Third Assessment Report
(TAR) - 2000
Comments and Reviews - Part 5
The Intergovernmental Panel on Climate Change (IPCC) has recently drafted a third assessment report (TAR-2000) intending it for circulation to experts for review.
This forum is open to expert and non-expert alike to submit comments. Comments should be addressed to firstname.lastname@example.org with `TAR-2000 Review' in the subject line.
|The 1920s warming due to re-instrumentation
Comparison of two Siberian stations - one of them faulty
Clarification of the Kahl et al paper re Arctic temperatures
Start dates for Siberian graph
Response to George Birchard re Siberian trends post-1930s
Explains post-WW1 warmings as due to station upgrades
`sea ice is melting, glaciers are retreating, Arctic is warming' etc
1998 was a transitory event caused by El Niño
More temperatures from the sub-Arctic. Again, no warming.
The atmosphere analogous to a thermos flask?
The significance of the station data from Arctic stations
Station data shows Arctic ice cannot melt significantly
Model explanation as to why Arctic temps may not rise yet
Greenhouse effect unable to explain Arctic trends
Response to Mike MacCracken re Arctic anomalies
You are not quite right about the surface temeprature record, which is compiled in the way I described in my recent paper.
The world is divided into 5°x5° squares (latitude by longitude) and the accepted monthly temperature readings in that square are amalgamated by a weighting procedure This average monthly figure is then subtracted from the average for the same grid for the same month for the reference period (1960-1990) to give the monthly anomaly. The monthly anomalies are then averaged to give the annual anomaly. . There are, in fact three different procedures for weighting, those of Jones et al, Hansen et al and Peterson et al. It is claimed that all three methods give the same results. Blank squares are just left blank, and they are not included in the average. Easter Island would fill one square.
My set of annual anomalies are the ones used by Karl in his map, for the period 1901-1996, and they include only those for which there was a minimum number of 72 annual readings out of the 96. I therefore do not have any reading for the Easter Island square.
I argue that the system is upwardly biased because the increase in human population over the years, plus the increase in buildings, and fuel consumption, inevitably influence the surface temperature readings, often in a subtle way that is difficult to detect. Every time the station is upgraded or changed, the temperature rises slightly. When the stations moved to airports after 1945 the temperature fell for a while. After 1976 the traffic at the airports increased steadily, many non-airport stations were closed; so we have our "unprecedented" "anthropogenically induced" temperature change.
As I have previously published on your website, the 1901-1996 annual anomalies cover only 938 out of the possible 2592 5°x5° grids, 45.7% of the global area.
I have recently calculated that 486 of these 938 grids (52%) were disrupted by the First World War, to the extent of providing no annual figure for at least one of the years 1914-1919. There are very few regions that had annual figures for each of these years. They include USA, Western Europe, India, Western Australia and New Zealand. All these countries showed no temperature increase for the years 1910 to 1945. I therefore deduce that the large increase of 0.5°C in surface temperature between 1910 and 1945, which has previously been attributed to "recovery from the little ice age", was actually caused by the disruption of Met stations because of the world war.
This effect shows up clearly, John, in your chart for Vardo, Norway, which suddenly jumped from a steady average of about 0.5°C between 1840 and 1914 to a steady average of about 1.5°C between 1920 and 1999. The improvements in the Arctic were needed more than elsewhere
I attach my most extreme example, Isfjord Radio, Svalbard which jumped 7.3°C between 1917 and 1920. No wonder the Arctic "warmed". It took place all at once, suddenly, when the Met stations were rehabilitated after the war.
simply do not know what stations are accepted or rejected by those who
compile the temperature record, but it might make sense when there are
many stations in a single grid. The problems arise for the many grids (the
majority) where there is only one or two stations, where spurious
readings are impossible to detect. Most of the Arctic stations and alsmost
all of the Siberian stations fall into this category. In central USA and
UK there are regular checks on calibration, accuracy and reliability. In
the wilds of the Arctic, supervision from headquarters is minimal, and
false readings are difficult to check. Nobody from the University of East
Anglia goes to Siberian airports to find where the instruments are kept,
or how the conditions have
changed over the past twenty years.
This is why the satellite readings are much more reliable. Regards
You mentioned in a previous message that you met a Russian researcher who told you that the data falsification by officials in the old Soviet Union may also have extended to meteorological data for purposes of influencing state fuel allocations to remote locations in Siberia.
Vincent demonstrated in his paper that Siberian records are not consistent with the satellite data for the same area, due to local artificial factors. The sheer size of Siberia results in its stations influencing the hemispheric and global average.
Below is a graph comparing two adjacent stations in the vast Siberian interior, Dzardzan (68.7N 124E, located on the giant Lena River) and Olenek (300 miles due west of dzardzan over flat country on the same latitude). Their geographical circumstances suggest they should exhibit both the same temperature and the same long-term trend. Both stations are designated as `rural' by GISS, thus heat islands are not at issue here.
Dzardzan is well-connected to Yakutsk (a large city to the south) via the Lena River, but Olenek is quite isolated from normal means of communication, particularly in winter.
What is remarkable is that Olenek is around 2°C colder than Dzardzan right up to 1966, but then we find a sudden `warming' jump in 1967, with the two stations then following closely in lockstep, as we would expect, with similar temperatures right up to the present (except for a brief hiccup in 1993-94).
While Dzardzan exhibits a very slight cooling over the 60 year record (consistent with other records from the Arctic we have been discussing), Olenek has a 2°C `warming' over the same period, all achieved in one jump in 1967. That warming is clearly not climatic, but artificial, since there was no similar permanent jump at Dzardzan.
Why this warming jump should occur is unknown. Instrument change, site change, procedural change, or data falsification previous to 1966 (to make Olenek 2°C colder to get more fuel) - any one of these explanations might apply. Whatever the reason, it is clearly artificial.
This raises the question as to what GISS or CRU did with Olenek. Did they put that data, complete with its artificial jump, into the statistical grinder to contaminate the overall regional results? Or was it corrected first as it clearly needs to be? The data is on the GISS website, so one can infer that it has been used as is. If so, we have a clear indication from the comparison between these two stations that Russian data may be highly suspect and that `warming' trends which may appear in the aggreggate results may be no more than the cumulative errors from places like Olenek. The fact that the satellites are inconsistent with the surface over Siberia suggests Olenek may not be an isolated case.
Dzardzan has a consistent record with no dramatic shifts in the long-term trend, thus making its record credible. If data falsification at Olenek was the reason for the pre-1966 disparity, its isolated location may have provided the motive, in contrast with the better river communications at Dzardzan. During the period of Soviet central economic planning, all kinds of economic and social data originating outside Moscow was falsified for economic and welfare reasons before being submitted to the planning authorities. Simple temperature data would hardly be immune from this practice.
As to the coastal and island stations in the Russian Arctic, these were probably military stations and thus much less likely to return faulty data.
Regards John Daly
You say: "I am also still puzzled by the fact that Richard gets excited, when he can find a minus-sign in the paper [by Kahl et al.], but manages to ignore the sentence "As predicted by theory, Arctic surface temperatures have shown accelerated warming trends during the past century.""
The cooling trend reported in the paper by Kahl et al. is the major conclusion of that paper. The sentence you quote is an introductory comment that 'sets the scene' for the work reported in that paper and does not relate to the work reported in that paper. I am not "excited" by the major conclusion of that paper: I am interested and informed by it. And I do not "ignore" any part of that paper: I scrutinise all the information I have to assess its pertinence. Your insinuation that I do other is not warranted.
< side remarks omitted>
You said "Where numerous stations in the same region show similar data, it then becomes material evidence that the region as a whole is trending the same way the stations in that region show it to be. In the case of the Arctic and Antarctic, that means a slight cooling".
Since when? You have clained that the late 30's were exceptionally warm. I agree. You have biased your conclusion by starting in the 30's on the graph you present of Siberian data.
George Birchard wrote:
"Where numerous stations in the same region show similar data, it then becomes material evidence that the region as a whole is trending the same way the stations in that region show it to be. In the case of the Arctic and Antarctic, that means a slight cooling".
Since when? You have clained that the late 30's were exceptionally warm. I agree. You have biased your conclusion by starting in the 30's on the graph you present of Siberian data.
The graph you refer to was the 3-in-1 graph. The reason for the start date in the 1930s was explained in the email, namely that two of the stations began their data then. The third, Turuhansk, began its data back in 1881, and I explained that it had shown a warming trend during the 1920s, a common feature of most stations with such a long record. Vincent Gray has already raised questions as to whether the 1920s warming was real or not.
I was not trying to bias the the conclusion, as two of the stations simply did not have any data before that date, and I did acknowledge the longer data span with Turuhansk, and the fact that it had warmed in the 1920s, when I could have just left it unsaid. Also Turuhansk is not strictly within the Arctic Circle, but about 100 miles south of it. Since the 1920s warming (assuming it to be real) is not even an issue these days (even IPCC scientists accept it was caused by the sun, not by greenhouse gases), I saw no reason to visually compress the post-1930s part of the graph (with all three stations) by taking up unnecessary horizontal screen space with the isolated pre-1930s Turuhansk data. It was just a judgement call, not an intent to deceive, which is why I explained what I had done in the text of the message.
The data comes from GISS at http://www.giss.nasa.gov/data/update/gistemp/station_data/ this URL having been quoted before on the debate file. I take the GISS raw data and graph it up myself using the spreadsheet module of Microsoft Works, which gives a better visual result than the unwieldy enlongated charts provided by GISS.
Regards John Daly
John L. Daly
"Still Waiting For Greenhouse"
The conventional wisdom as to what is happening to the Arctic comes from the surface temperature record.
The surface temperature record is compiled from individual station records. The world is divided into 5°x5° latitude/longitude squares and the weighted mean measured temperature in that square subtracted from the average of the same weighted mean for the same squares for 1960-1990.
The Arctic Circle is 67°. On the TAR map for 1901-1996 there are 26 grids above 65°. Most of these contain only one individual measurement station, in most cases renewed and renovated many times over the period. Surprisingly, from my analysis, most had continuous records through World War 1.
The reported temperature rises 1901-1996 for this group are, on average, positive. But they are highly variable. There is one large negative value of -2.12°C for the Greenland Sea, and a huge positive, of +4.06°C for Svalbard, only 1500 metres away which gave an unprecedented jump of over 7°C during World War 1. See my chart for Isfjord Radio. As John Daly has shown with Vardo, Norway, several others also show the WW1 jump. In the case of Vardo, the site shows subsequernt cooling, but the surface record shows an overall rise since 1901, because of the 1914-18 jump.
The pattern for the second TAR map, for 1976-1998 is different. There are still only 26 or so stations above 65°N. The trend over this period is positive, because of the 1998 El Niño, and probably also because of subtle upgrading over the period. The cover is not exactly thorough. The MSU satellite units do not detect a cooling from 1979-2000. The Antarctic is undoubtedly cooling from both measurements.
Do you prefer proxies?
Regards Vincent Gray
As I've said before, I don't consider the measurement of latent heat by virtue of ice and snow quantities as proxies. Latent heat is real. Also real is the relative heat capacity of the water in the Arctic Ocean. Do you consider the overall Arctic Ocean temperature rise to be a proxy as well?
I've also explained how the North Atlantic Current is forced more into the Arctic now from less sea ice up north and therefore more friction from wind currents, robbing the Antarctic circumpolar flow of the warmed Indian Ocean flow, which is why "The Antarctic is undoubtedly cooling". This is compounding the warming so the rates of northern sea ice melt and current increases should continue to increase until the ice is gone, and some new mechanism will control the thermohaline circulation, although at a much higher rate of heat transfer into the arctic. I believe this will give us a climate jump similar to the ones that show up frequently in the paleoclimatic record. However, if you don't even believe that sea ice is melting, glaciers are retreating and the Arctic Ocean water is warming I suppose that's all lost on you.
So, back to basics. You blame the rise in 1976 -1998 arctic temperatures on the 1998 El Nino. I don't take that as a given. I believe they could both be the result of something else.
And, as I've said before, single station data is anecdotal evidence. If most of the 26 5x5 grids north of 65 degrees latitude have at least one measurement station, why is it you and John keep citing the same 4 or 5 stations over and over?
Do you prefer anecdotes?
If the ocean is really warming I would not deny it. How many sites have measured it? Is it also based on "anecdotes"? Or is it measured by satellites? But you do not believe satellite measurements, or do you?
I merely pointed out that the surface record is certainly based on anecdotes, and confusing ones at that.. You can have all 26 of them if you want. Don't tempt me.
Whatever the cause of the 1998 temperature rise, it was detected by both the surface record and by the MSU satellites. In fact the latter showed a stronger effect. The MSU temperature has now reverted to its usual zero
The 1998 temperature rise was transitory. Since the MSU units have not detected an overall temperature rise since 1979, the 1998 event could not have been caused by the greenhouse effect. What is your candidate, if it not El Niño?
Dear Vincent (and you NASA MSU guys),
I believe in the MSU data. Did you see my thermos explanation of why the stratosphere is cooling? I'm going to modify it a little bit. Put a small light bulb inside a thermos mostly full of water (the water's the modification), partially evacuated with a valve and hooked to a vacuum pump. Turn the light bulb on and the outside of the thermos will come up to that equilibrium temperature in which energy in through the cord equals energy out from the temperature differential and heat loss of the thermos shell to the surroundings. Then evacuate the thermos more. The temperature of the outside of the thermos will drop until the inside of the thermos comes up to the new internal equilibrium temperature in which transfer through the wall and out of the system again equals the energy coming in, which is when the outside of the thermos is at the same temperature as before.
The outside of the thermos is analogous to the stratosphere. The water is analogous to the massive (compared to the atmosphere) heat capacity of the oceans. The amount of vacuum (insulation, or R value) of the thermos is analogous to the amount of greenhouse gases in the atmosphere. The shell cools because the insulation has increased and the water inside has not yet come up to equilibrium temperature. This shows that since the stratosphere is cooling, the insulation of the atmosphere below it is increasing. The reason the troposphere is not heating much is because the ocean has to come up to the equilibrium temperature.
Therefore, stratospheric cooling SHOULD be more apparent than tropospheric warming because of its very small heat capacity compared to the oceans.
Does that not make sense?
It has been claimed here that station records from islands and coastal locations in the Arctic Ocean region may be exhibiting cooling due to latent withdrawal from the atmosphere from ice melt, although this is possibly an excuse to explain away the lack of warming registered at the weather stations there.
However, latent heat withdrawal cannot explain similar temperature patterns inland in Russia and Finland where the influence of latent heat withdrawal will be absent. Here is Turuhansk, Siberia (just east of the Urals, and just below the Arctic Circle), a record very similar to the coastal stations and hardly supportive of the big warmings predicted for the polar/sub-polar regions. Taken in conjunction with Dzardzan (shown above in an earlier message), this is clear evidence that the lack of warming is not restricted to the Arctic Ocean (and its alleged latent heat effects), but is also evident in the inland regions.
Further west we have a pair of stations, again just below the Arctic Circle, Kajaani in central Finland close to the Russian border and Reboly just across the border in Russia. Note how the two stations are in close lockstep. This demonstrates that stations such as these cannot be considered as isolated cases, or `anecdotal' oddities, but indicative of conditions over broad regions. Their close match not only provides mutual validation, but also demonstrates that station records are in no way `anecdotal' (an inappropriate and meaningless term in this context) but solid evidence of a zero trend throughout the Finnish-Russian borderland.
There's a simple message here. At high latitude regions, where warming should be greatest (according to the models and the theory), and where ocean latent heat effects are non-existent, there has been no warming, there is no warming, and there is no hint of any future warming.
Regards John Daly
I write to provide some clarity to the recent argument concerning the usefulness of Arctic surface station temperature data. You say to Vincent Grey: "And, as I've said before, single station data is anecdotal evidence." And you ask: "If most of the 26 5x5 grids north of 65 degrees latitude have at least one measurement station, why is it you and John keep citing the same 4 or 5 stations over and over?"
In the context of the debate prior to your above statement, the single station data is definitive proof and NOT "anecdotal evidence". The reason for this is as follows.
Kahl et al. had determined that the Arctic Basin had cooled in the period 1950 to 1990. Franz Gerl suggested that the cooling detected by Kahl et al. was probably constrained to the Central and Western Arctic Ocean because cooling was not exhibited by land-based stations outside these regions. Gerl's suggestion would be disproven if cooling between 1950 and 1990 were indicated by any one Arctic land-based station outside the Central and Western Arctic Ocean. Grey and Daly have responded by providing data from 6 such stations that exhibit the cooling. This is definitive proof that Gerl's suggestion is not correct. Q.E.D.
Also, please note that Grey and Daly are suggesting distortion of Arctic station temperature data from anthropogenic artifacts that falsely indicate warming. The distortion would occur if such an artifact existed in data from only one of the Arctic stations because all the data sets are agregated in the GISS record. Grey and Daly have demonstrated that 6 of the data sets contain such an artifact. This is definitive proof that the GISS record falsely indicates warming as a result of anthropogenic artifacts. Again, Q.E.D.
Also, Grey and Daly have demonstrated that the distortion to the GISS record is widespread because the 6 stations they cite are widely seperated geographically. And the distortion to the GISS record is likely to be large because the 6 distorted data sets are 20% of the 26 total data sets in the Arctic region.
I hope this clarifies the matter for you.
All the best Richard
Steve Hemphill said:
... the rates of northern sea ice melt and current increases should continue to increase until the ice is gone, and some new mechanism will control the thermohaline circulation...
This a dramatic and sweeping assertion, unproved, unreferenced, unsubstantiated in any way - just asserted as if mere assertion was sufficient to establish its truth. But let's indulge the ridiculous for a moment and take a reality check as to what it would take to achieve a total melt of the Arctic Ocean.
Here we have Ostrov Kotel, a Russian island in the Arctic Ocean off the northeast coast of Siberia. This place is permanently icebound and the reason is obvious from the chart. The temperature is at around -14°C.
First point to make is the lack of warming - any warming.
That not-so-minor point aside, look what it would take to get Ostrov Kotel ice free. It would have to warm up a full 12°C just to reach the melting point of seawater at -2°C (the baseline of the graph). Now look at the trend of the graph in the context of already rising CO2. The effect of CO2 has been, not to put too fine a point on it, zero.
We can see that Kotel has a long, long way to go for its temperature to lift that massive 12°C just to get to that critical melting point. In fact, it has not even started.
At this rate, we could be waiting until Hell freezes over. At least Hell would find itself in the right place.
Cheers John Daly
To all--I am very interested in this discussion about the Arctic (my thesis three decades ago involved climate modeling of ice ages, including the Ewing-Donn hypothesis about the Arctic driving glacial cycling), but have been unable to really participate (or even keep up with all the msgs). Just a few thoughts:
1. The conventional wisdom that the Arctic should now be warming much more rapidly than lower latitudes originated from doubled CO2 equilibrium runs--so what the climate would be like a long time after CO2 doubling had occurred. A great mistake in all the discussion has been to think that this warming might occur uniformly over time--that is that one could linearly interpolate in time. This is simply not the case. One likely gets a slow lengthening of the melt season (and temps don't rise much if at all during this period--in fact the sfc temp of salt water is less than of the freshwater on the ice, so one could imagine very little rise in temp in the summer. The cold season starts to get shorter, but one would keep getting the very cold temps below the inversion for a long time--until the insulating effect of the sea ice goes away--and that means it has to get very thin. Thus, this expectation of seeing very rapid warming in the high latitudes is, to a large extent, a misinterpretation of the model results. The new transient models also suggest that changes in ocean mixing, etc. can lead to cooling in some high latitude regions early on, so again, the expectation that everything will be linear is, I believe, largely mistaken and one needs to be rechecking one's perceptions against the transient climate model runs--not interpolating from old equilibrium models.
2. This all also means that one must be very careful in thinking about trends at particular points--both due to representativeness problems, but also in time, as the sfc temp will stay low for a long time as the ice thins, then once it melts and the open sea can absorb solar radiation in the summer, the open water will keep the inversion from forming at all as intensely, and we'll see a huge jump in sfc temp records (basically, a much longer open water season and shorter winter--that is how the temp will change--not a rise of a lot in summer temp or an increase of a lot in the coldest winter temp--but mostly at first changes in season lengths).
3. I think we can all agree one has to be careful of urban warming or other potential biases in the records, and attempts are made to remove these, but some may well still remain. However, there are other indications than temp msmts that climate change is occurring--among them that the sea ice is melting. We are having a USGCRP seminar today on this topic (see notice included at the end of the msg). The sea ice can thin without the temp rising--in fact it should be that way. It is not the temp that melts the sea ice, but the energy fluxes that are entering (think of your iced drink--it stays cold until all the ice melts--but heat is flowing in).
4. There have been suggestions that we should use the MSU data to validate the sfc data. Nothing could be further from the truth, it seems to me. In the Arctic, the sfc and troposphere are dynamically decoupled. Look at the old model results and you will see the sfc warming is high as the sfc inversion is weakened and the troposphere warming is small (less than in the tropics). Analysis of the MSU records and the sfc records shows that their correlations range from close to 1 to close to -1 in particular places, and this occurs for very valid scientific reasons--a long story we should have a discussion on this (but basically, the atmosphere moves--and so takes whatever anomaly one might have and mixes it around, also changing lapse rates with the circulation changes. I hope no one believes, for example, that the atmosphere could sustain the same anomaly as the eastern Pacific during an El Nino, when the anomaly is a couple of degrees--the atmosphere moves, and does so much faster than the oceans--and so moves the anomaly around). The only purpose we should put the MSU records to is to get at the long-term trend, and to do that we need to filter out the ENSO and volcanic influences (which it takes models to do)--or go long enough that we have enough of those events to average over (so wait longer than 2 decades). It makes no sense at all to be comparing the MSU anomaly over the former Soviet Union to the surface anomaly--sorry, but the spatial pattern of the MSU tells us much more about circulation (and so lapse rate) changes in the atmosphere than about thermodynamic changes at the surface. That there has been all this discussion about patterns of changes (by those of all sides on this debate) seems to me a great mistake.
Well--sorry for the short interjection and then bailing out for a bit, but I just sense the discussion is way off track.
Michael C. MacCracken, Ph.D.
National Assessment Coordination Office
Suite 750 400 Virginia Avenue
Washington DC 20024
Tel: (202) 314-2230 (Main number for NACO and for Robert
Cherry, Admin. Asst.) Tel: (202) 314-2233 (office and voicemail) Fax: (202)
488-8681 or (202) 488-8678
E-mail: email@example.com USGCRP Home Page: http://www.nacc.usgcrp.gov/
Dear Steve, and others
We agree, do we, that the amalgamated surface temperature record is a bunch of unreliable anecdotes, mostly biased upwards by changes in the surroundings of the measurement stations. John Daly has been trying to find examples where the effects of the surroundings were minimal, but it is a difficult task.
Then, the MSU units measure temperature in the lower troposphere, where there has been no overall change between 1979 and 2000. I just tried out their results for the Arctic region. They show no overall change between 1979 and 1998, but a rise of around 0.5°C from 1998 until now. So there has been some recent Arctic warming? will it persist?
Steve says that the ice is thinner, so there is a temperature change there.
Then there is the stratosphere. It has been cooling.
So we are down to a difference in semantics. Our "environment", around our individual dwelloings, cities and villages, is warming. The Arctic sea ice is warming. The "climate" as represented by the lower troposphere, is steady.. . The upper climate, represented by the stratosphere, is cooling.
How are these to be explained. Surely not the greenhouse effect!.
Regards Vincent Gray
1. The conventional wisdom that the Arctic should now be warming much more rapidly than lower latitudes originated from doubled CO2 equilibrium > runs--so what the climate would be like a long time after CO2 doubling had occurred.
As you yourself point out, that piece of conventional wisdom originated with the modellers, and was presumably based on their understanding of how CO2 behaves, an understanding which was then incorporated in the models.
A great mistake in all the discussion has been to think that this warming might occur uniformly over time--that is that one could linearly interpolate in time. This is simply not the case. One likely gets a slow lengthening of the melt season (and temps don't rise much if at all during this period--in fact the sfc temp of salt water is less than of the freshwater on the ice, so one could imagine very little rise in temp in the summer. The cold season starts to get shorter, but one would keep getting the very cold temps below the inversion for a long time--until the insulating effect of the sea ice goes away--and that means it has to get very thin. Thus, this expectation of seeing very rapid warming in the high latitudes is, to a large extent, a misinterpretation of the model results.
That is a restatement of the latent heat argument put earlier. It's a sound view in itself as even a glass of water containing ice cubes will not warm until the cubes have melted.
But there are two problems with that interpretation as far as the Arctic is concerned
Firstly, the lack of warming at the stations is not confined to the Arctic Ocean coasts and islands, but there is also similar lack of warming far inland in Siberia and Finland, well away from the latent heat effects of sea ice. See Kajaani (Finland), Ruboly (Russia), Turuhansk (Russia) and Dzardzan (Russia) to see what I mean. All these stations are far inland from the Arctic ocean and its ice, yet they too show no warming.
Secondly, many stations exist in ice-free parts of the Arctic, such as Akureyri (Icland), Vardo (Norway), Jan Mayen Island, Bear Island, and since there is no ice around, there also cannot be any latent heat effect from melting ice. But these too are showing the same neutral trend as the others.
The new transient models also suggest that changes in ocean mixing, etc. can lead to cooling in some high latitude regions early on, so again, the expectation that everything will be linear is, I believe, largely mistaken and one needs to be rechecking one's perceptions against the transient climate model runs--not interpolating from old equilibrium models.
The question which would spring to mind there is whether the models re-created this cooling as a natural outcome of the machine calculations, or whether this outcome was an artifact of changed model inputs to make them accord more closely with observed conditions. If the latter was the case, the model is really retrodicting, not predicting, and thus not to be treated as a predictive tool.
... the sfc temp will stay low for a long time as the ice thins, then once it melts and the open sea can absorb solar radiation in the summer, the open water will keep the inversion from forming at all as intensely, and we'll see a huge jump in sfc temp records (basically, a much longer open water season and shorter winter--that is how the temp will change--not a rise of a lot in summer temp or an increase of a lot in the coldest winter temp--but mostly at first changes in season lengths).
You say `will' on the basis of what a model is telling you. But without observational evidence from stations to support that, it is just a modelling conjecture. And I repeat that this latent heat effect you postulate will not apply inland or in ice-free areas, both of which also show no warming, when by your analysis they should have begun warming strongly by now. Indeed, Vardo and Akureyri should be showing rampant warming by now based on the theory you state above, but they too show no warming.
3. I think we can all agree one has to be careful of urban warming or other potential biases in the records, and attempts are made to remove these, but some may well still remain.
I gave an earlier example of Ankara (pop. 2.5 million) where the GISS urban correction amounted to a mere 0.2°C over 100 years. That told me one thing - GISS's urban corrections are woefully inadequate. It allows them to say with authority "We have corrected for urbanisation", with little inkling by the general public or policymakers just how small these corrections are, so small as to be meaningless. This is why I have called for a rigorous independent review of the surface record of the kind the MSU has been subjected to. Current peer review is simply not good enough, and is not seen to be rigorous and impartial.
4. There have been suggestions that we should use the MSU data to validate the sfc data. Nothing could be further from the truth, it seems to me. In the Arctic, the sfc and troposphere are dynamically decoupled.
I presume you include the Antarctic in that. Because we have the same lack of interior warming there too. Indeed, while Arctic ice extent is reported to be shrinking in recent years, the Antarctic sea ice is reported to be expanding. The notion that decoupling between the surface and troposphere in the Arctic might explain the lack of warming at the stations has now been extended by IPCC scientists to basically include the whole earth. This is their novel way to explain away the disparity between the satellites and the GISS surface record. What began as a reasonable explanation to account for lack of warming in the icy part of the Arctic Ocean (though not the ice-free part, nor the inland), has now been adopted as a means to explain away the satellite-surface divergence globally.
Their reasoning goes something like - "The satellites are right, but so is the surface. So the divergent trend between them must be due to a global change in the lapse rate or other similar decouplings between surface and troposphere". You must excuse me if I find that just a tad far-fetched especially given that the whole troposphere is characterised in the models as a well-mixed layer and thus incapable of this kind of decoupling. That reasoning by the IPCC is `explaining away', not explaining. It's too neat and convenient and avoids the central issue as to whether the GISS surface record is any good or not. That's why an independent and rigorous review is so important, to determine once and for all if the surface record can be trusted and to make modifications to it if necessary as was done to the satellites when some small errors were found with them. Until that review is done, passing off the divergence as some quirk in the lapse rate does the science of climate no credit at all.
John L. Daly
"Still Waiting For Greenhouse"
As this is the fifth file (average 100K each) in this TAR-2000 debate, there will be no further messages posted here as all the substantive issues have been thoroughly aired. My thanks to all contributors - until the next `Open Review' - JD