Recent Comments
Prev 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 Next
Comments 111401 to 111450:
-
Tom Dayton at 03:07 AM on 1 September 2010Hockey stick is broken
Deep Climate has done a detailed analysis of the McShane and Wyner paper. -
HumanityRules at 02:59 AM on 1 September 2010Sea level rise: the broader picture
38.Ned If +5.4 mm/year or +4.9 mm/year compare well with +3.3 mm/year then how different is 1.8mm/year (overall 20thC) from 3.3mm/year (past twenty years)? We seem to be arguing this second difference is significant but it doesn't look so different from a couple of numbers you say "compare very well". -
TOP at 02:58 AM on 1 September 2010Temp record is unreliable
The answer might be no. A STATISTICAL ANALYSIS OF MULTIPLE TEMPERATURE PROXIES: ARE RECONSTRUCTIONS OF SURFACE TEMPERATURES OVER THE LAST 1000 YEARS RELIABLE? McShane and Wyner. Submitted to the Annals of Applied Statistics One of the conclusions:
...we conclude unequivocally that the evidence for a ”long-handled” hockey stick (where the shaft of the hockey stick extends to the year 1000 AD) is lacking in the data.
In other words, there might have been other sharp run-ups in temperature, but the proxies can't show them. The hockey stick handle may be crooked, but the proxies can't show it one way or the other.Moderator Response: Not the same topic. Try this thread for a better place to discuss McShane and Wyner:
Is the hockey stick broken? -
Alexandre at 02:45 AM on 1 September 2010Urban Heat Islands: serious problem or holiday destination for skeptics?
but... but... I was sure if you dropped urban stations there would be an obvious decline in global temperatures! (just kidding) -
omnologos at 02:44 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Just to clarify: we are talking about USHCN criteria, not mines. Seemingly I have started off something 'round here. Since I find it less than kind to hijack somebody else's blog with a salvo of comments, apologies to all but I'll suspend things here, and perhaps post a note or two later today, plus something in my own blog. I strongly suspect we have all made our points, sometimes more than once, and could continue for weeks. That's part of the good/bad nature of the 'net. Thank you all, and to John Cook and the other authors, for the intellectual challenge. -
HumanityRules at 02:23 AM on 1 September 2010Sea level rise: the broader picture
22.Dappledwater It's not just the Wenzel paper that shows no acceleration. It looks like the 18 satellite data doesn't show one. So you're saying the acceleration occured around 1950 and since then things have been moving at a steady rate? That's what I see from the figures you show. But the most prominant temperature rises and land ice melt have occured post 1970 and post 2000. It doesn't seem to fit. I don't think land/sea temp is as relevant as OHC. Try this old, white, male deniers presentation of the NODC data on OHC. Shouldn't we be seeing acceleration in teh 18 year satellite data?Moderator Response: The topic of ocean heat content is better explored on an appropriate thread. A robust discussion may be found here:
Robust warming of the global upper ocean -
Jim Meador at 02:23 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Hi again omnilogos "My example is about going to see 100 doctors, _all_ making mistakes in measuring your blood pressure." Right, your example is about absolute accuracy. In this case, the right example would be going to see these 100 doctors every day for 30 years, and the doctors all make consistent errors in measuring your blood pressure. At the end maybe you would not know your absolute blood pressure, but you would certainly be able to tell how it had gone up or down over the years. You would know a lot about the trend. You CAN get useful trend information from imperfect data, and that is the point of the Menne paper. -
Tom Dayton at 02:23 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
omnologos, all your doctors' blood pressure readings of you are "wrong" by your criteria! You will never be able to find a doctor meeting your criteria, because there is no such doctor! If you research the definition of "blood pressure," you will discover that the concept is useless without an operationalization. Then you will be dismayed to discover that there are in fact a bunch of operational definitions, involving the patient sitting, lying, standing; the instrument being on the wrist or arm or other body part or even inserted in a blood vessel; the instrument being a microphone or a health care provider's ears in a stethoscope or a direct pressure sensor; and even the strength of the sound that marks the trigger for the measurement has multiple values. Have you ever noticed that none of your doctors has ever been consistent in demanding your posture be precisely "correct" for precisely one of the many requisite durations before measuring your blood pressure, let alone insisting that you be at the doctor's office at precisely the same time of day? -
Klaus Flemløse at 02:14 AM on 1 September 2010Urban Heat Islands: serious problem or holiday destination for skeptics?
As I recall, IPPC states: In the past warming from UHI is insignifikant. In the future it may be signifikant. Please consider to included this in writing about UHI. -
Berényi Péter at 02:14 AM on 1 September 2010Urban Heat Islands: serious problem or holiday destination for skeptics?
#1 RSVP at 00:10 AM on 1 September, 2010 "...results in a slightly warmer envelope of air over urbanised areas when compared to surrounding rural areas.." Why not express this in degrees centrigrade? It is about 0.2-0.3°C per doubling of local population density for a very wide range of initial population densities. This rule also works for sites considered "rural", although the exact value of the coefficient should depend on level of economic development as well. This is the reason behind similar trends in urban vs. rural sites. That is, even if scientists have been very careful to ensure that UHI is not influencing the temperature trends, they could only make sure the influence was about the same over all kinds of sites. Population density distribution is always fractal-like and population on average grows by the same percentage everywhere. The net result is "urbanization", when ever higher proportion of the population lives in really densely populated areas. But as UHI is proportional to the logarithm of local population density, it is no wonder its effect on trend is not smaller for low population density areas. In fact it is expected to be a bit stronger there, because much smaller absolute numbers are needed to increase population density twofold. The only correct check for the actual magnitude of the temporal UHI effect is to calculate temperature trends for sites where local population density has decreased for an extended period and compare them to the rest. Of course it is not easy to find such regions, because global population has doubled twice since the beginning of the last century (therefore about 0.4-0.6°C of the global trend is due to UHI). However, the quest is not impossible. For example several regions of the US experienced multi-decadal population decrease (southern West Virginia, Northern Maine, many regions of the mid-west). There are also excellent census data in the US, so it is pretty easy to locate such regions. Basically one should choose all the counties where population has decreased for the last thirty years, with a decreasing population in all the neighboring counties as well and which have climatic data for the entire period (something like Beckley city, Raleigh County, West Virginia). That's the job to be done for scientists, provided of course there are some who really want to be careful. -
MichaelM at 02:12 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
It is ironic that omnologos cites Galileo who I would posit never made 'accurate' measurements. He had no timepiece, his compass was probably only accurate to a degree, his lenses were imperfectly ground etc, but his contribution to science is immeasureable. -
JMurphy at 02:10 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
omnologos wrote : "If a doctor doesn't use the diagnostic instruments properly, I go to a doctor that does, don't you?" How do you know when you've found a doctor that uses that instrument properly ? What do you judge the result against ? For me, if a doctor didn't use the instrument properly (but I didn't know - I was suspicious because I got a result I didn't like) and gave me bad news, I would go to another. If the next one told me the same bad news, do I suspect that he/she doesn't know what he/she is doing too, or do I face up to reality ? -
Jim Meador at 02:08 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Hi omnologos, You aren't making a distinction between absolute measurements and trends, and this is something many people overlook. Although on any given day a perfect measurement is great to have, it is not required for trend analysis. If every measurement is off by 100 degrees, every day, over the entire record, then this 100 degree offset does not matter for the trend analysis. It was present at the beginning and at the end, so it does not affect the trend. The graph shows the "Anomaly" which is the difference from a reference point. (Notice that the scale is near zero.) Absolute measurement errors that are present throughout the record don't make it into the anomaly. You mention "poor data" but the point of the Menne study was to show that "poorly sited" stations still provide useful data for trend analysis. "Poorly sited" does not mean "poor data" when it comes to trends. -
Doug Bostrom at 02:03 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
I don't think you'll be able to find a doctor you're comfortable with, omnologos. How many have you visited? Have you compared readings? How can you tell if the readings are correct? -
CBDunkerson at 02:03 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
So... when confronted with three matching data sets; A, B, and A+B which all show virtually identical results your conclusion is that they must all be wrong? Because some guy with a blog found that set B don't meet a list of criteria which were established for measuring something other (absolute temperatures) than the data sets are looking at (relative temperature changes). When we then add in matching anomaly readings from weather balloons does that mean those are ALSO wrong? The matching satellite temperature record? Thrown off by urban heat islands in space? The various matching proxy temperature sets? Has EVERY temperature reading in the history of the world been wrong? Or are you being ridiculous? -
The surprising result when you compare bad weather stations to good stations
Omnologos - I stand by my statements regarding accuracy and precision. If your doctors instruments are poorly calibrated they will be inaccurate (offsets from ground truth). If, however, they are precise, they will give the same readings over time. The various issues raised about temperature collection seem to all be about accuracy - arguing that there are offsets for individual instruments due to siting. But these issues have nothing to do with precision, which is driven by consistent use of an instrument, maintaining correct offsets when changing thermometers, changing the time of day for recording values, or moving an instrument - those are all procedural issues, and the only complaints I've seen on those is "It's complicated!" from a couple of posters. Trend analysis only requires precision. -
michael sweet at 02:01 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Omnologos, Scientists HAVE analized the data very carefully over decades and they say that there is a trend in temperature. If you want to claim that "it's because there is something else affecting the measurements, and maybe there is no trend" then you need to suggest what that something might be. Otherwise you are just saying "I doubt it" and you have no argument. What do you suggest is causing this trend, in the good and the bad stations, that is not AGW? It needs to explain glacier retreat, seasons changing, sea level rise also. I look forward to your explaination. -
omnologos at 01:55 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
@KR: "This actually gets at the difference between accuracy and precision". No. My example is not about accuracy and precision. My example is about going to see 100 doctors, _all_ making mistakes in measuring your blood pressure. If a doctor doesn't use the diagnostic instruments properly, I go to a doctor that does, don't you? -
omnologos at 01:52 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
@michael sweet: "You will see that the 'good' data give the same result as the 'bad' data". And that _is_ the problem. The answer could either be (a) "it's because 'good' and 'bad' don't have a meaning in matters of trends" (the most popular take, around here). Or (b) "it's because there is something else affecting the measurements, and maybe there is no trend". It's up to science to figure out which is which. If instead we assume it's (a), then it's an assumption, not a finding. And it's not science. If well-constructed clocks drift just as much as poorly-constructed clocks, we can all assume time is dilating, or try to understand what is actually happening. -
The surprising result when you compare bad weather stations to good stations
Omnologos - Given your doctor example here, if you've visited 100 doctors and received varying results for blood pressure, that means that the accuracy of the various instruments vary a lot. Perhaps some of the are better at calibrating their instruments than others? If you then visit the same 100 doctors a year later, using the same instruments, and see that on all of them your blood pressure is 20 points higher than it was the year before, a trend change - are you going to throw out that trend change because of the accuracy issue? Or are you going to get a prescription for a blood pressure lowering medicine? This actually gets at the difference between accuracy and precision. The adjustments guidelines are intended to maintain both accuracy (calibrate correct offsets for site conditions) and precision (when changing thermometers, adjust for instrument differences). Trend analysis only requires precision and sufficient data, not accuracy for each instrument. -
dsleaton at 01:40 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Forecasters: "The barometer on station X seems to be off by about 100mb, but it is still trending downward, in line with the other three stations. Observers also report increasing wind and wave activity. Our models indicate a 90% chance that a major hurricane will make landfall here tonight." Omnologos: "Right, whatever! One out of four stations is totally off! That's significant error! 90%? Not good enough! I believe I'll be partying on the beach tonight. Bloody alarmists." -
michael sweet at 01:36 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Omnologos: Check out the graph at the top of this post. You will see that the "good" data give the same result as the "bad" data. Deniers like Watts would have us throw out the entire data collection, "bad" and "good", because of the issues you are raising. Scientists have corrected the "bad" data as much as possible. The graph here shows that it is not necessary to correct the "bad" data-- they are the same uncorrected! Looking at the graph in the post I have to ask "what is Omnologos concerned about?" What in this graph makes you doubt the reliability of the temperature record? No data is perfect. We do the best we can with the data we have. -
Doug Bostrom at 01:33 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Nice attempt at circularity there, omnologos. Let's express it a different way: You're wrong, so why does Jim have to post this blog at all? The answer is that folks like you will continue bashing away with your failed argument about how absolute accuracy in temperature measurements is necessary to identify a trend in temperature and its proportional relevance. This is basically the same pointless distraction as asking "if we can't predict weather, how can we predict climate?" For reasons we cannot know, maintaining confusion here is terribly important to you, enough so that you'll ignore everything you're told regarding the matter. Using your example, if you visit 100 doctors with sphygmomanometers and all of those show similarly high readings, what's the main concern? Will you worry over the absolute accuracy of all of those instruments, or should you instead be motivated to explore whether you may have a pathology causing your blood pressure to rise? You'll choose risking a stroke, for some reason. Personally, I find it ironic to discover how people worried about temperatures trending upwards show no interest whatsoever in getting those trends right (in a scientific sense). Really? How extraordinarily obdurate. Here's a summary of things of which you're probably already aware, but you can still say people are not interested in getting it right? -
omnologos at 01:19 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
@michael sweet: "what should we do with data that is not perfect". I am sure it is not news to you that no data is perfect. Some data, however, are (much) better than others. If "good" data show a trend, then "good data show a trend", and "poor" stations should be transformed into "good ones" at once, to contribute to the analysis of the trend. If somebody wants to use "poor data", they should therefore show how the "poor data" can be transformed into "good data". Unless that is done, there is as little meaning in using "poor data" to understand warming trends as in issuing speeding tickets with the use of an incorrectly-placed traffic enforcement camera. No court of law would allow those tickets, no scientist should concern about the "poorly sited" data. -
CBDunkerson at 01:14 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
omnilogos #19, you just really can't understand can you? "I find it ironic to discover how people worried about temperatures trending upwards show no interest whatsoever in getting those trends right" No. False statement. Everyone here is absolutely interested in getting the temperature trends right. You just can't seem to understand that your objection about thermometer citing is irrelevant to that issue. Here, try a simple thought experiment. Let's say in your yard you put a thermometer out on the black asphalt driveway under direct sunlight and another one in the most heavily shaded area you can find. Two thermometers less than a hundred feet apart, but they are going to show significantly different ABSOLUTE temperatures. Which is 'right'? Both of them are correct for the location they are in. Neither would meet the 'station siting guidelines'. Yet both are perfectly valid for measuring TRENDS. As the day gets hotter BOTH thermometers will register higher temperatures. The absolute temperature readings will differ, the trends over time will not. -
omnologos at 01:06 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
@doug_bostrom: if you were right, then there would have been no point in Jim Meador posting this blog at all. Personally, I find it ironic to discover how people worried about temperatures trending upwards show no interest whatsoever in getting those trends right (in a scientific sense). Imaging going to see the doctor if your concern is blood pressure, only to accept the diagnosis even if the sphygmomanometer has been obviously used incorrectly. Worse: imagine going to see 100 doctors, all making mistakes in measuring your blood pressure, only to conclude there must be something wrong because they all said so. Sanity means discarding the opinion of poorly-trained doctors. All of them. Actually, the entire "body" of modern science is based on maths, therefore on numbers, therefore on measurements. Hence, incorrect measurements can only lead to incorrect science. A "poorly sited" station is a serious issue, and the untold number of hours spent by Galileo in getting his measurements right were not a waste of time. Once again: maybe the USHCN guidelines don't matter, regarding warming trends. I am not excluding that possibility. All I am stating is that somebody should show that _that_ is the case. If it is instead considered as an assumption _because of_ the trends, we are in the realm of circular reasoning. -
michael sweet at 00:47 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Another issue relating to Omnilogos questions is what should we do with data that is not perfect? Obtaining 100 years of temperature data is an immense amount of time and effort. Deniers would have us discard all this data because it might have some small flaws. Scientists generally try to correct the record and/or make the best use that they can of the existing data. It makes no sense to throw data away when it can be corrected. As this example shows, the "bad" data replicates the "good" data. -
Doug Bostrom at 00:37 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
Absolutely pointless to say it, but a thermometer does not have to produce an absolute, accurate or precise reading in degrees anything to identify a trend. We have a vast network of thermometers we can view for this purpose without referring to units at all. This network indicates an upward trend in temperature. This is not very complicated, really. The same network used to produce absolute temperature indications can be exploited for a different purpose, to identify a trend. The trend application benefits in some ways from being expressed in units but actually only needs to be assessed from the perspective of proportionality to derive a useful conclusion. Perhaps the solution is to avoid referring to these instruments as thermometers in this context so as to avoid igniting neurotic obsessions. How about "bulk caloric trend indicator?" But, as I say, pointless to mention it down here. Fortunately normal people who are simply looking for a simple explanation of what's going on the world will not make their way down here into the deep dark, where omnologos is prattling away about guidelines. -
CBDunkerson at 00:28 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
omnologos #15, they have this little handbook which says, 'for best results place your thermometer in an area like this'. Watts decided those were 'guidelines' or 'criteria' and any site which did not meet all of them was 'bad'. All of which is completely irrelevant as those instructions were meant to limit any sort of error in ABSOLUTE temperature measurements... whereas the global anomaly values are based on RELATIVE change in measurements over time... which aren't impacted by most of the siting issues Watts (and now apparently you) obsessed about... as subsequent studies have proved repeatedly. -
omnologos at 00:24 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
@Tom Dayton: I was quoting @Dappledwater. In fact, I used quotes 8-) and prefer "poor" to "bad" anyway. Let's go back to the root of the problem here: what are those guidelines about if we are to take on "poorly sited" measurement systems in a cavalier manner? -
CBDunkerson at 00:22 AM on 1 September 2010Climate's changed before
bobconsole, actually many of the 'think tanks' (i.e. propaganda units) disputing global warming also insisted that cigarettes do NOT cause lung cancer. Go figure. -
Doug Bostrom at 00:18 AM on 1 September 2010Sea level rise: the broader picture
Is the article wrong, or did you unconsciously inflate it because you thought that it had to fit the ballpark figure you had in your mind. Is this a case where failing to focus on tiny portions of data might lead to false conclusions? "Focus on tiny portions of data." Too funny in light of more recent posts. I'm flabbergasted that folks are able to view a graph such as the Church display in the "basic" article above and end up rattling down a funnel of perspective to become entirely obsessed with a single tide gauge or a particular region of the world. You and Miekol could inoculate yourselves from appearing so stubbornly myopic by actually reading things such as the paper I cited for Peter, above: Closing the sea level rise budget with altimetry, Argo, and GRACE Instead, chanting over a web page from the Queensland maritime safety agency works better for maintaining ignorance. You really, really do -not- want to see the large picture, it seems. Sorry, JohnD, I've got a major bout of sciatica going on, I'm on Torquemada's rack right now. I goofed my decimal points. Gasp! I'ts a plot! Doug Bostrom is controlling a global network of oceanographers but has been exposed when he flubbed a coverup! Congratulations, and thanks especially for helping to illustrate a case example of what I was referring to above, the synthesis of misleading and pointless distractions. Beyond helping. -
CBDunkerson at 00:15 AM on 1 September 2010Urban Heat Islands: serious problem or holiday destination for skeptics?
RSVP, presumably because it varies, this is a 'basic' writeup, and the urban - rural anomaly is completely irrelevant? As the article explains, absolute temperature of any given location doesn't impact anomaly readings... we are looking at the change in temperature over time, and that is consistent between urban and rural locations. You might as well ask why not list the difference in temperature between Nome and Miami? Because that is no more irrelevant than the urban vs rural difference you ask for. -
RSVP at 00:10 AM on 1 September 2010Urban Heat Islands: serious problem or holiday destination for skeptics?
"...results in a slightly warmer envelope of air over urbanised areas when compared to surrounding rural areas.." Why not express this in degrees centrigrade? -
Tom Dayton at 00:03 AM on 1 September 2010The surprising result when you compare bad weather stations to good stations
No, omnologos, the definition of a "bad" site does not mean the site fails to correctly measure anything. The siting guidelines do not even include the label "bad." The label "bad" was invented by Watts to exaggerate the shortcomings of some stations, in Watts's desperate attempt to discredit AGW. -
omnologos at 23:49 PM on 31 August 2010The surprising result when you compare bad weather stations to good stations
@adelady and @Dappledwater: I understand my description of the logical issue is too short and cryptic. I will try to reformulate it. Basically if a measurement site needs to follow certain guidelines, then if it doesn't follow those guidelines it should not be included with those sites that do follow those guidelines. Or alternatively, if those guidelines are shown not to matter, then what is the point of the guidelines in the first place? "The warming signal is so significant it shows up in both good and bad sites" is not an answer: if a site is "bad", then by definition it does _not_ correctly measure anything. Same regarding "consistent measurement": a consistently-bad measurement is consistently wrong, again by definition. If a trend appears and it is "right", it's just pure luck with no scientific value or basis (unless, once again, somebody is able to show that the guidelines really don't matter). The above is valid for any measurement setting, not just temperatures. Hopefully nobody wants to destroy the foundations of "measurement science" this side of Galileo. -
bobconsole at 23:34 PM on 31 August 2010Climate's changed before
Lung cancer existed before cigarettes were invented too. Does that fact invalidate cigarette induced lung cancer? -
Rob Painting at 23:28 PM on 31 August 2010Sea level rise: the broader picture
Ken, that set of posts was entertaining. And yet here you are still banging on the same old drum. -
Ken Lambert at 23:10 PM on 31 August 2010Sea level rise: the broader picture
And lo and behold: Here's the quote from #42: "Chris and I have debated the TOPEX - Jason satellite transition - and if the two trends are linearized - there is a flattening in the Jason record closer to 2.0mm than 3.2mm per annum often quoted over the combined record. There is also the probably of an offset error in the transition." -
The empirical evidence that humans are causing global warming
Unrecovered - Exactly, I was talking about the transient state. However, given the thermal inertia of ocean warming, that transient can last for a long time... The 0.9 W/m^2 is on the edge of measurement tech right now, but the difference in emission spectra is measurable and clear. And, very importantly, indicates that conditions on Earth have to change in order to radiate the same amount of energy as before that spectral change (reduction), primarily by warming up. -
Rob Painting at 23:08 PM on 31 August 2010The surprising result when you compare bad weather stations to good stations
Omnologos @ 10 - "The graphs show there is something effecting an upwards trend in both "well sited" and "poorly sited" stations." Yes, global warming. The warming signal is so significant it shows up in both good and bad sites. A more interesting question is why the bad sites show less warming, which is why you should read the intermediate version here -
Ken Lambert at 22:56 PM on 31 August 2010Sea level rise: the broader picture
BP #19 Another interesting post. We dealt with the Jason-Topex in the 'Astronomical Cycle' thread here: http://www.skepticalscience.com/news.php?p=3&t=141&&n=232 I came up with a similar flattening of Jason and an offset between the two linear trends of Jason and Topex from a pencil and ruler analysis. -
adelady at 22:54 PM on 31 August 2010The surprising result when you compare bad weather stations to good stations
omnologos Correct measurement of temperature isn't the prime requirement for climate trends. Consistent measurement is the important issue. For climate trends we're only interested in the change in temperatures rather than the temp itself. That's why it's important when a new and better thermometer is installed at a site, the first thing that is done is to note the records accordingly. The new temperature recordings mustn't be allowed to falsely show a change in actual temperatures. -
muoncounter at 22:51 PM on 31 August 2010Human CO2: Peddling Myths About The Carbon Cycle
#14: "there is far greater capacity to sequester CO2 than is fully realised. " If, by sequester, you mean temporarily house within the plant growth for half of the annual cycle. Take those cycles out and you're still left with an uptrend -- which is not, as far as I can see, part of the annual cycle. What is left that is also increasing from year to year and is also the right multiple of the annual atmospheric increase? -
Rob Painting at 22:34 PM on 31 August 2010Sea level rise: the broader picture
JohnD - note the familiar figures (although unrelated) , from Church et al 2006: Sea level rise around the Australian coastline and the changing frequency of extreme sea-level events "Australian sea-level records for the period 1920-2000 clearly indicate a rise in relative mean sea level. Averaged around Australia, the rate of increase is about 1.2 mm per year. This value is less than the global increase in eustatic sea level for two reasons. First, the sea-level rises presented here are relative sea level and do not include any correction for ongoing crustal motion. To estimate eustatic sea-level change from the data from the Australian sites, the rates of sea-level rise would typically need to be increased by about 0.3 mm per year. Second, at least for the period 1950 to 2000, sea-level rise off western Australia less than the global average (Church et al. 2004), possibly as a result of the trend to more frequent, persisent and intense ENSO events since the mid-1970s." -
omnologos at 22:28 PM on 31 August 2010The surprising result when you compare bad weather stations to good stations
This poses a logical issue. The graphs show there is something effecting an upwards trend in both "well sited" and "poorly sited" stations. But if "poor siting" doesn't matter, _either_ the USHCN guidelines are useless/unimportant for a correct measurement of temperatures and there is no such a thing as "poor siting" (i.e. "warming" can be detected whatever the quality of the siting) _or_ the USHCN guidelines are useful/important for a correct measurement of temperatures, and therefore the upwards trend can't be ascribed to an actual "warming". ?? -
Rob Painting at 21:52 PM on 31 August 2010Sea level rise: the broader picture
Almost forgot, these guys also found: An Anomalous Recent Acceleration of Global Sea Level Rise Some quotes: "If the early twentieth-century acceleration is global in scale, we speculate that it differs from the recent acceleration in that the Northern Hemisphere apparently contributed significantly to the early event but little to the recent event." BP take note of the above. "1) the global sea level rise rate has accelerated from 1.5 mm yr prior to 1990 to a present day rate close to 3.2 mm yr" & "2) the acceleration in global sea level is accounted for primarily by the tropical and southern oceans, because of a phase change in the way the two regions covary: out of phase during relatively steady global sea level trends, and in phase during the trend increase" Which gels with the satellite altimetry and Seaframe stations mentioned in Ned's post @ 38. -
johnd at 21:43 PM on 31 August 2010Sea level rise: the broader picture
Dappledwater at 21:09 PM, the article referenced was about the establishment of the Tidal Reference Frame For Queensland. It utilised the established rate of sea level rise averaging 0.0003 metres, 0.3mmm per annum for the Australian continent (Mitchell, 2002). It detailed the implementation of the tidal datum epoch (TDE) which is the interval recommended for the calculation of datums. The TDE is normally longer than 18.6 years in order to include a full lunar nodal cycle, so looking for trends should only be looked for over multiple TDE's. From another source, the Permanent Committee for Tides and Mean Sea Level (PCTMSL) recommended that a 20-year TDE, 1992-2011 inclusive, be adopted for the determination of Lowest Astronomical Tide and Highest Astronomical Tide, so you see we are not at the end of that period yet. -
CBDunkerson at 21:26 PM on 31 August 2010Can humans affect global climate?
Singer is the worst of the worst. Cigarettes are not harmful. Asbestos is good for you. Acid rain does not exist. There is no ozone hole. Name a just plain crazy wrong 'scientific' position of the last thirty years and chances are he has been one of its prime proponents. That said, the statement quoted above is classic Singer. Pretty much everything he says takes some minor element of 'truth' and uses it as the foundation for a complete lie. In this case, "one-twentieth of a degree by 2050" is an extremely low, but not completely impossible, estimate while "imperceptible effect on future temperatures" is an outright lie. This is because most of the warming between now and 2050 is already 'locked in'. Complying with the Kyoto benchmark (1990 emissions levels by 2012) and sticking to that level would very likely have more than a 0.05 C impact on the 2050 global temp anomaly, but it would also mean at least 1 C difference in 2100 and even more further out. -
Ned at 21:10 PM on 31 August 2010Sea level rise: the broader picture
Miekol writes: Is science sea levels different to island sea levels? and links to Jo Nova's website, where Ms Nova writes about sea level data measured at a network of stations on various South and Southwest Pacific islands. We discussed that network of stations recently in another thread. If you ignore Ms Nova's selective and misleading analysis, and look at the full data for all of the islands in the network, you will find individual trends ranging from +3.2 to +8.6 mm/year (ignoring one outlier of +16.8 mm/year). The mean is +5.4 mm/year and the median is +4.9 mm/year. This compares very well with the satellite-derived sea level trend, which is only +3.3 mm/year (over the past two decades) for the world as a whole but higher in the Southwest Pacific. For some unfathomable reason, Ms Nova fails to quote the annual reports from this network of stations: The net sea level trends are positive at all sites, which indicates sea level in the region has risen over the duration of the project. The sea level rise is not geographically uniform but varies spatially in broad agreement with observations taken by satellite altimeters over a similar timeframe. [...] The sea level trends from SEAFRAME stations are mostly higher than the global average rate, but this is consistent with higher rates in the southwest Pacific measured by satellite altimeters
Prev 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 Next