Well, nobody seemed to notice my previous post on this subject, so I’ll repeat it. It is meaningless to argue about the decadal trend 1998-2007. There is no statistically significant trend during that time, and in fact, there is no ten year period in the cited record which is statistically different than zero.
So there’s nothing to discuss.
w.
Ah, that’ll teach me to post without preview. I meant, of course, that there is no ten year period in the cited record whose trend is statistically different than zero.
w.
Clearly not true, not even close, based on the facts and historical temperature anomalies even since 1 AD. The only modern day 20 year ending period to make into the top 10 was 2003 and it was 10th at +0.029C / year!
Moberg (1 AD - 1979 AD), NCDC UAH (1980 AD = 2006 AD), NH Temperature Anomalies
Year Ending, Ending Year NH Temperature Anomaly, 20 Years Prior LS Fit - Degrees C / Year
560 -0.071 0.036
561 -0.306 0.036
34 -0.001 0.034
33 -0.081 0.032
559 -0.138 0.031
35 -0.012 0.031
1560 -0.541 0.030
1491 -0.314 0.030
1561 -0.441 0.030
2003 0.388 0.029
562 -0.68 0.029
1490 -0.029 0.029
1087 0.133 0.028
32 -0.232 0.028
1088 0.078 0.028
1831 -0.422 0.027
1562 -0.393 0.027
1086 0.107 0.027
1559 -0.645 0.027
1492 -0.577 0.026
2001 0.279 0.026
2002 0.332 0.026
898 0.072 0.026
2004 0.289 0.026
1089 0.055 0.025
941 -0.055 0.025
1832 -0.469 0.025
558 -0.23 0.025
897 0.3 0.025
1563 -0.377 0.025
Example Analysis Calculation:
Year Ending 560 AD, NH Temperature Anomaly, 20 Year LS Fit Slope - Degrees C / Year
541 -0.755
542 -0.822
543 -0.846
544 -0.868
545 -0.862
546 -0.805
547 -0.682
548 -0.500
549 -0.390
550 -0.367
551 -0.373
552 -0.436
553 -0.651
554 -0.795
555 -0.587
556 -0.324
557 -0.237
558 -0.230
559 -0.138
560 -0.071
Regression Output:
Constant = -20.5043
Std Err of Y Est = 0.15516
R Squared = 0.669
Correlation = 0.818
No. of Observations = 20
Degrees of Freedom = 18
X Coefficient(s) = 0.03627
Std Err of Coef. = 0.006017
ftp://ftp.ncdc.noaa.gov/pub/data/paleo/contributions_by_author/moberg2005/nhtemp-moberg2005.txt
http://vortex.nsstc.uah.edu/data/msu/t2lt/uahncdc.lt
You hit the nail on the head. The average annual global temperature anomaly of +0.23C over the past 10 years is not significantly different than 0.00C. There should not be any discussion about unnatural global warming based on this.
However, the discussion always ends up involving the rate of change of the anomalies. The rate of change of the anomalies based on a 20 year period shows that current rates of change are less, or at least no more at best, than historical rates of change, see post above.
This added to the fact that you mentioned about the significance should be the end of the global warming saga in terms of unnatural causes based on the facts and analysis of the data. However, I am sure it will continue and it promotes interesting analysis of the data.
Global Temperature Anomalies in Perspective, 17000 BC to 2006 AD
This analysis consolidates the SH Vostok (17000 BC to 0 BC), NH Moberg ( 1 AD to 1979 AD), and the NCDC G ( 1980 AD to 2006 AD) temperature anomalies. SH - Southern Hemisphere, NH - Northern Hemisphere, G - Global, BC is BC and not BP - Before Present.
The Moberg NH anomalies were converted to Global anomalies via:
Global Anomaly = -0.00534 + 0.744429 x Northern Anomaly (r = 0.97, LS NCDC 1979 - 2006)
The Vostok SH anomalies were converted to Global anomalies via:
Global Anomaly = 0.03243 + 0.92725 x Southern Anomaly (r = 0.86, LS NCDC 1979 - 2006)
The NCDC Global anomalies are as reported.
Due of the large amount of data to display graphically, the data from 1 AD to 1990 AD is represented every 40 years, in line with much of the Vostok data. The data from 1990 to 2006 is by year.
This analysis clearly shows that during the current warming cycle, since the glacial period, the global temperature has warmed +8C to 0.00C over the period 17000 BC to 9000 BC. The global temperature anomalies from 9000 BC to 2006 AD have not been, on average, significantly different than 0.00C. The “hockey stick” is nothing but a miniscule blip in the annals of global temperature history and the current warming cycle.
Moberg - ftp://ftp.ncdc.noaa.gov/pub/data/paleo/contributions_by_author/moberg2005/nhtemp-moberg2005.txt
NCDC - http://vortex.nsstc.uah.edu/data/msu/t2lt/uahncdc.lt
Vostok - http://cdiac.esd.ornl.gov/ftp/trends/temp/vostok/vostok.1999.temp.dat
How is Chronos’s statement untrue? I would have phrased it differently, but if you select not only the highest peak in your data set as a starting point, but the highest peak for decades around, you’re bound to see a decrease due to your starting point being so high. It’s also unreasonable to expect a new, higher peak in the last ten years if 1997/1998 itself was the highest peak in the last 30 years.
What do you find if you start your time series at 1990? Or 1970?
This is going from bad to worse. First of all, I don’t think anybody believes that the proxy temperature reconstructions such as Moberg et al. are accurate or precise enough to do the sort of data analysis you are doing with them…You are taking the data way beyond what is justifiable.
Second of all, I was not arguing that there have been no other 10 year periods where temperatures have gone up at this rate. I don’t think one has the data to say with any certainty what has happened over all 10 year periods in the past. What I was originally arguing against was your attempt to lump in the modern warming with that from the ice age by pointing out:
(1) The warming out of the last ice age stopped many thousands of years ago.
(2) While the planet did warm several degrees C, that happened over several thousand years so that the average rate of warming was only about 0.2C / century, a fact that is confirmed by the plot you made. (In fact, it looks like the warming at Vostok was ~0.2 C per century during that time and, because of the polar amplification, the global rate of warming was presumably somewhat lower.)
(3) The current rate of warming, which has now persisted for about a 30 year period, is close to 0.2 C per decade.
Finally, I was never making the claim that the fact that the warming has occurred so rapidly over the last 30 years alone shows it to be unnatural. The most important evidence that the warming is not natural lies in careful detection and attribution studies. It basically involves looking at the fingerprint of the warming and the various known forcings of the climate system.
Also, we know how much radiative forcing is being produced by the elevated greenhouse gas levels and it would require significant changes in our understanding of the climate system…both in terms of the climate models…and in terms of paleoclimate records to explain away how this will not cause at least a reasonably significant warming.
Finally, noone argues that the amount of warming that has occurred to far is comparable to that which occurred at the end of the last ice age…which I suppose is comforting since that made the difference between the current climate here in Rochester, NY and a climate where we were covered by 2 mile thick ice sheets! However, the warming that is occurring has just begun (as the levels of CO2 above the pre-industrial baseline value of ~280ppm is already ~2.5 X what it was in 1970). It is estimated that another 0.5 C would occur even if the level of greenhouse gases in the atmosphere did not continue to rise because the rapid increase in these gases has put the earth out of radiative balance. And, if ~0.2 C / decade warming continues into the future, it won’t be long before we will be committed to profound changes.
Well, if we work that out, it comes to an average rate of warming of 1 C per millenium or 0.1 C per century over that period. I should also add once again that your plot exaggerates this warming somewhat because it plots the Vostok temperature, which likely has close to a 2X larger factor of warming than the global temperature. Furthermore, not all of the temperature changes that occurred at Vostok occurred synchronously at other places around the world. In fact, some are evidently regional changes or shift of heat between the two hemispheres, likely mediated by changes in ocean currents.
Okay, ClimateGuy, let’s say, for the sake of argument that you’re right, global warming is, in fact, junk science. What, exactly, do you propose that we do about it? Next, I have to ask, have you considered all of the benefits that come from reducing greenhouse gases, even if there is no global warming? Benefits such as these:[ul]
[li]Technological advancement. Dirt cheap fuel cells, for example, would last longer than batteries in things like laptops, MP3 players, etc.[/li][li]More money in the pocket of consumers. A device that uses less energy, but does the same amount of work, means I have more money in my pocket to spend on more toys.[/li][li]Lower levels of environmental pollution from things like oil spills and the like.[/li][li]Job creation. New technologies are going to create new employment opportunites for people, many of them at higher rates of pay.[/li][/ul]Even if global warming is a myth, can you think of a way to stimulate a concerted effort which will give those same benefits?
Wevets, thanks for your post. Unfortunately, your (and Chronos’s) assumption that a peak is necessarily followed by a decreasing trend is not true. Here is an imaginary ten year series with a positive trend:
YEAR TEMP
1200 4.0
1201 -4.0
1202 -2.0
1203 1.0
1204 -1.0
1205 2.0
1206 2.5
1207 2.0
1208 2.5
1209 2.5
The linear trend for this series is +0.36° per year, despite the fact that the first year (1200) has a temperature anomaly of +4.0°, far and away the largest in the decade. So starting from a peak does not guarantee a falling trend in the following decade.
Isn’t math weird?
Regarding your second point, the decade before 1998 had two record high years (1991 and 1995), while the decade following 1998 has had none …
w.
That’s all well and good, but you’re using a different data set!
I think this is a miscommunication. While in the general sense, it is possible to select or construct a data set where starting from a peak does not result in a falling trend, that is not true for the particular data set posted by ClimateGuy and jshore (see posts #1, 44 and 55.) My statement (and Chronos’ but I don’t want to presume to speak for him) was not true for any and all data sets, but it is true for the particular data set found in posts #1, 44, and 55.
To be clear, let’s look at a regression of the temperature anomalies – my back-of-the-envelope calculations give:
For regression of temperature anomalies 1990 to 2006 inclusive (relative to 1998):
About +0.09° per year
A positive regression coefficient of +0.0058
An explained sum of squares of 0.05 (how much of the linear variation in temperature anomaly is accounted for as time advances – presumably with global warming, but the mathematical technique does not demonstrate causation)
t[sub]s[/sub] = 1.03 (P is between 0.2 and 0.4 – not statistically significant, just like ClimateGuy’s result – but it still reinforces my point)
Isn’t math weird?
Given the data posted above, do the 1991 and 1995 peaks reach the level of the 1997/1998 peak? I don’t see that they do. See posts #44 or #55.
1991: +0.14 anomaly and –0.36C compared to 1998
1995: +0.13 anomaly and –0.37C compared to 1998
(from way upthread, post #36)
Then why get so hot and bothered over a 10 year period showing no global warming when you admit that it’s not statistically significant? Look at longer periods of time so that you can find significant relationships.
The major problem I have with the logic presented in the OP is that it does this:
It seems to me that a rational approach to this claim would be to examine a longer time period and start at a different year and see if the conclusion holds – which it does not appear to. It always raises a big red flag for me when the OP does not want to examine a larger data set.
But by your reckoning, the “average body temperature” of this patient is being figured from temperature readings that were only taken while the patient is in the midst of a fever. If the patient’s temperature averages 101.3 degrees F, and the last three hours the patient’s temperature has only been 100.9F, that doesn’t mean the patient is necessarily healthy.
You need a longer baseline to determine an “average” temperature. We have lifetimes of data on billions of humans to determine that 98.6F is “average” body temperature, and not 101.3F. Similarly, we have historical temperature records going back many, many decades (not just the most recent 30 years) as well as other methods of determining temperatures even farther back (ice cores, etc.).
All of the data show that the earth is in the midst of one raging global fever.
Approximately 15C (59F) versus 15.23C (59.4F with the last 10 year average NCDC anomaly) is a raging fever? I don’t think based on the variances in the global temperature anomaly that this can even be shown to be slightly significant. Maybe you have an analysis supporting your statement.
wevets, thanks for the reply. Actually, despite your claim, it is true for the particular decade referred to in the OP. The first data point (1998) is the largest one, but despite that, the trend is positive (+0.09°C/decade). I just made up an imaginary dataset because I thought I could avoid the rancor that seems to surround the real one. My bad.
Isn’t math weird?
Say what? I’m not hot and bothered, I’m advising people not to get hot and bothered about a non-significant trend. The only thing we can conclude from the decade’s data is that the trend is not significantly different from zero.
And as to why not look at a longer time period, it’s because I don’t want to hijack the thread. I agree with you that looking at longer time spans is much more productive, but you did notice the part about "The Decade, 1998 to 2007" in the thread title, didn’t you?
w.
Maybe you could present analyses quantifying these issues as they relate to global warming. The purpose of this thread was not to debate issues but to perform analysis of published data relative to the OP.
Even using the limited time scale presented by the Hadley data, the following graph is a global temperature anomaly analysis from 1850 to 2006. Note that 1911 was “cooler” relative to the 2 sigma limit than any recent years were “warmer.”
This analysis also dispels the notion that the 1998 global temperature anomaly was “unusual” relative to the OP.
1.) The analysis from 17000 BC through 2006 AD spans 190 centuries.
2.) Maybe the warmest but not statistically different than any other year and no warmer than 1911 was cooler. See above post.
3.) I have seen no analyses to date presented here that shows conclusively that abnormal global warming exists based on the the published data.
Even using the limited Hadley data, the warming in the period 1911 to 1944 was greater than the current rate of change in global temperature anomalies
Climateguy, what, exactly, are you saying (since your OP presented no question, nor a position in a debate topic):
In the current warming cycle, since the end of the glacial period, the Earth is getting no warmer than might be expected based on all of the data. This thread was started in GQ in order to promote analyses of all of the available data to show one way or another if this is true or not, not debate nor opinion.
If it is not true, then you can assume we “pesky” humans, as you said, are just riding the currents of nature and not vice versa.