No twentieth century warming

ice out day of Lake Winnipesaukee

ice out day of Lake Winnipesaukee

The earliest ice outs, and thus the warmest years, were in the 1950s.  The warming from 1975 to 1998 is real, but not very large – just a blip in the overall cooling trend from 1950.

During the 1950s, Winnipesaukee ice outs were generally around Julian day 100, which is usually April 10.  These days, it is around Julian day 107, which is usually April 17 – the ice is taking longer to melt, therefore, these days the weather is cooler than it was in the 1950s.

Data and calculations here – you will not find that replicability if you look at “peer reviewed” research.

If one selects the lakes with the best data, no sign of twentieth century warming.

Similarly for the weather station record. which indicates no twentieth century warming

Sea ice areas are unchanged, and sea levels rise is small and slowing, consistent with the overall shape of the above graph.

The vertical axis is the Julian day number of the ice outs.  The little green dots are the julian numbers of ice out days for particular years, the graph is the seven year rolling average of the ice out day.  Smaller numbers, representing warmer years, are at the top of the graph, larger numbers, representing cooler years, are at the bottom.  Lake Winnipesaukee was selected because lots of people who have no interest in global warming have an interest in this ice out day and report it, because it enables them to do business and get to their properties, hence the ice out day is a valid number.  It is the top hit in google for ice outs, excluding global warming related hits.

Lakes whose ice out day is subject to less interest are likely to have the same reliability problems as the instrumental temperature record.  If one has to look hard for historical data, that data is unlikely to be accurate because few people were monitoring it.  Since the data is likely to be inaccurate, one can always cherry pick data that proves anything one wants to prove, which cherry picking is apt to slowly become making up data outright – since one already supposedly knows what the data should show, searching overly hard for data that one knows must be true is apt to become outright forgery, as the Climategate documents directory shows happened with the instrumental temperature record.

If there are gaps in the data, as there generally is with the instrumental record and with many lake iceouts, that means few people are monitoring it.  If few, then perhaps none, perhaps the data is consciously or unconsciously fraudulent, and whether legitimate or not, no way to prove it legitimate.  Incompleteness is a symptom of other problems.  If a lake is so obscure that one cannot find a lake ice out for this year on the internet from boating enthusiasts, did anyone really find the lake ice out for 1950, or did they just make it up?

It is hard to estimate global climate from the instrumental record, but the most plausible evidence, if we exclude cities for the urban hot spot effect, and use only climate stations with good stability, refraining from efforts to patch together lots of fragmentary climate station records of unclear provenance, is that the 1950s were the warmest period in the twentieth century, and that the warming from 1975 to 1998 was just a small fluctuation in the long term cooling trend since 1950

The documents directory of the climategate files reveals that Harry could not derive global temperatures from the instrumental data, that the hadcrut global temperatures, the supposed instrumental record, had come right out of Tim’s ass.

Various people have attempted to reconstruct the global temperature record from the instrumental record but the data are not of quality that would enable this to be done.  The jumps resulting from moves of weather stations and suchlike are much larger than the climate changes one is trying to detect.  You cannot get a silk purse out of a sow’s ear.  Weather stations keep moving their thermometers, changing the way they record data, and so on and so forth.  Until 2004 no one was trying to do measurements that would be suitable for evaluating climate change – some would say they were trying from 1998, but they surely were not succeeding until 2004.

The only instrumental global temperature records are, like the climategate instrumental record, the result of someone’s secret sauce which they will not reveal, for no one who will reveal his method of calculation can produce anything from weather station data that they will claim to be plausible or credible.  Lots of people have tried.

For some examples of the difficulties encountered, see  Watts up with that, xx, xx. The GHCN adjustments were obviously fraudulent and intended to create fake global warming in the instrumental record, but it is impossible to say what adjustment would be reasonable and uncontroversial.  There is room enough in the gaps between one weather station and the next to manufacture global warming, global cooling, or have temperatures dance the watutsi.  Obvious the CRU was wrong to exclude those weather stations that showed world temperatures falling – but if the world climate trend depends on what stations you exclude and what you include, one can have little confidence in the trend derived from any particular set of weather stations.

Science is replication, not peer review.

What happens behind the scenes in peer review was revealed in the climategate emails

Peer review, as revealed in the climategate emails, is in practice theological review.  If it is peer reviewed, it is a lie. If it appears in “Nature” or “Science” it is a lie.  The truth is not allowed.

Which fact is obvious from the fact that non peer reviewed reports show their data and method of calculation, and peer reviewed reports on political  topics do not.  If it is peer reviewed, it cannot be replicated.  If it cannot be replicated, is not science.

Working through these non peer reviewed reports, replicating them, one can prove that various peer reviewed reports are criminal frauds. When we got the inside info, when the climategate files came out, we found in the documents directory the programs that did what we had already proven had been done.  Climategate confirmed what replicated research had already proven.  Replicated is the gold standard.  Peer reviewed is not.

We had already proven that global warming was criminal fraud.  Then we got confirmation from inside in the Climategate files, proving that our account of how global warming was cooked up was indeed how it was cooked up.

10 Responses to “No twentieth century warming”

  1. Looked-at-the-data says:


    I looked at the ice-out data and see a clear warming trend. Over the whole data set, over the past 50 years…any way I cut it. There is no “warming trend” since the 1950’s.

    Your selection of a 7 year moving average puzzles me. Can you explain how this would help show a long term trend like global warming?

    Just use a simple linear regression, and you’ll see a negative slope. The functionality is right there in Excel; just chart your data, and then add a linear trendline. It is the default option, and the most useful.

    It shows that every few years ice-out comes on average a little sooner.

    I also see you claim the 1950’s had the average earliest date for ice-out. And it does contain two low values of Julian day 93 for April 3rd. But that value appears again in 2006. But climate analysis is not about any one year, or finding a couple of outliers, but rather long term trends. So when I calculate the average for each decade I get 107.1 for the 1950 – 1959 years, but the slightly earlier (warmer) numbers of 106.7 and 106.9 for the 1980’s and 1990’s respectively.

    (If you use the technical definition of a decade starting in the year ending in 1, then the 1950’s are 1951 thru 1960 — you still get an average of 107.1 as 1950 and 1960 have the same Julian date of 110, and now the 1990’s move down to 105.8 as 1990’s 112 is replaced with 2000’s 100.)

    The point is that the 1950’s do not represent a warm time, followed by a cooling trend. The data simply does not support this claim.

    I would encourage all of your readers to spend 10 minutes looking at the actual data set. I believe they will find your analysis tortured and misleading.

    • Looked-at-the-data says:


      My first reply should read “cooling trend” in the first paragraph. You claim there is a cooling trend since the 1950’s when nothing in the data supports this claim.

      • jim says:

        Ice outs indicate it is cooler now than in 1950 at Lake Winnipesaukee. If you want to say that the fact is warmer now than in 1975 is a warming trend, then the fact that it is cooler now than in 1950 is a cooling trend. Of course, the fact that the trend depends on the start date means there is no significant twentieth century trend.

        • Looked-at-the-data says:

          Why do you say it is cooler now than in 1950’s? The dates of ice out do not show this.

          • jim says:

            Taking the rolling seven year average of ice out dates, ice out in 1950 was around day 101, ice outs today are around day 109

            Taking the raw data, (displayed in spreadsheet you get by clicking on the graph) the earliest ice out in recent times was day 96 in 2006, while the earliest ice out near 1950 was day 89 in 1946

            So the ice melts later these days, than it did back then.

          • Looked-at-the-data says:

            I believe you need to invent crazy patterns like 7 year averages to get the data to fit the conclusion you want to reach.

            The irony is that this is exactly what the Global-warming-doubters accuse the “mainstream” science community of doing. When in fact, the shoe is on the other foot.

            A linear trendline from 1950 – 2009 shows a slope of negative .06 as the average date has moved up 3 days in 60 years. You might not like the line, or the slope — but there it is. The slope over the past 30 years is negative .03. Over the past 75 years it is negative .03, over the past hundred years is negative .013. Over the whole historic record it is negative .04.

            The slope of every trendline is negative, as the average ice-out date is earlier each year.

            You need to pick crazy bulls eyes and distract the readers mind to make them see “cooling trends” in a data set that clearly shows warming.

            • jim says:

              I believe you need to invent crazy patterns like 7 year averages to get the data to fit the conclusion you want to reach.

              The data is all there in the spreadsheet you get when you click on the graph. If you think a seven year rolling average is a “crazy pattern” try a non crazy pattern.

              A linear trendline from 1950 – 2009 shows a slope of negative .06 as the average date has moved up 3 days in 60 years.

              The large fluctuations from trend, mean that you can get any trend you want, depending on how you do it, mean that the trend line is meaningless.

              If climate change is caused by anthropogenic CO2, rather than natural fluctuations, we would not be seeing such large deviations from trendline. The argument is that climate change fits well to anthropogenic CO2. Manifestly it does not fit.

              The argument is that current warm temperatures are unprecedented, and therefore a threat to life on earth. Manifestly they are not. Things were warmer a few decades back. See also Just Data which gives a very similar graph from surface station readings.

    • jim says:

      A mark one eyeball indicates that the fluctuations are so large as to render twentieth trend line meaningless.

  2. […] Jim’s Blog Liberty in an unfree world « No twentieth century warming […]

Leave a Reply