Another warming question

Gokul43201

Your "observation" is unrelated to your "prediction". Furthermore, it is flat out wrong. Warming over the last decade has been stronger than over the last 50 years. And your interpretation of the claims (though listed under "observation") is also at best, misleading.

Andre

Perhaps not. Recheck this post

Gokul43201

Andre you know that those linear trends are extremely sensitive to the '98 data point. What happens to the trend once you've established reasonable insensitivity to end-points? It is meaningless to speak of a trend otherwise.

Andre

The trend is also extremely sensitive to removing the deep dip immediately following the spike of 1998. But perhaps we could conclude that it's not safe to say that warming over the last decade has been stronger than over the last 50 years

Gokul43201

1. That's a spurious argument and you ought to know that a lot better than most of us here. There is no "deep dip" after 1998. The appearance of a deep dip is merely an artifact of clipping the dataset conveniently at 1998. Your own posts in other threads establish that there is no "deep dip" after 1998 compared to temperatures just before 1998.

2. I've looked at the data with and without outliers of different strengths (2, 3, 4 standard deviations away from mean), and with each additional removal of outliers, the slope either increased significantly or changed very little. But even if you look at the 5 trend-lines as you have them in the linked post, with all outliers included, the average slope is 0.5K/cent, which is the same as the warming trend over the previous 50 or so years. Eliminating outliers or expanding the dataset and thus reducing the error bar on the trend to actually meaningful values gives a number that is roughly twice as big.

I admit that I should have been a lot more careful with my statement about the recent warming data. While the trend is positive to a high degree of confidence, the confidence level is much poorer in the difference with the trend over the previous 50 years.

Art

The IPCC predicted a temp rise that hasn't happened so what is your argument?? Even AGWs most ardent supporters now say it is likely we are in a 10 - 30 year cooling cycle. (which takes the heat off them for a few decades ) for eg http://www.theaustralian.news.com.au...9-7583,00.html

It seems your argument is a statistical one based on if you squint your eyes a certain way whilst looking at gently massaged, cherry picked data you can at a stretch perhaps argue the temp hasn't actually fallen. That's what I meant about AGW proponents desperately clinging to a failed hypothesis.

btw What did you do with the 'abnormal' global cooling figures of the 70's when creating your 50 year trend lines?

Andre

Perhaps have a look at the maximum extent of the data, the last 29 years, when in 1979 the satelites became operational.

http://gallery.myff.org/gallery/2451...rected-29y.PNG

Some see a bit of a cycle emerging here, with lows in the late 1980s and highs in the early 2000's. of course in a cycle you can generate any trend.

More interesting still is to figure out why GHCN and NOAA keep on warming while the other three clearly break with the trend.

Attached Thumbnails

 

Gokul43201

My argument is that your post was factually incorrect in addition to being scientifically meaningless.

No, it is not, but I like how you walk into a science forum and argue that statistical analysis is the same as squinting and cherry picking, while your gross misrepresentation of the science is perfectly justified.

Sounds almost like you are addressing me.


I didn't fit trend lines for 50 year data, only for the last decade or so. For the long term trend, I used the most often cited number (it was also used in one of the anti-warming blogs that was cited recently in one of these threads). What I did use long term data for was evaluating the distribution of noise in the data.

Gokul43201

There is a high frequency mode in the analysis by Xian et al. (posted in the Global Cooling thread) that has a period of about 8 years and a magnitude of about twice that of the multi-decadal oscillation. And it has a dip near 1990 and a peak near 2000. This is probably what is reflecting the cycle you mention. There is no mention in that paper of any modeling that predicts a cycle with this frequency, but that is understandable given the state of sophistication of the science so far.

There is a large difference for the relative sizes of the anomalies near the 1998 peak (why this is, I have no idea), but probably a bigger difference is in the data from the last 2-3 years.

You could probably start a thread about just this. I haven't seen a thread that is dedicated directly to how the different datasets are generated, and I don't think I'm the only one that is interested in learning about this.

Art

Gokul as you are a statistician can you explain how you can make comparisons between 'corrected' data from the past 10 years and 'uncorrected' data from the previous 50 and then draw a conclusion as to how one relates to the other as it seems to a non-statistician like me to be like a comparison between apples and oranges. The 1951 - 1980 data set commonly used to establish the base line of 14 C includes the 'manmade' (sic) anomalous cold period of the 70's. What happens to the 50 year trend if you 'correct' the data to remove these 'outliers' ?

I would also like to know how you identify an outlier anyway. Is there a presumption that some special event caused an extraordinary variation? but if so how do you know the 'normal' variation data points aren't themselves made up of two or more extraordinary events which happen to 'almost' cancel each other out especially in such a complicated system as our climate? It is the inclusion of some 'anomalous' data points and the targeted or spurious removal of others that I referred to as cherry picking.

Andre

In a while, after having a look at the April 08 data. Only UAH (0.015) and RSS (-0.069) are in. We first need HADCRUT and GHCN/GISS

fuzzyfelt

Sorry that I haven’t been able to get back to this before now. Thanks for the mention of the peer reviewed cooling thread which I hadn’t properly taken in. I see the word ‘cooling’, taken from the link, in the OT, is debatable as are the patterns and aberrations. To debate this is beyond me, but I am very grateful for the valuable discussion, and look forward to further discussion on more data.

Andre

And we will keep the discussion alive Ff, as for instance we talked about Antarctica earlier. This would also add to that part:

http://www.ucar.edu/news/releases/2008/antarctica.jsp

Study here (abstract)

Andre

Update, full PDF available here:

http://polarmet.mps.ohio-state.edu/P...07GL032630.pdf

latecommer

It appears that those here who insist that the Earth is warming are being very selective in their choice of source material.
Of the four global temperature gathering groups, three of the four report stasis in the last 10 years. Hadley Cru (sp?) from the UK show no warming since 1998. Both satallite measuring sites show stasis to slight cooling. Only Nasa shows warming based on questionable ground based sites (mostly in the United States...hardly any in the Southern Hemisphere). If all four sites are used one would seriously have to doubt the validity of any statement that the Earth is warming at all in the last 10 years despite the increase in the perported "driver" CO2.
So is it truely scientific to cherry pick the site that agrees with ones opinion?

Count Iblis

Write up an article entitled: "Recent measurements contradict AGW" and submit it to Nature.

latecommer

I have actually done that in my field which is Geology with Paleo emphasis. It was rejected on the grounds that it didn't meet current understanding of climate forcers.
I have discovered that both Nature and Science will not accept any paper that seriously disputes the company line of AGW.