Here is the HADCRUT3 global temperature record. As you can see, there were in fact some years where a temperature record held for several years. For example, the record set in 1973 (well, it was a record since the early 1940s anyway) wasn’t tied until 1980 or exceeded until 1981. Likewise, the record set in 1990 wasn’t exceeded until 1995.
However, the other thing of note is that this plot shows that the 1998 record was a particularly dramatic one. It was 0.2 C higher than ANY previous year (and, if you exclude the previous year of 1997, which was also influenced by the El Nino, it was ~0.27 C higher than anything before 1997)! This is a very large difference compared to most of the other temperature records. For example, the 1990 record was only 0.075 C higher than the previous record. So, it is not surprising that the 1998 record will hold for several years before being broken. And, while the HADCRUT3 data set shows 1998 as still the warmest year, the last 6 years occupy the 2nd through 7th places!
Choice B. It is highly unlikely that our knowledge of conditions 50,000 years ago is as good as our knowledge of today’s conditions.
I merely used Hansens numbers, and included a conservative uncertainty (±1°C) for the temperature. For example, if the error in the 6.6 W/m2 of forcing is 1.5W/m2, then the 95% confidence interval for the forcing is from 2.2 to 11.0 W/m2.If you don’t like Hansen’s numbers, complain to Hansen, not to me.
Absolutely not. Curiously, in this case, you’re the one that’s claiming that Hansen is wrong. I’m making no claim either way, just pointing out that your citation of Hansen does not support your earlier statement that estimates of climate sensitivity from the ice age data were in “good agreement” with modern estimates. According to Hansen, while they encompass the modern data, they are not in “good agreement” with the modern data.
intention, you can’t warp a person’s arguments around and then still attribute it to them. Hansen calculated a result of roughly 3°C ± 1°C for the climate sensitivity on doubling CO2. You got your different result by saying:
(1) Well, I don’t like 1-sigma confidence intervals. So I am going to assume that what Hansen gave were in fact 1-sigma confidence intervals and but then I am going to do the computation using 2-sigma confidence intervals (i.e., requiring 95% confidence).
(2) I don’t like the fact that Hansen didn’t give any error bars on the temperature. So, I am going to arbitarily make one up…and going to call it the 1-sigma error bar. So, then I double it to get my 2-sigma error bar. And, I am not going to worry about the fact that the lower bound I am using for the global temperature rise of 3°C from the last interglacial may be lower than anybody else would find reasonable. [I am not sure it is completely unreasonable…but it sure as hell sounds a lot lower than anything I have heard of as an estimate.]
So, your result is your result, not Hansen’s result. You may believe it more than you believe Hansen’s result but you ought not to necessarily expect us to.
[P.S. - I also don’t think that the numbers you quoted are the correct 2-sigma range on the forcing numbers. They should be 3.6 to 9.6 W/m2 (assuming for your sake that Hansen’s numbers were 1-sigma). However, as near as I can tell (and assuming that you did use ±1°C for your one-sigma error bar in temperature as I stated above), you did use these correct numbers in your calculation.]
Let’s try this. Do you have a marker handy? If not, grab one. Now, write “The '70s” on your left hand. On your forehead, write “1998”. Now, on your right shoulder, write “2006” and “The Future” on your right hand.
Stand up. Extend both arms like you are an airplane. Vroooom, vroooom! You’re flying! But now, oh no, a sea gull, you must bank to the left! So drop your left arm and raise your right arm. Got it? Now hold it right there. Your left hand is lower than your head and your right arm is higher than your head, yes? Well, that is what the data is. It’s been going up since the '70s. Even though 1998, your forehead, is way taller than most of the data, your right hand, “The Future”, is higher still.
Now put the top back on the marker, sit down, and have a zen moment.
Your statement is obviously wrong. 1998 ranks 8th in YTY anomaly changes since 1850, nothing new here and certainly there is no basis to reject 1998 as anything unusal.
Continuing with peak analysis in addition to the prior YTY anomaly change analysis. This also goes to the question raised earlier that a peak always leads to an extended lower period, which is false.
Frequency of peak years 1979 through 1998 = 9/20 - p = 0.45
Probability of no new peak years 1999 through 2006 (8 years) - p = 0.0084
It is clear with a high level of confidence, that the supposed “acceleration” in global warming has been abated. Even more so if 2007 continues to track lower than 1998 then p = 0.0046 for 9 years.
Year, Global Temperature Anomaly, Degrees C - Hadley:
What is relevant in determining how long a record is likely to hold is not how much it differs from the previous year but how much it differs from either the average of or (more simply) how high it is above the previous record. This is where 1998 stands out.
And, just to say again, we are not “rejecting” 1998. We are just not promoting it to “most favored year status” whereby we start all our analysis with this year (and just do connect the dots between years rather than any sort of running average).
Sometimes if your thesis is competely f-ed up, it is better to admit it is wrong than to defend it with more and more ludicrous arguments.
This again is a bogus calculation. What it shows is simply that 1998 was a strong anomaly. Why don’t you compute the probability that the last 6 years all rank in the top 7 warmest years of the instrumental temperature record!?!
So we had a big increase in average global temperatures in 1998, followed by a couple of smaller increases, followed by a bunch of medium-sized increases. Ergo, the Earth is still getting hotter.
Is is just me, or is that third column a red herring?
The previous record to 1998 was 1997 and that anomaly change was hardly dramatic based on history as I showed. I am sorry that you can’t objectively evaluate the data.
They are not increases, they are the anomalies relative to the 0.00C datum baseline. There has been no increase in the annual global temperature since 1998.
Ahhh…Where did you show that? In fact, your post #66 shows just the opposite! Look at the peak years since 1979 that you starred. Before 1998, the maximum increase over the previous peak year was at most 0.065. By contrast, 1998 was 0.179 higher than the previous peak year of 1997. And, in fact, since 1997 and 1998 were strongly correlated by the El Nino over that period, this rather understates the jump over previous records. 1998 was actually 0.256 above any year prior to 1997.
And, your previous post to that dealt with the difference in the anomaly from that of the previous year…which is not what is relevant.
My guess is the one who is evaluating the data in a way that contradicts the conclusions of almost all of the world’s climate scientists is the one who is failing to evaluate it objectively.
At an absolute anomaly average of about +0.20C (NCDC) versus the temperature change of +8C in the current warming cycle? We can either analyze absolute temperatures or the “acceleration.”
I typically analyze the “acceleration” issue since no one yet has been able to explain the +8C change in global temperature during the current warming cycle since the glacial period and how it relates to mankind. It is usually that the temperatures are rising faster now, acceleration, than anytime since the glacial period.
Another clue that your method of evaluating the data is bogus: Any method of evaluating the data should not rely excessively on just one data point. In particular, let us consider the method whereby we eliminate the data for one single year. (We could imagine that a freak accident destroyed the data for that year.) If we were to eliminate the year 1998 from the data, then your whole thesis would come crashing down.
By contrast, any reasonable method that relies in looking at trends by doing a running average of something of the sort would conclude that global temperatures are continuing to rise, independent of which year of data was eliminated.
That because the rise between the glacial and interglacial period that occurred about 10000-15000 years ago (which, as I noted, occurred at roughly an order of magnitude slower pace than the rate of warming over the last ~30 years) does not relate in any way to mankind. You have created a complete “straw man”.
Why don’t you show an acceleration analysis of the Vostok data using a 30 year period to show that the current acceleration is signidicantly different than historical accelerations to prove your point?
Just by glancing at the figure, it is clear that the only rival that 1998 has is the 1877-1878 years. Neither of these years individually saw a jump in the record that is quite as great as the jump in 1998…However, if you take them as a unit (i.e., look at 1878 relative to any year before 1877 in the instrumental record), you have a jump of about the same size as that of 1998 relative to any year previous to 1997. Note that it took over 60 years before the 1878 record was exceeded at all (and about 100 years before it was exceeded with regularity).
jshore, always a pleasure to hear from you. When numbers are given with errors, they are by convention assumed to be one sigma unless specified otherwise. In addition, it would be ludicrous to think that Hansen’s numbers are 2 sigma values, as that would make the ice age errors smaller in all cases than the modern errors.
Unfortunately, the IPCC numbers are in some vague form of 2 sigma, with a value below the low end value being “very unlikely” (less than 10%) and a statement that values above the high end “cannot be excluded” but there is little model agreement on the high values.
So to compare the two as you wish, we have to adjust either one or the other. I figured the IPCC values were approximately two sigma, so I adjusted Hansen’s numbers. But we can’t compare them directly as you seem to think, that’s comparing apples and oranges.
You still haven’t dealt with the problem that Hansen claims to compute the overall error, but has not included an error on the temperature. It’s not that I “don’t like the fact” that Hansen didn’t give an error estimate for the temperature – it’s that it is incorrect and misleading (intentional or otherwise) to claim to calculate the total error when you have not included all the individual errors.
I picked a number I thought was reasonable for the temperature error. You’re free to disagree, and if so, please suggest what you think might be a better value. Or you could just agree that Hansen’s numbers without an error on the temperature are not meaningful, and we could forget the whole thing
But no matter how you do it, we still end up with Hansen’s values encompassing the modern values, which does not provide anything but the most minimal support for the modern estimate, and does not add any new information or allow us to narrow the range of the modern estimate.