Dear Kenneth (I can not be sure that you are the real Kenneth Towe),

This will be my last answer to your posts. I would not answer but there will be non-specialist people, who may be reading this line of posts.

I do think that, you are intentionally distorting the information, or you really don't know and try to learn (as a paleobiologist, the latter condition would be interesting).

Yes, that is true that Antarctic ice sheets began to form at the end of the Eocene (around 36 million years ago). On the other hand, first ice caps emerged much earlier when CO2 levels were much higher (like 45 million years ago, most probably if you had found this reference at Google, you would have given this. But you must have missed it while googling). However, the conditions like today was not established until the Late Miocene-Early Pliocene ( around 6 million years ago). CO2 levels were similar to today's levels at the Early Pliocene. The reference you mentioned in your previous post (Pearson et al., 2009) is just a small adjustment to the dates and CO2 levels for Late Eocene. At the end of its abstract it adds: Overall, our results confirm the central role of declining pCO2 atm in the development of the Antarctic ice sheet (in broad agreement with carbon cycle modelling12) and help to constrain mechanisms and feedbacks associated with the Earth’s biggest climate switch of the past 65 Myr.

Moreover, if a student who has attended to an introductory climatology course would know that, especially, Antarctica's climate and development of the ice sheets on the Antarctica Continent can not be explained solely by CO2 levels. Opening of the Drake Passage, Formation of the Antarctic Circumpolar Current, Rise of the major mountain ranges (such as the Andes and Himalayas), closure of Panama and many other paleogeographic features most probably I forget to mention here.

All of these mean, you are not giving correct examples and what you write is disinformation.

Other than Kenneth Towe's statements, there is also an extended abstract about your question, which I find interesting. It claims, with the methods applied through this abstract, it should be more than 30 years (which is a "rule").

Here is the link:

http://iopscience.iop.org/article/10.1088/1755-1315/11/1/012028/meta

Thanks so much @Niels @Kenneth @Kezi I've found all your answers very helpful.

@Kenneth u mean some institutions controlling the data are constantly destroying it? How do you mean?

@Kenneth, data did and will always have uncertainties and measurement errors. That is the nature of data based science.

@Kenneth, you say: "Data are not constantly being destroyed. Data are constantly being adjusted. British data at CRU, East Anglia were destroyed.".

However, to be honest, I didn't understand the same thing from the link you have sent (maybe couldn't understand you). Especially gridded climate data differs from model to model, which depends on the station data used, interpolation/extrapolation methods, homogenization of the data etc.

For example, since I know the numbers in Turkey, there are more than 160 meteorology stations. Most of them are active from the 1950's (of course, there are years without any data). On the other hand, CRU reconstructions only use 7 of these station data. I am not able to give other example, but let's think about another reconstruction which uses 15 different data of totally different stations, it will make a difference. Moreover, CRU for example, does not actively follow the station data and their homogenization process may not be able to solve, for example, the transfer of a station. These all make the condition complicated.

Interesting question! The article (mentioned by Mr.Zeki) says 30 year sampling period based on probability density functions (PDF) for key meteorological variables. Is it an accepted rule? Regards

@Asit, it is a convention accepted in 1935 by the world meteorological organization as stated in the first paragraph of the extended abstract of Larsson (the link in my first post).

Dear Kenneth, it is just a report of Deutsche Bank, it seems. There are of course error terms. But within the confidence intervals, there is a trend. About the policy-makers, I am not sure that if it is good for them to know what is confidence interval:)

Attached is a table from IPCC-5, page 187. You can see there are different estimates (of course with error terms in parenthesis), but all show a trend. You can find a really deep analysis in IPCC-5, chapter 2.

Dear Kenneth, forecast models and reconstruction of the past are two different subjects. I think you know that and I will not get into details, also the subject is getting more out of my speciality. Even if you change the past climate values, it will not change future projections seriously.

The values of temperatures, whether real or anomalies, change over time when data is added, deleted or adjusted. This is not an attempt to manipulate the data, but to ensure the temperature measurements are valid as well as reliable. For example, over time it is discovered that a temperature station reading is affected by urban sprawl, which cause an increase in the temperature at the station due to the heat island effect. So, that is corrected and the temperature would be adjusted (lowered in this case). Sometimes it is discovered that the temperature devises being uses are providing inaccurate measures, and so the temperature readings are adjusted. Sometimes stations are removed and others are added, which can affect the mean. There are scientific reasons such adjustments are made. If these adjustments were not made, then we would say that the validity of the data is being ignored. Thus, the fact that the temperature mean in 1921 is estimated as X from Z stations in 1941, does not imply that the estimated means from the existing data was valid, or that the measures used were valid. Would you want to keep the 1921 mean the same as estimated in 1941 if you discovered the data were not valid?

Good discussion - I appreciate the remarks of Dr. Zeki, Dr. Kenneth and Dr. Michael. Regards

@Kenneth : wow... What could be the motive of NCDC?

SO, my suggestion that the differences in the data over time have to do with data validation is "a very strong assertion," and then, what should we make of your claim that these differences are a purposeful manipulation? The latter doesn't strike you as a "strong assertion" ? If what you suggests happens, that's a fraud. Do you have evidence of fraud? Or are you just assuming fraud because the temperature measures differ? Shouldn't scientists check their data and correct invalid measures?

Dear Kenneth, why do you keep giving examples from the first half of the 20th century? On the other hand, the strong hypothesis here is the greenhouse effects of CO2 and CH4, which, I think, will not be falsified (or for many years). If someone wants to blame the global climate change assertion, first they need to falsify the theory on greenhouse gases. The rest are just small parts of the story.

The discussion fell apart from the question. Dear Cyril, most of the forecast models, as far as I know, use the years 1960-1990 or 1970-2000 as training sets. I hope this helps.

@zeki: yes it does. Thanks so much

Dear Kenneth (I can not be sure that you are the real Kenneth Towe),

This will be my last answer to your posts. I would not answer but there will be non-specialist people, who may be reading this line of posts.

I do think that, you are intentionally distorting the information, or you really don't know and try to learn (as a paleobiologist, the latter condition would be interesting).

Yes, that is true that Antarctic ice sheets began to form at the end of the Eocene (around 36 million years ago). On the other hand, first ice caps emerged much earlier when CO2 levels were much higher (like 45 million years ago, most probably if you had found this reference at Google, you would have given this. But you must have missed it while googling). However, the conditions like today was not established until the Late Miocene-Early Pliocene ( around 6 million years ago). CO2 levels were similar to today's levels at the Early Pliocene. The reference you mentioned in your previous post (Pearson et al., 2009) is just a small adjustment to the dates and CO2 levels for Late Eocene. At the end of its abstract it adds: Overall, our results confirm the central role of declining pCO2 atm in the development of the Antarctic ice sheet (in broad agreement with carbon cycle modelling12) and help to constrain mechanisms and feedbacks associated with the Earth’s biggest climate switch of the past 65 Myr.

Moreover, if a student who has attended to an introductory climatology course would know that, especially, Antarctica's climate and development of the ice sheets on the Antarctica Continent can not be explained solely by CO2 levels. Opening of the Drake Passage, Formation of the Antarctic Circumpolar Current, Rise of the major mountain ranges (such as the Andes and Himalayas), closure of Panama and many other paleogeographic features most probably I forget to mention here.

All of these mean, you are not giving correct examples and what you write is disinformation.

If you’re life is throu environmental facts threaten , your now climate have changed .