation which still explains the limited data points would involve undersampling and aliasing. Is this likely? You do not suggest it is, but you propose it is possible. It is not possible to know if it is likely given the data presented. But it is certainly possible and given the data presented, it's just as likely. "Without exception they display the poor resolution which you would rightly criticize if any study was the single source of our evidence." The Gehrels paper (annoyingly $25 USD) cited in the article has decent resolution through most of the period examined. (Coincidentally it lacks resolution through most of the period examined by the Donnely paper discussed). Another article which I still haven't read yet by Gehrels et. al. Quaternary Science Reviews 24 (2005) 2083–2100. Appears to have high resolution over the period discussed. But I think that as long as error estimates are high, comparisons to short term recent trends will be undermined. "One of the best that is free (I honestly did not select by author!) is Donnelly 2006. It is possibly close enough geographically to the 2004 data set so that gaps in the record in each could be reduced," Admittedly I haven't read the article yet but having skimmed over it and focusing on figure 7 and taking into account the height uncertainties, I can't see how either paper helps to validate the other. Donnely 2004 only covers a small portion of the Donnely 2006 reconstruction between samples R5 and R6 and has ~20cm total height error at 95% confidence (you can read the actual +/- ~10cm intervals in the text). Donnely 2006 has ~30cm total height errors and so cannot serve to further refine Donnely 2004. Donnely 2004 cannot serve to refine 2006 beyond possibly reducing the error margins to ~20cm total. "Likewise, archaeological and historical evidence on sea level changes, from around the world, taken in isolation, means little - and could easily be written off using local crustal depression etc." Certainly the overall global picture is not represented even if all of the papers we've discussed thus far are put together, I agree. Is crustal depression such a factor in any of these papers? I thought that the more recent papers tried to address these issues, Gehrels 2006 seems to. But I may have my (depression mixed up with my subsidence, my ignorance shows - but I could read up on that Doug) "It is the integrated wider evidence based picture that emerges when researchers try to put all this stuff together that proves persuasive to many, and to me." But the wider picture in terms of SLR thus far seems to be an integration of poor data (in the correctly calculated error estimate sense) and or conclusions. "A further point is that researchers like Donnelly cannot help but acquire a great deal of background knowledge or expertise, and will be aware of large amounts of evidence that might not even be in publication, but nevertheless adds to the overall common sense probability based conclusion. Is this conclusion overwhelmingly robust? I couldn’t say without analysis, but it is consistent with the majority of recent and emerging published data." I hope you don't get annoyed when I say that i think this comment again boils down to an ad hominem type argument roughly equivalent to "They're experts, just trust em, they know what they're doing." I wonder about the quality of all of that published data both on SLR and other climate factors.
45. KR at 00:30 AM on 3 July, 2010
    daniel - While it's possible that there are high frequency changes in temperature missed by a particular low-resolution sample set, it's really completely unreasonable to postulate that this indeed is the case based on that evidence. If I permit more degrees of freedom in my fitting than are supported by my data, I can draw whatever curve I like - including one that indicates the Earth cycled between absolute zero and plasma temperatures during a 30-day period between samples. I could also postulate that such temperature swings were driven by invisible pink unicorns, but I don't have samples that actually indicate that. In the universe of possible data fits, a randomly chosen fit is NOT as likely as the simplest one that fits the data. It's a rudimentary basis of data analysis that you don't over-fit your samples - that falls into the aspects of parsimony, or Occams razor. Given the samples present in the papers you have been referring to, it's reasonable to state that there's a linear historic trend passing through those data points, with a later steeper trend passing through the much denser data points of recent records. Are there excursions outside that linear trend that don't fall upon the sample points, that weren't sampled? Perhaps. That would take more data - the data presented doesn't support that hypothesis. If you take into account the multiple lines of evidence, the many data sets containing samples at different (and overlapping) timepoints along historic record, the hypothesis of a fairly linear trend for the 1400-1850 period, with a steepening incline after that, is still the most reasonable, parsimonious explanation that fits the data. And with no unicorns...
46. daniel at 19:44 PM on 3 July, 2010
    To KR #45 "daniel - While it's possible that there are high frequency changes in temperature missed by a particular low-resolution sample set, it's really completely unreasonable to postulate that this indeed is the case based on that evidence." Yep, don't remeber talking much about temperature changes. Mostly about SLR. "If I permit more degrees of freedom in my fitting than are supported by my data, I can draw whatever curve I like - including one that indicates the Earth cycled between absolute zero and plasma temperatures during a 30-day period between samples." Geez KR don't go overstating what I said or nothin. Although maybe I guess you're right, slight short term increases in SLR (not temperature) are about on par with absolute zero to plasma level temperatures (not SLR) aren't they. I'm so glad I've got you around to keep my feet on the ground. Thanks KR ;) "I could also postulate that such temperature swings were driven by invisible pink unicorns, but I don't have samples that actually indicate that. In the universe of possible data fits, a randomly chosen fit is NOT as likely as the simplest one that fits the data. " Uh huh..... unicorns...... got it. I think you may be a little stressed having to strain to understand that I don't dispute the long term linear trend just it's comparison to the short term uptrend. It's called an invalid comparison :) "It's a rudimentary basis of data analysis that you don't over-fit your samples - that falls into the aspects of parsimony, or Occams razor. Given the samples present in the papers you have been referring to," Yes I know you shouldn't "overfit" your samples as you say. You also shpuldm't just assume short term linear trends from such noidy data. It's cute that you know what parsimony is I wonder if you're aware that it doesn't always apply to reality (a concept you claim to have a better grasp of than me). It's also cute when people go on about Occam's razor, a phrase recently popularised by the movie "Contact" but as I j ust said, not always applicable. " it's reasonable to state that there's a linear historic trend passing through those data points, with a later steeper trend passing through the much denser data points of recent records." But not to say that the recent uptrend has been shown to be unusually high given the sparse, noisy data available. "Are there excursions outside that linear trend that don't fall upon the sample points," Eh? Please clarify this rant. "If you take into account the multiple lines of evidence, the many data sets containing samples at different (and overlapping) timepoints along historic record," Yes I know they're claiming the trend extends into the instrumental record. Lucky really. My complaints are perfectly valid and need to be addressed. "the hypothesis of a fairly linear trend for the 1400-1850 period, with a steepening incline after that, is still the most reasonable, parsimonious explanation that fits the data. And with no unicorns..." I agree, there was definitely a long term trend of 1mm/year that overlaps with a few decades of the instrumental record but what were the short term trends in that period. What do you have against unicorns in science! They are just as able to understand your jibberish as I am and any reference to them as "imaginary", "mythological" or "unparsimonious" I take as an ad hominem attack!
47. KR at 02:08 AM on 4 July, 2010
    Actually, daniel, I think that would be insulting unicorns would be ad Unicornis, as best I can tell. I'll apologize now to the greater Unicornis community... Sorry about the mis-reference to temperature, instead of sea level - proof that I sometimes don't proof-read enough! And that perhaps I'm taking too much cold medicine at the moment :) "Are there excursions outside that linear trend that don't fall upon the sample points" means that the sample points in that Donnelly paper mark, within the errors on vegetation prevalence and radiocarbon dating, points on the sea level record. The simplest fit justified by the record is a piece-wise linear fit running through each data point. The least justified fit is a line that avoids your data points. Given the noise in that simple reconstruction, it's reasonable to time-average data points, especially for the recent (dense, somewhat noisy) data points. Note that the core samples have some implicit time averaging - it takes time for vegetation to grow, and the sample investigated is not going to be a 2D core slice; the thickness of it (and is the sediment flat there?) will introduce some time averaging. I didn't see that explicitly stated in the paper, but that's a known element for core analysis - you don't tend to see day-to-day changes in them! Either way - the reconstruction best justified from the evidence in this experiment should pass through or very close to each of the data points or averages. The data "anchors" the reconstruction there, and any large deviation from the trend (excursion) would have to either (a) show as a shifted data point, or (b) occur between data points - and vanish again before the next one. However, there is in this experiment actual evidence against offsets from the reconstructed sea level trend around the data points themselves.
48. Peter Hogarth at 03:42 AM on 4 July, 2010
    daniel at 18:21 PM on 2 July, 2010 I think you may have missed or misunderstood the point. If there were short term variations of the magnitude which you suggest between the sparse points then the probability of all of these randomly sampled points fitting any smooth long term curve is small. Statistically, your alternative is most certainly not "just as likely"! Any extra points we find which also fit the curve increases the probability that the curve is a good model, and constrains other probable models to those with low amplitude variations. With respect, if this is lost on you, then I understand why you keep re-iterating your point, and you should address this. I would not argue that you should not question the work of experts. I am arguing that Donnelly is presenting work that is specialist. His data is site specific and is intended to add a small piece to the unfolding picture which is science, rather than act as first line defence against "climate skepticism". That you accept that drivers of sea level should be accounted for is a good step, yet you still do not appear to modify your suggestion of "likely" high sea level variations in light of this. This is not scientific.
49. Doug Bostrom at 06:42 AM on 4 July, 2010
    Two Daniels?
50. daniel at 18:49 PM on 5 July, 2010
    doug_bostrom at 06:42 AM on 4 July, 2010 There is only one Daniel KR at 02:08 AM on 4 July, 2010 "Sorry about the mis-reference to temperature, instead of sea level - proof that I sometimes don't proof-read enough!" Yes, but it's not just you that's doing it both Doug and Peter have also skimmed over what I've said and quickly responded with fervour without actually understanding my point. I would like to highlight the fact that these issues are highly emotive and the fears people have from your side of the argument are clouding your judgment. This is occurring both on this and other points of the debate and is clearly evidenced by all of your comments during this discussion. "The least justified fit is a line that avoids your data points. " Have I proposed such a fit? I have proposed short term fits within the error estimates of the data points. "Given the noise in that simple reconstruction, it's reasonable to time-average data points, especially for the recent (dense, somewhat noisy) data points.” I agree but I don’t see the relevance, you cant compare that recent, directly measured, high resolution, short term to the uncertain, low resolution, long term data set like that. “Note that the core samples have some implicit time averaging - it takes time for vegetation to grow, and the sample investigated is not going to be a 2D core slice; the thickness of it (and is the sediment flat there?) will introduce some time averaging. I didn't see that explicitly stated in the paper, but that's a known element for core analysis - you don't tend to see day-to-day changes in them!" I can't say I follow you here. My understanding is that the time uncertainties are from the C14 analysis. The researchers can only obtain a date range (from a non-Gaussian probability function) using this method. It doesn't give you a range on the order of days or months but years. "Either way - the reconstruction best justified from the evidence in this experiment should pass through or very close to each of the data points or averages. The data "anchors" the reconstruction there, and any large deviation from the trend (excursion) would have to either (a) show as a shifted data point, or (b) occur between data points - and vanish again before the next one. " I'm sorry but you are not addressing the long term / short term issue. I will say again that I agree with the proposed long term linear trend and the data allows for short term deviations not too far from the data points that would undermine Donnelly’s conclusions. "However, there is in this experiment actual evidence against offsets from the reconstructed sea level trend around the data points themselves." Explain. Peter Hogarth at 03:42 AM on 4 July, 2010 "If there were short term variations of the magnitude which you suggest between the sparse points then the probability of all of these randomly sampled points fitting any smooth long term curve is small." Maybe you are finally understanding my point. You're right that on the short term scale the probability of the long term trend is small, thankyou. :) "Any extra points we find which also fit the curve increases the probability that the curve is a good model, and constrains other probable models to those with low amplitude variations." There aren't any more data points provided and if more data points showed that there was a low amplitude variation from the linear trend then the recent uptrend would look less alarming, more precedented or natural and much less anthropogenic or induced by CO2. "With respect, if this is lost on you, then I understand why you keep re-iterating your point, and you should address this." It's not lost on me Peter, as far as I can see you are trying to use wordy rebuttals that don't amount to much. There is not enough resolution to determine that thre is a long term linear trend that barely deviates on the short term. More importantly as long as there are large enough uncertainty levels the recent uptrend will never be shown to be unusual. If you reconstruct the data Peter from table 1., just use the absolute centres of the boxes, you will see that using sample 1 (dated ~1975) along with the other data points the trend stays much the same (possibly even lowers a little) and so the entire trend over the last 700 years is still ~1mm/yr at Barn Island. I hope that addresses your “undersampling” or “Unlikely wild deviations” tack. The instrumental record is showing us that the Donnely reconstruction may in fact be an undersampling of a natural higher amplitude trend. "I am arguing that Donnelly is presenting work that is specialist." Ad hominem "His data is site specific and is intended to add a small piece to the unfolding picture which is science, rather than act as first line defence against "climate skepticism" Undoing the poor science of climatology is the first line of attack when it comes to this debate. Their methods may be scientific in nature but their conclusions seem to be biased, driven by an unfounded fear of gloom and doom. "That you accept that drivers of sea level should be accounted for is a good step, yet you still do not appear to modify your suggestion of "likely" high sea level variations in light of this. This is not scientific" It has not been shown from this data that the uptrend is un-natural and therefore it is not necessarily anthropogenic. To claim otherwise is unscientific.

Prev  1  2  3  Next

Post a Comment

Political, off-topic or ad hominem comments will be deleted. Comments Policy...

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.