Hiding The Decline - Part 4

In another guest post at Watts Up With That, Willis Eschenbach looks at some emails between Phil Jones and Kevin Trenberth of the CRU to/from Professor Wibjorn Karlen who had asked some very pointed questions about the CRU and IPCC temperature results for Europe and other parts of the world. Again what he found was more of the same from the CRU, incomplete, incorrect and at times misleading answers to the questions put to them. A strategy that has become all too familiar. To ease the reading of this I have written Willis' comments and general information from his article in blue with any highlights in red. Karlen's comments in purple, Trenberth's comments in green and Jones' comments in maroon. That way it is clear who is commenting and when.

One of the claims in this hacked CRU email saga goes something like “Well, the scientists acted like jerks, but that doesn’t affect the results, it’s still warming.”

I got intrigued by one of the hacked CRU emails, from the Phil Jones and Kevin Trenberth to Professor Wibjorn Karlen. In it, Professor Karlen asked some very pointed questions about the CRU and IPCC results. He got incomplete, incorrect and very misleading answers. Here’s the story, complete with pictures. I have labeled the text to make it clear who is speaking, including my comments.

From Jones and Trenberth to Wibjorn Karlen, 17 Sep 2008 (email # 1221683947).

[Trenberth]Hi Wibjorn

It appears that your concern is mainly with the surface temperature record, and my co lead author in IPCC, Phil Jones, is best able to address those questions. However the IPCC only uses published data plus their extensions and in our Chapter the sources of the data are well documented, along with their characteristics. I offer a few more comments below (my comments are limited as I am on vacation and away from my office).

[Karlen to Trenberth]Uppsala 17 September 2008,

Dear Kevin,

In short, the problem is that I cannot find data supporting the temperature curves in IPCC and also published in e.g. Forster, P. et al. 2007: Assessing uncertainty in climate simulation. Nature 4: 63-64.

[My comments] Here is the figure from Nature, Assessing uncertainty in climate simulations, Piers Forster et al., Nature Reports Climate Change , 63 (2007) doi:10.1038/climate.2007.46a

Original Caption: Figure 1: Comparison of observed continental- and global-scale changes in surface temperature with results simulated by climate models using natural and anthropogenic forcings. Decadal averages of observations are shown for the period 1906 to 2005 (black line) plotted against the centre of the decade and relative to the corresponding average for 1901–1950. Lines are dashed where spatial coverage is less than 50%. Blue shaded bands show the 5–95% range for 19 simulations from five climate models using only the natural forcings due to solar activity and volcanoes. Red shaded bands show the 5–95% range for 58 simulations from 14 climate models using both natural and anthropogenic forcings. SOURCE: http://www.nature.com/climate/2007/0709/full/climate.2007.46a.html

Here is the IPCC figure he is referring to, Fig. 9.12, once again with the black lines showing the instrumentally measured temperatures:

Original Caption: Figure 9.12. Comparison of multi-model data set 20C3M model simulations containing all forcings (red shaded regions) and containing natural forcings only (blue shaded regions) with observed decadal mean temperature changes (°C) from 1906 to 2005 from the Hadley Centre/Climatic Research Unit gridded surface temperature data set (HadCRUT3; Brohan et al., 2006). The panel labelled GLO shows comparison for global mean; LAN, global land; and OCE, global ocean data. Remaining panels display results for 22 sub-continental scale regions (see the Supplementary Material, Appendix 9.C for a description of the regions).…

Note that around the globe, temperatures are shown as rising from 1900 to about 1930, falling or staying level until the mid ’70s, and then rising sharply after that.

So these are the curves that Professor Karlen is attempting to reconstruct. Note that the IPCC chapter identifies these as “sub-continental regions” and shows separate data for ocean regions.

[Karlen] In attempts to reconstruct the temperature I find an increase from the early 1900s to ca 1935, a trend down until the mid 1970s and so another increase to about the same temperature level as in the late 1930s.

A distinct warming to a temperature about 0.5 deg C above the level 1940 is reported in the IPCC diagrams. I have been searching for this recent increase, which is very important for the discussion about a possible human influence on climate, but I have basically failed to find an increase above the late 1930s.

[Trenberth] This region, as I am sure you know, suffers from missing data and large gaps spatially. How one covered both can greatly influence the outcome.

In IPCC we produce an Arctic curve and describe its problems and

character. In IPCC the result is very conservative owing to lack of

inclusion of the Arctic where dramatic decreases in sea ice in recent

years have taken place: 2005 was lowest at the time we did our assessment but 2007 is now the record closely followed by 2008.

Anomalies of over 5C are evident in some areas in SSTs but the SSTs are not established if there was ice there previously. These and other indicators show that there is no doubt about recent warming; see also chapter 4 of IPCC.

[My comment] As I will show below, everything he says about the ocean and the sea ice and the sea surface temperatures (SSTs) is meaningless. The IPCC figure is solely for the land.

[Karlen] In my letter to Klass V I included diagram showing the mean annual temperature of the Nordic countries (1890-ca 2001) presented on the net by the database NORDKLIM, a joint project between the meteorological institutes in the Nordic countries. Except for Denmark, the data sets show an increase after the 1970s to the same level as in the late 1930s or lower. None demonstrates the distinct increase IPCC indicates. The trends of these 6 areas are very similar except for a few interesting details.

[Trenberth] Results will also depend on the exact region.

[My comments] I cannot find the NORDKLIM graphic he refers to, so I have calculated it myself. I used the NORDKLIM dataset available at http://www.smhi.se/hfa_coord/nordklim/data/Nordklim_data_set_v1_0_2002.xls. I removed the one marine record from “Ship M”. To avoid infilling where there are missing records, I took the “first difference” of all of the available records for each year and averaged them. Then I used a running sum to calculate the average anomaly. I did not remove cities or adjust for the Urban Heat Island (UHI) effect. Here is the result:

You can see that, as Professor Karlen said, this does not show what the “Northern Europe” part of the IPCC graph shows. It is exactly as Professor Karlen stated, in the NORDKLIM data it rises until 1930, there is a drop from 1930 to 1970, followed by an increase after the 1970s to a temperature slightly lower than the 1930s. (In fact, the rise from 1880 until 1930 dwarfs the recent rise since the 1970’s). Here, for comparison, is a blowup of the “Northern Europe” graph from Fig. 9.12 above:

This claims that there is a full degree temperature rise from 1970 to 2000, ending way warmer than the 1930s. You can see why Professor Karlen is wondering how the IPCC got such a different answer.

[Karlen] I have in my studies of temperatures also checked a number of areas using data from NASA. One, in my mind interesting study, includes all the 13 stations with long and decent continuously records north of 65 deg N.

The pattern is the same as for the Nordic countries. This diagram only shows 11-yr means of individual stations. A few stations such as Verhojans and Svalbard indicate a recent mean 11-year temperature increase up to 0.5 deg C above the late 1930s. Verhojansk, shows this increase but the

temperature has after the peak temperature decreased with about 0.3 deg C during the last few years. The majority of the stations show that the recent temperatures are similar to the one in the late 1930s.

In preparation of some talks I have been invited to give, I have expanded the Nordic area both west and east. The area of similar change in climate is vast. Only a few stations near Bering Strait deviates (e.g. St Paul, Kodiak, Nome, located south of 65 deg. N).

My studies include Africa, a study which took me most of a summer because there are a large number of stations in the NASA records. I found 11 stations including data from 1898-1975 and 16 stations including 1950-2003.

The data sets could in a convincing way be spliced. However, I noticed that some persons were not familiar with ’splicing’ technique so I have accepted to reduce the study to the 7 stations including data from the whole period between 1898-2003. The results are similar as to the spliced data set and

also, surprisingly similar to the variability of the Nordic data.

Regression indicates a minor (if any) decrease in temperature (I have used all stations independent of location, city location or not).

[Trenberth] Africa is notorious for missing and inaccurate data and needs careful assessment.

[Karlen] Another example is Australia. NASA only presents 3 stations covering the period 1897-1992. What kind of data is the IPCC Australia diagram based on?

If any trend it is a slight cooling. However, if a shorter period (1949-2005) is used, the temperature has increased substantially.

The Australians have many stations and have published more detailed maps of changes and trends.

There are more examples, but I think this is much enough for my present point:

How has the laboratories feeding IPCC with temperature records selected stations?

[Trenberth] See our chapter and the appendices.

[My comment] I have looked at these. The source for Fig. 9.1.2 is given as “(HadCRUT3; Brohan et al., 2006)”. HadCRUT3 is produced jointly by CRU and the Hadley Centre.

[Karlen] I have noticed that major cities often demonstrate a major urban effect (Buenos Aires, Osaka, New York Central Park, etc). Have data from major cities been used by the laboratories sending data to IPCC? Lennart Bengtsson and other claims that the urban effect is accounted for but from what I read, it seems like the technique used has been a simplistic

[Trenberth] Major inner cities are excluded: their climate change is real but very local.

[My comment] It is true that the IPCC Chapter 3 FAQ says this:

Additional warming occurs in cities and urban areas (often referred to as the urban heat island effect), but is confined in spatial extent, and its effects are allowed for both by excluding as many of the affected sites as possible from the global temperature data and by increasing the error range (the blue band in the figure).

To check this claim, I took the list of temperature stations used by CRU (which I had to use an FOI to get), and checked them against the GISS list. The GISS list categorizes stations as “Urban” or “Rural”. It also uses satellite photos to categorize the amount of light that shows at night, with big cities being brightest. It puts them into three categories, A, B, and C. C is the brightest.

It turns out that there are over 500 cities in the CRU database that the GISS database categorizes as “Urban C”, the brightest of cities. These include, among many others:

AUCKLAND, NEW ZEALAND

BANGKOK METROPOLIS, THAILAND

BARCELONA, SPAIN

BEIJING, CHINA

BRASILIA, BRAZIL

BRISBANE, AUSTRALIA

BUENOS AIRES, ARGENTINA

CHRISTCHURCH, NEW ZEALAND

DHAKA, BANGLADESH

FLORENCE, ITALY

GLASGOW, UK

GUATEMALA CITY, GUATEMALA

HANNOVER, GERMANY

INCHON, KOREA

KHARTOUM, SUDAN

KYOTO, JAPAN

LISBON, PORTUGAL

LUXOR, EGYPT

MARRAKECH, MOROCCO

MOMBASA, KENYA

MOSKVA, RUSSIAN FEDERA

MOSUL, IRAQ

NAGASAKI, JAPAN

NAGOYA, JAPAN

NICE, FRANCE

OSAKA, JAPAN

PRETORIA, SOUTH AFRICA

RIYADH, SAUDI ARABIA

SAO PAULO, BRAZIL

SEOUL, KOREA

SHANGHAI, CHINA

SINGAPORE, SINGAPORE

STOCKHOLM, SWEDEN

TEGUCIGALPA, HONDURAS

TOKYO, JAPAN

VALENCIA, SPAIN

VOLGOGRAD, USSR

So the CRU is using Tokyo? Beijing? Seoul? Shanghai? Moscow? Their claim is entirely false. In other words, once again the good folk of the CRU are blowing smoke. I can understand why it took me a Freedom of Information request to get the station list.

[Karlen] Next step has been to compare my results with temperature records in the literature. One interesting figures is published by you in:

Trenberth, K., 2005: Uncertainty in Hurricanes and Global Warming. Science 308: 1753-1754.

As you obviously know, the recent increase in temperature above the 1940s is minor between 10 deg N and 20 deg N and only slightly larger above the temperature maximum in the early 1950s. Both the increases in temperature in the 1930s and in the 1980s to 1990s is of similar amplitude and similar steepness, if any difference possibly slightly less steep in the northern area than in the southern (the eddies slow down the warm water|transport?).

Your diagram describes a limited area of the North Atlantic because you are primarily interested in hurricanes. The complexity of sea surface temperature increases and decreases is seen in e.g. Cabanes, C, et al 2001 (Science 294: 840-842).

[Trenberth] As we discuss, there is a lot of natural variability in the North Atlantic but there is also a common component that relates to global changes. See my GRL article with Shea for more details. Trenberth, K. E., and D. J. Shea, 2006: Atlantic hurricanes and natural variability in 2005. Geophys. Res. Lett., 33, L12704, doi:10.1029/2006GL026894.

[Karlen] One example of sea surface temperature is published by:

Goldenberg, S.B., Landsea, C.W., Mestas-Nuoez, A.M. and Gray, W.M., 2001: The recent increases in Atlantic hurricane activity: causes and implications. Science 293: 474-479.

Again, there is a marked increase in temperature in the 1930s and 1950s (about 1 deg C), a decrease to approximately the level in the 1910s and thereafter a new increase to a temperature slightly below the level in the1940s.

One example of published data not supporting a major temperature increase during recent time is:

Polyakov, I.V., Bekryaev, R.V., Alekseev, G.H., Bhatt,U.S., Colony,

R.L., Johnson, M.A., Maskshtas, A.P. and Walsh, D., 2003: Variability and Trends of Air Temperature and Pressure in the Maritime Arctic, 1875-2000. Journal of Climate: Vol. 16 (12): 2067ñ2077.

He included many more stations than I did in my calculation of temperatures N 65 N, but the result is similar. It is hard to find evidence of a drastic warming of the Arctic.

It is also difficult to find evidence of a drastic warming outside urban areas in a large part of the world outside Europe. However the increase in temperature in Central Europe may be because the whole area is urbanized (see e.g. Bidwell, T., 2004: Scotobiology – the biology of darkness. Global change News Letter No. 58 June, 2004).

So, I find it necessary to object to the talk about a scaring temperature increase because of increased human release of CO2. In fact, the warming seems to be limited to densely populated areas. The often mentioned correlation between temperature and CO2 is not convincing. If there is a factor explaining a major part of changes in the temperature, it is solar irradiation. There are numerous studies demonstrating this correlation but papers are not accepted by IPCC. Most likely, any reduction of CO2 release will have no effect whatsoever on the temperature (independent of how expensive).

[Trenberth] You can object all you like but you are not looking at the evidence and you need to have a basis, which you have not established. You seem to doubt that CO2 has increased and that it is a greenhouse gas and you are very wrong. But of course there is a lot of variability and looking at one spot narrowly is not the way to see the big picture.

[My comment] Professor Karlen was quite correct. The claims made by the CRU, and repeated in the IPCC document, were false. Karlen was looking at the evidence.

[Karlen] In my mind, we have to accept that it is great if we can reduce the release of CO2 because we are using up a resource the earth will be short of in the future, but we are in error if we claims a global warming caused by CO2.

[Trenberth] I disagree.

[My comment] No comment.

[Karlen] I also think we had to protest when erroneous data like the claim that winter temperature in Abisko increased by 5.5 deg C during the last 100 years. The real increase is 0.4 deg C. The 5.5 deg C figure has been repeated a number of times in TV-programs. This kind of exaggerations is not supporting attempts to save fossil fuel.

I have numerous diagrams illustrating the discussion above. I don’t include these in an e-mail because my computer can only handle a few at a time. If you would like to see some, I can send them by air mail.

I am often asked about why I don’t publish about my views. I have. Just one example of among 100 other I could select is: Karlen, W., 2001: Global temperature forces by solar irradiation and greenhouse gases? Ambio 30(6): 349-350.

Yours sincerely

Wibjorn

Geografiska Annaler

Professor em Wibjorn Karlen

Department of Social and Economic Geography

Geografiska Annaler Ser. A

Box 513

SE-751 20 Uppsala

SWEDEN

[Trenberth] I trust that Phil Jones may also respond

Regards

Kevin Trenberth

___________________

Kevin Trenberth

Climate Analysis Section, NCAR

From: P.Jones

To: trenbert

Subject: Re: Climate

Date: Wed, 17 Sep 2008 16:39:07 +0100 (BST)

Cc: Wibjorn Karlen

[Jones to Professor Karlen, same email]Wibjorn,

I’m in Athens at the moment. Unless you’re referring specifically to the Arctic the temperature curves in IPCC Ch 3 all include the oceans.

[My comment] Absolutely not. The legend for Fig. 9.1.2 (see above) says “(see the Supplementary Material, Appendix 9.C for a description of the regions)” Appendix 9.C in turn describes the calculations:

6. Apply land/ocean mask on observations. Plots describing observed changes in land or ocean areas were based on observed data that was masked to retain land or ocean data only (necessary to remove islands and marine stations not existent in models). This masking was performed as in Step 3, using the land area fraction data from the CCSM3 model.

Note that the ocean is entirely masked out of the observations.

And the regions are described as:

Note 2: List of Regions

The regions are defined as the collection of rectangular boxes listed for each region. The domain of interest (land and ocean, land, or ocean) is also given.

REGION, DESIGNATOR, COVERAGE, DOMAIN

Global, GLO, 180W to 180E, 90S to 90N, land and ocean

Global Land, LAN, 180W to 180E, 90S to 90N, land

Global Ocean, OCE, 180W to 180E, 90S to 90N, ocean

North America, ALA, 170W to 103W, 60N to 72N, land

North America, CGI, 103W to 10W, 50N to 85N, land

North America, WNA, 130W to 103W, 30N to 60N, land

North America, CNA, 103W to 85W, 30N to 50N, land

North America, ENA, 85W to 50W, 25N to 50N, land

South America, CAM, 116W to 83W, 10N to 30N, land

South America, AMZ, 82W to 34W, 20S to 12N, land

South America, SSA, 76W to 40W, 56S to 20S, land

Europe, NEU, 10W to 40E, 48N to 75N, land

Europe, SEU, 10W to 40E, 30N to 48N, land

Africa, SAR, 20W to 65E, 18N to 30N, land

Africa, WAF, 20W to 22E, 12S to 18N, land

Africa, EAF, 22E to 52E, 12S to 18N, land

Africa, SAF, 10E to 52E, 35S to 12S, land

Asia, NAS, 40E to 180E, 50N to 70N, land

Asia, CAS, 40E to 75E, 30N to 50N, land

Asia, TIB, 75E to 100E, 30N to 50N, land

Asia, EAS, 100E to 145E, 20N to 50N, land

Asia, SAS, 65E to 100E, 5N to 30N, land

Asia, SEA, 95E to 155E, 11S to 20N, land

Australia, NAU, 110E to 155E, 30S to 11S, land

Australia, SAU, 110E to 155E, 45S to 30S, land

So no, that excuse won’t wash. Once again Professor Karlan is quite correct. The observations simply don’t match the CRU/IPCC claims. Phil Jones’ story about the regions including the ocean is false.

[Jones] Fennoscandia is just a small part of the NH. When I’m back next week, I’ll be able to calculate the boxes that encompass Fennoscandia, so you can compare with this region. As you’re aware Anders did lots of the update work in 2001-2002 and he included all the NORDKLIM data. I can send you a list of the Fennoscandian data if you want – either the sites used or their data as well.

I guess you’re attachments are in your direct email, which I come to later.

One final thing – we are getting SST data in from some of the new sea-ice free parts of the Arctic. We are not using these as we’ve yet to figure out how to as we don’t have normals for these ‘mostly covered by sea ice in the 1961-90′ areas.

Cheers

Phil

[My comments]Now, I have not taken a stand on whether the machinations of the CRU extended to actually altering the global temperature figures. It seems quite clear from Professor Karlen’s observations, however, that they have gotten it very wrong in at least the Fennoscandian region. Since this region has very good records and a lot of them, this does not bode well for the rest of the globe …

In what seems like a continuing trend the raw data is adjusted to show warming that doesn't appear to be there. When caught out, the team at CRU begin a campaign of misinformation and claims to the effect of "you are wrong" or "you don't have the information only we do" to silence the critic. No matter how based in evidence their criticism may be. Next we will examine Alaska.

Hiding The Decline - Part 3

As I alluded to at the end of part three the NIWA alterations were not alone. The following is at a site in Darwin Airport, Australia known as Darwin Zero. What guest writer for Watts Up With That, Willis Eschenbach found was quite extraordinary:

People keep saying “Yes, the Climategate scientists behaved badly. But that doesn’t mean the data is bad. That doesn’t mean the earth is not warming.”

Let me start with the second objection first. The earth has generally been warming since the Little Ice Age, around 1650. There is general agreement that the earth has warmed since then. See e.g. Akasofu . Climategate doesn’t affect that.

The second question, the integrity of the data, is different. People say “Yes, they destroyed emails, and hid from Freedom of information Acts, and messed with proxies, and fought to keep other scientists’ papers out of the journals … but that doesn’t affect the data, the data is still good.” Which sounds reasonable.

There are three main global temperature datasets. One is at the CRU, Climate Research Unit of the University of East Anglia, where we’ve been trying to get access to the raw numbers. One is at NOAA/GHCN, the Global Historical Climate Network. The final one is at NASA/GISS, the Goddard Institute for Space Studies. The three groups take raw data, and they “homogenize” it to remove things like when a station was moved to a warmer location and there’s a 2C jump in the temperature. The three global temperature records are usually called CRU, GISS, and GHCN. Both GISS and CRU, however, get almost all of their raw data from GHCN. All three produce very similar global historical temperature records from the raw data.

So I’m still on my multi-year quest to understand the climate data. You never know where this data chase will lead. This time, it has ended me up in Australia. I got to thinking about Professor Wibjorn Karlen’s statement about Australia that I quoted here:

Another example is Australia. NASA [GHCN] only presents 3 stations covering the period 1897-1992. What kind of data is the IPCC Australia diagram based on?

If any trend it is a slight cooling. However, if a shorter period (1949-2005) is used, the temperature has increased substantially. The Australians have many stations and have published more detailed maps of changes and trends.

The folks at CRU told Wibjorn that he was just plain wrong. Here’s what they said is right, the record that Wibjorn was talking about, Fig. 9.12 in the UN IPCC Fourth Assessment Report, showing Northern Australia:

Figure 1. Temperature trends and model results in Northern Australia. Black line is observations (From Fig. 9.12 from the UN IPCC Fourth Annual Report). Covers the area from 110E to 155E, and from 30S to 11S. Based on the CRU land temperature.) Data from the CRU.

One of the things that was revealed in the released CRU emails is that the CRU basically uses the Global Historical Climate Network (GHCN) dataset for its raw data. So I looked at the GHCN dataset. There, I find three stations in North Australia as Wibjorn had said, and nine stations in all of Australia, that cover the period 1900-2000. Here is the average of the GHCN unadjusted data for those three Northern stations, from AIS:

Figure 2. GHCN Raw Data, All 100-yr stations in IPCC area above.

So once again Wibjorn is correct, this looks nothing like the corresponding IPCC temperature record for Australia. But it’s too soon to tell. Professor Karlen is only showing 3 stations. Three is not a lot of stations, but that’s all of the century-long Australian records we have in the IPCC specified region. OK, we’ve seen the longest stations record, so lets throw more records into the mix. Here’s every station in the UN IPCC specified region which contains temperature records that extend up to the year 2000 no matter when they started, which is 30 stations.

Figure 3. GHCN Raw Data, All stations extending to 2000 in IPCC area above.

Still no similarity with IPCC. So I looked at every station in the area. That’s 222 stations. Here’s that result:

Figure 4. GHCN Raw Data, All stations extending to 2000 in IPCC area above.

So you can see why Wibjorn was concerned. This looks nothing like the UN IPCC data, which came from the CRU, which was based on the GHCN data. Why the difference?

The answer is, these graphs all use the raw GHCN data. But the IPCC uses the “adjusted” data. GHCN adjusts the data to remove what it calls “inhomogeneities”. So on a whim I thought I’d take a look at the first station on the list, Darwin Airport, so I could see what an inhomogeneity might look like when it was at home. And I could find out how large the GHCN adjustment for Darwin inhomogeneities was.

First, what is an “inhomogeneity”? I can do no better than quote from GHCN:

Most long-term climate stations have undergone changes that make a time series of their observations inhomogeneous. There are many causes for the discontinuities, including changes in instruments, shelters, the environment around the shelter, the location of the station, the time of observation, and the method used to calculate mean temperature. Often several of these occur at the same time, as is often the case with the introduction of automatic weather stations that is occurring in many parts of the world. Before one can reliably use such climate data for analysis of longterm climate change, adjustments are needed to compensate for the nonclimatic discontinuities.

That makes sense. The raw data will have jumps from station moves and the like. We don’t want to think it’s warming just because the thermometer was moved to a warmer location. Unpleasant as it may seem, we have to adjust for those as best we can.

I always like to start with the rawest data, so I can understand the adjustments. At Darwin there are five separate individual station records that are combined to make up the final Darwin record. These are the individual records of stations in the area, which are numbered from zero to four:

DATA SOURCE: http://data.giss.nasa.gov/cgi-bin/gistemp/findstation.py?datatype=gistemp&data_set=0&name=darwin

Figure 5. Five individual temperature records for Darwin, plus station count (green line). This raw data is downloaded from GISS, but GISS use the GHCN raw data as the starting point for their analysis.

Darwin does have a few advantages over other stations with multiple records. There is a continuous record from 1941 to the present (Station 1). There is also a continuous record covering a century. finally, the stations are in very close agreement over the entire period of the record. In fact, where there are multiple stations in operation they are so close that you can’t see the records behind Station Zero.

This is an ideal station, because it also illustrates many of the problems with the raw temperature station data.

• There is no one record that covers the whole period.

• The shortest record is only nine years long.

• There are gaps of a month and more in almost all of the records.

• It looks like there are problems with the data at around 1941.

• Most of the datasets are missing months.

• For most of the period there are few nearby stations.

• There is no one year covered by all five records.

• The temperature dropped over a six year period, from a high in 1936 to a low in 1941. The station did move in 1941 … but what happened in the previous six years?

In resolving station records, it’s a judgment call. First off, you have to decide if what you are looking at needs any changes at all. In Darwin’s case, it’s a close call. The record seems to be screwed up around 1941, but not in the year of the move.

Also, although the 1941 temperature shift seems large, I see a similar sized shift from 1992 to 1999. Looking at the whole picture, I think I’d vote to leave it as it is, that’s always the best option when you don’t have other evidence. First do no harm.

However, there’s a case to be made for adjusting it, particularly given the 1941 station move. If I decided to adjust Darwin, I’d do it like this:

Figure 6 A possible adjustment for Darwin. Black line shows the total amount of the adjustment, on the right scale, and shows the timing of the change.

I shifted the pre-1941 data down by about 0.6C. We end up with little change end to end in my “adjusted” data (shown in red), it’s neither warming nor cooling. However, it reduces the apparent cooling in the raw data. Post-1941, where the other records overlap, they are very close, so I wouldn’t adjust them in any way. Why should we adjust those, they all show exactly the same thing.

OK, so that’s how I’d homogenize the data if I had to, but I vote against adjusting it at all. It only changes one station record (Darwin Zero), and the rest are left untouched.

Then I went to look at what happens when the GHCN removes the “in-homogeneities” to “adjust” the data. Of the five raw datasets, the GHCN discards two, likely because they are short and duplicate existing longer records. The three remaining records are first “homogenized” and then averaged to give the “GHCN Adjusted” temperature record for Darwin.

To my great surprise, here’s what I found. To explain the full effect, I am showing this with both datasets starting at the same point (rather than ending at the same point as they are often shown).

Figure 7. GHCN homogeneity adjustments to Darwin Airport combined record

YIKES! Before getting homogenized, temperatures in Darwin were falling at 0.7 Celcius per century … but after the homogenization, they were warming at 1.2 Celcius per century. And the adjustment that they made was over two degrees per century … when those guys “adjust”, they don’t mess around. And the adjustment is an odd shape, with the adjustment first going stepwise, then climbing roughly to stop at 2.4C.

Of course, that led me to look at exactly how the GHCN “adjusts” the temperature data. Here’s what they say in An Overview of the GHCN Database:

GHCN temperature data include two different datasets: the original data and a homogeneity- adjusted dataset. All homogeneity testing was done on annual time series. The homogeneity- adjustment technique used two steps.

The first step was creating a homogeneous reference series for each station (Peterson and Easterling 1994). Building a completely homogeneous reference series using data with unknown inhomogeneities may be impossible, but we used several techniques to minimize any potential inhomogeneities in the reference series.

…

In creating each year’s first difference reference series, we used the five most highly correlated neighboring stations that had enough data to accurately model the candidate station.

…

The final technique we used to minimize inhomogeneities in the reference series used the mean of the central three values (of the five neighboring station values) to create the first difference reference series.

Fair enough, that all sounds good. They pick five neighboring stations, and average them. Then they compare the average to the station in question. If it looks wonky compared to the average of the reference five, they check any historical records for changes, and if necessary, they homogenize the poor data mercilessly. I have some problems with what they do to homogenize it, but that’s how they identify the inhomogeneous stations.

OK … but given the scarcity of stations in Australia, I wondered how they would find five “neighboring stations” in 1941 …

So I looked it up. The nearest station that covers the year 1941 is 500 km away from Darwin. Not only is it 500 km away, it is the only station within 750 km of Darwin that covers the 1941 time period. (It’s also a pub, Daly Waters Pub to be exact, but hey, it’s Australia, good on ya.) So there simply aren’t five stations to make a “reference series” out of to check the 1936-1941 drop at Darwin.

Intrigued by the curious shape of the average of the homogenized Darwin records, I then went to see how they had homogenized each of the individual station records. What made up that strange average shown in Fig. 7? I started at zero with the earliest record. Here is Station Zero at Darwin, showing the raw and the homogenized versions.

Figure 8 Darwin Zero Homogeneity Adjustments. Black line shows amount and timing of adjustments.

Yikes again, double yikes! What on earth justifies that adjustment? How can they do that? We have five different records covering Darwin from 1941 on. They all agree almost exactly. Why adjust them at all? They’ve just added a huge artificial totally imaginary trend to the last half of the raw data! Now it looks like the IPCC diagram in Figure 1, all right … but a six degree per century trend? And in the shape of a regular stepped pyramid climbing to heaven? What’s up with that?

Those, dear friends, are the clumsy fingerprints of someone messing with the data Egyptian style … they are indisputable evidence that the “homogenized” data has been changed to fit someone’s preconceptions about whether the earth is warming.

One thing is clear from this. People who say that “Climategate was only about scientists behaving badly, but the data is OK” are wrong. At least one part of the data is bad, too. The Smoking Gun for that statement is at Darwin Zero.

So once again, I’m left with an unsolved mystery. How and why did the GHCN “adjust” Darwin’s historical temperature to show radical warming? Why did they adjust it stepwise? Do Phil Jones and the CRU folks use the “adjusted” or the raw GHCN dataset? My guess is the adjusted one since it shows warming, but of course we still don’t know … because despite all of this, the CRU still hasn’t released the list of data that they actually use, just the station list.

Another odd fact, the GHCN adjusted Station 1 to match Darwin Zero’s strange adjustment, but they left Station 2 (which covers much of the same period, and as per Fig. 5 is in excellent agreement with Station Zero and Station 1) totally untouched. They only homogenized two of the three. Then they averaged them.

That way, you get an average that looks kinda real, I guess, it “hides the decline”.

Oh, and for what it’s worth, care to know the way that GISS deals with this problem? Well, they only use the Darwin data after 1963, a fine way of neatly avoiding the question … and also a fine way to throw away all of the inconveniently colder data prior to 1941. It’s likely a better choice than the GHCN monstrosity, but it’s a hard one to justify.

Now, I want to be clear here. The blatantly bogus GHCN adjustment for this one station does NOT mean that the earth is not warming. It also does NOT mean that the three records (CRU, GISS, and GHCN) are generally wrong either. This may be an isolated incident, we don’t know. But every time the data gets revised and homogenized, the trends keep increasing. Now GISS does their own adjustments. However, as they keep telling us, they get the same answer as GHCN gets … which makes their numbers suspicious as well.

And CRU? Who knows what they use? We’re still waiting on that one, no data yet …

What this does show is that there is at least one temperature station where the trend has been artificially increased to give a false warming where the raw data shows cooling. In addition, the average raw data for Northern Australia is quite different from the adjusted, so there must be a number of … mmm … let me say “interesting” adjustments in Northern Australia other than just Darwin.

And with the Latin saying “Falsus in unum, falsus in omis” (false in one, false in all) as our guide, until all of the station “adjustments” are examined, adjustments of CRU, GHCN, and GISS alike, we can’t trust anyone using homogenized numbers.

It would seem that the influence of the CRU and it's preference for adjusting data wasn't restricted to the Southern Hemisphere. As I will show in the subsequent posts on this topic.

Hiding The Decline - Part 2

After the Climategate emails were released, people began to look more closely at their own temperature data sets and also to see if there was any "CRU" influence involved. One of the first to be uncovered was New Zealand's National Institute of Water and Atmospheric Reaserch (NIWA). Note the use of chartmanship on the first graph.

New Zealand’s National Institute of Water & Atmospheric Research (NIWA) is responsible for New Zealand’s National Climate Database. This database, available online, holds all New Zealand’s climate data, including temperature readings, since the 1850s. Anybody can go and get the data for free. That’s what we did, and we made our own graph. Before we see that, let’s look at the official temperature record. This is NIWA’s graph of temperatures covering the last 156 years: From NIWA’s web site -

Mean annual temperature over New Zealand, from 1853 to 2008 inclusive, based on between 2 (from 1853) and 7 (from 1908) long-term station records. The blue and red bars show annual differences from the 1971 - 2000 average, the solid black line is a smoothed time series, and the dotted [straight] line is the linear trend over 1909 to 2008 (0.92C/100 years).

This graph is the centrepiece of NIWA’s temperature claims. It contributes to global temperature statistics and the IPCC reports. It is partly why our government is insisting on introducing an ETS scheme and participating in the climate conference in Copenhagen. But it’s an illusion.

Dr Jim Salinger (who no longer works for NIWA) started this graph in the 1980s when he was at CRU (Climate Research Unit at the University of East Anglia, UK) and it has been updated with the most recent data. It’s published on NIWA’s website and in their climate-related publications.

The actual thermometer readings

To get the original New Zealand temperature readings, you register on NIWA’s web site, download what you want and make your own graph. We did that, but the result looked nothing like the official graph. Instead, we were surprised to get this:

Straight away you can see there’s no slope - either up or down. The temperatures are remarkably constant way back to the 1850s. Of course, the temperature still varies from year to year, but the trend stays level - statistically insignificant at 0.06C per century since 1850. Putting these two graphs side by side, you can see huge differences. What is going on?

Why does NIWA’s graph show strong warming, but graphing their own raw data looks completely different? Their graph shows warming, but the actual temperature readings show none whatsoever! Have the readings in the official NIWA graph been adjusted?

It is relatively easy to find out. We compared raw data for each station (from NIWA’s web site) with the adjusted official data, which we obtained from one of Dr Salinger’s colleagues. Requests for this information from Dr Salinger himself over the years, by different scientists, have long gone unanswered, but now we might discover the truth.

Proof of man-made warming? What did we find? First, the station histories are unremarkable. There are no reasons for any large corrections. But we were astonished to find that strong adjustments have indeed been made. About half the adjustments actually created a warming trend where none existed; the other half greatly exaggerated existing warming. All the adjustments increased or even created a warming trend, with only one (Dunedin) going the other way and slightly reducing the original trend.

The shocking truth is that the oldest readings have been cranked way down and later readings artificially lifted to give a false impression of warming, as documented below. There is nothing in the station histories to warrant these adjustments and to date Dr Salinger and NIWA have not revealed why they did this.

See much more of this detailed analysis here. NIWA responds to the charges here but Anthony Watts uses instrument photo located at NIWA headquarters to cast doubt on their claims here. See also how the government is hell bent on moving forward with carbon emission schemes choosing to believe agenda driven government scientists here.

This was not to be an isolated incident, as Anthony Watts discovered in a guest post by Willis Eschenbach when he examined the temperature in Darwin Australia. I will cover this in Part three.

Hiding The Decline - Part 1

In amongst the most damning of the Climategate emails was one from CRU head Phil Jones in which he boasts of using "Mike's trick" to “hide the decline” that would have otherwise spoiled his graph showing temperatures soaring ever-upward. This was met with howls of protest from Jones, Mann, Schmidt and all their supporters that this comment was being taken "out of context" and that the term "trick" referred to nothing sneaky. Indeed the term "hiding the decline" wasn't what it appeared either, or was it?

In the Climategate context it referred to the Briffa tree ring reconstruction and how it showed a late 20th Century decline which was considered "very inconvenient" by the remainder of this close climate clique. Steve McIntyre did what the conspirators of Climategate demanded and put these comments into context as can be seen in full here.

Much recent attention has been paid to the email about the “trick” and the effort to “hide the decline”. Climate scientists have complained that this email has been taken “out of context”. In this case, I’m not sure that it’s in their interests that this email be placed in context because the context leads right back to a meeting of IPCC authors in Tanzania, raising serious questions about the role of IPCC itself in “hiding the decline” in the Briffa reconstruction.

Relevant Climategate correspondence in the period (September-October 1999) leading up to the trick email is incomplete, but, in context, is highly revealing. There was a meeting of IPCC lead authors between Sept 1-3, 1999 to consider the “zero-order draft” of the Third Assessment Report. The emails provide clear evidence that IPCC had already decided to include a proxy diagram reconstructing temperature for the past 1000 years and that a version of the proxy diagram was presented at the Tanzania meeting showing the late twentieth century decline. I now have a copy of the proxy diagram presented at this meeting.

The emails show that the late 20th century decline in the Briffa reconstruction was perceived by IPCC as “diluting the message”, that “everyone in the room at IPCC” thought that the Briffa decline was a “problem” and a “potential distraction/detraction”, that this was then the “most important issue” in chapter 2 of the IPCC report and that there was “pressure” on Briffa and other authors to show a “nice tidy story” of “unprecedented warming in a thousand years or more”.

...Climategate Letters, Sep 22-23, 1999

The Climategate Letters contain a flurry of correspondence between Mann, Briffa, Jones and Folland (copy to Tom Karl of NOAA) on Sep 22-23, 1999, shedding light on how the authors responded to the stone in IPCC’s shoe. By this time, it appears that each of the three authors (Jones, Mann and Briffa) had experimented with different approaches to the “problem” of the decline.

Jones appears to have floated the idea of using two different diagrams - one without the inconvenient Briffa reconstruction (presumably in the Summary for Policy-makers) and one with the Briffa reconstruction (presumably in the relevant chapter). Jones said that this might make it “somewhat awkward for the reader trying to put them into context”, with it being unclear whether Jones viewed this as an advantage or disadvantage:

....Thus, when Mann arrived at work on Sep 22, 1999, Mann observed that he had walked into a “hornet’s nest”. (Mann Sep 22, 1999, 0938018124.txt). In an effort to resolve the dispute, Mann said that (subject to the agreement of Chapter Authors Karl and Folland) he would add back Briffa’s reconstruction, but pointed out that this would present a “conundrum”:

....Mann went on to say that the skeptics would have a “field day” if the declining Briffa reconstruction were shown and that he’d “hate to be the one” to give them “fodder”:

...By the following day, matters seem to have settled down, with Briffa apologizing to Mann for his temporary pangs of conscience. On Oct 5, 1999, Osborn (on behalf of Briffa) sent Mann a revised version of the Briffa reconstruction with more “low-frequency” variability (Osborn, Oct 5, 1999, 0939154709.txt), a version that is identical up to 1960, this version is identical to the digital version archived at NCDC for Briffa et al (JGR 2001). (The post-1960 values of this version were not “shown” in the version archived at NCDC; they were deleted.)

As discussed below, this version had an even larger late-20th century decline than the version shown at the Tanzania Lead Authors’ meeting. Nonetheless, the First Order Draft (Oct 27, 1999) sent out a few weeks later contained a new version of the proxy diagram, a version which contains the main elements of the eventual Third Assessment Report proxy diagram . Two weeks later came Jones’ now infamous “trick” email (0942777075.txt).

The IPCC Trick

Mann’s IPCC trick is related to the Jones’ trick, but different. (The Jones trick has been explained in previous CA posts here, here and consists of replacing the tree ring data with temperature data after 1960 – thereby hiding the decline – and then showing the smoothed graph as a proxy reconstruction.) While some elements of the IPCC Trick can be identified with considerable certainty, other elements are still somewhat unclear.

The diagram below shows the IPCC version of the Briffa reconstruction (digitized from the IPCC 2001) compared to actual Briffa data from the Climategate email of October 5, 1999, smoothed using the methodology said to have been used in the caption to the IPCC figure (a 40 year Hamming filter with end-point padding with the mean of the closing 20 years).

Versions of the Briffa Reconstruction in controversy, comparing the original data smoothed according to the reported methodology to a digitization of the IPCC version.

Clearly, there are a number of important differences between the version sent to Mann and the version that appeared in the IPCC report. The most obvious is, of course, that the decline in the Briffa reconstruction has, for the most part, been deleted from the IPCC proxy diagram. However, there are some other frustrating inconsistencies and puzzles that are all too familiar.

There are some more technical inconsistencies that I’ll record for specialist readers. It is very unlikely that that the IPCC caption is correct in stating that a 40-year Hamming filter was used. Based on comparisons of the MBH reconstruction and Jones reconstruction, as well as the Briffa reconstruction, to versions constructed from raw data, it appears that a Butterworth filter was used – a filter frequently used in Mann’s subsequent work (a detail that, in addition, bears on the authorship of the graphic itself).

Second, the IPCC caption stated that “boundary constraints imposed by padding the series with its mean values during the first and last 25 years.” Again, this doesn’t seem to reconcile with efforts to replicate the IPCC version from raw data. It appears far more likely to me that each of the temperature series has been padded with instrumental temperatures rather than the mean values of the last 25 years.

Finally, there are puzzling changes in scale. The underlying annual data for the Jones and Briffa reconstructions are expressed in deg C (basis 1961-1990) and should scale simply to the smoothed version in the IPCC version, but don’t quite. This may partly derive from errors introduced in digitization, but is a loose end in present replication efforts.

The final IPCC diagram (2.21) is shown below. In this rendering, the Briffa reconstruction is obviously no longer “a problem and a potential distraction/detraction”and does not “dilute the message”. Mann has not given any “fodder” to the skeptics, who obviously did not have a “field day” with the decline.

IPCC Third Assessment Report Figure 2.21: Comparison of warm-season (Jones et al., 1998) and annual mean (Mann et al., 1998, 1999) multi-proxy-based and warm season tree-ring-based (Briffa, 2000) millennial Northern Hemisphere temperature reconstructions. The recent instrumental annual mean Northern Hemisphere temperature record to 1999 is shown for comparison. Also shown is an extra-tropical sampling of the Mann et al. (1999) temperature pattern reconstructions more directly comparable in its latitudinal sampling to the Jones et al. series. The self-consistently estimated two standard error limits (shaded region) for the smoothed Mann et al. (1999) series are shown. The horizontal zero line denotes the 1961 to 1990 reference period mean temperature. All series were smoothed with a 40-year Hamming-weights lowpass filter, with boundary constraints imposed by padding the series with its mean values during the first and last 25 years.

Contrary to claims by various climate scientists, the IPCC Third Assessment Report did not disclose the deletion of the post-1960 values. Nor did it discuss the “divergence problem”. Yes, there had been previous discussion of the problem in the peer-reviewed literature (Briffa et al 1998) – a point made over and over by Gavin Schmidt and others. But not in the IPCC Third Assessment Report. Nor was the deletion of the declining values reported or disclosed in the IPCC Third Assessment Report. [Dec 11.- IPCC TAR does contain a sly allusion to the problem; it mentions "evidence" that tree ring density variations had "changed in their response in recent decades". Contrary to claims of realclimate commenters, this does not constitute disclosure of the deletion of the post-1960 values in the controversial figure or even of the decline itself.] The hiding of the decline was made particularly artful because the potentially dangling 1960 endpoint of the Briffa reconstruction was hidden under other lines in the spaghetti graph as shown in the following blow-up:

No real surprises then that these people would resort to visual trickery to conceal what they should be showing. In the other parts to this series I intend to show how the CRU has has influence over other countries temperature sets and they too have hidden the decline or indeed just created a rise where none was seen before.