How will the global temperature develop?

[Philip]

According to your best understanding, how will the global temperature develop over the remainder of the 21st century?

[Philip]

Here is my own opinion. I think the global temperature anomaly is dominated by natural cycles with a contribution from anthropogenic effects. From this, I think it is likely that the anomaly will increase at the end of the 21st century by at most 0.4 K. This is basically the argument due I think to Akasofu.

The main driver is a rebound from the LIA, due to a millennial scale natural cycle of unknown cause. Since the cycle has such a long duration, it is reasonable to approximate it linearly over the period 1900-1950, when CO2 contributions were small. Scafetta has performed a quadratic fit to 20th century temperatures, which can be adapted here by matching at 1900 and 1950. If DT[LIA] is the temperature anomaly relative to 1900 and t is the year, then:-

[1] DT[LIA](t) = 0.0044(t - 1900) K

The LIA rebound is modulated by a 60-year oscillation, evident in the HadCRU temperature record, which is thought to be associated with oceanic cycles (AMO and/or PDO). The simplest approximation for this is a sin wave of period 60 years, fitted to the observed temperatures by setting to zero in 1925 and with amplitude 0.2 K:-

[2] DT[AMO](t) = 0.2sin[2pi(t - 1925)/60] K

Since 1950, there has been an increase in atmospheric C02 concentration. To estimate this effect, let's linearly extrapolate C02 concentrations from Mauna Loa. In 1970, the measured value was about 325 ppmv and by 2000 it was 370 ppmv. This is fitted by [ 295 + 1.5*(t - 1950) ] ppmv. From this, the concentration in 1950 is 295 ppmv and by 2100 it is 520 ppmv, an increase of 76%. If we accept that the associated temperature change is related to the logarithm of C02 concentration, and the Lindzen estimate that a doubling of C02 results in a 0.5 K temperature increase then:-

[3] DT[CO2](t) = 0 K for t < 1950 ; 0.72ln[1 + 0.0051(t - 1950)] K for t > 1950

prediction2.jpg (58.38 KiB) Viewed 1143 times

The above diagram shows temperature curves starting in 1900 and continuing through to 2100 (click on the diagram to expand it). The series 1 curve is the smoothed HadCRU instrumental temperature, which therefore terminates in 2010. The series 2 curve includes the effect of the LIA rebound [1] and the series 3 curve includes the effect of LIA and oceanic modulations [1]-[2]. Although the match between series 1 and 3 is good for 1900-1950, there is a divergence up to 2000 that can be resolved by adding in the effect of increased CO2 [3], as shown in series 4.

Overall, the sum of the effects listed above (series 4) provides a good match to observed temperatures (series 1) over the course of the 20th century and the beginning of the 21st. During the remainder of the current century, it is possible that the LIA rebound will peak and decline, and it also seems likely that C02 emissions will peak as alternative energy sources become cost effective. Hence, a straightforward extrapolation of series 4 up to 2100 (as shown in the diagram) provides a reasonable upper bound for the temperature change in 2100 of 0.4 K above the value in 2010.

[Mike Davis]

Philip:
You need to subtract the effects of human induced error caused by poor siting and corrupted methods for the surface temperature records. While the 1500 year cycle is a factor there is also the longer term cycles that need to be accounted for in your equation. Because we only have satellite lower troposphere temperature records for a portion of a climate cycle we would need to wait for the major cycles to complete to have a better understanding of recent climate because it is possible that the recent records for surface temperature are primarily an artifact of measurement errors of 51 to 90% for both land and ocean temperatures.
I would look for another major step down either this century or next as I do not think the global temperature will recover the warmth that was experienced during the Medieval warm period. If the globe is lucky it will end the century without losing what it gained from the LIA so I will give a range of .5 to -1.5 at the end of the century. I will place more weight on the colder side because that is the direction long term natural patterns are headed.

[Ravenna]

According to your best understanding, how will the global temperature develop over the remainder of the 21st century?

It will go up and will go down.

[questioner]

Philip,
Your projections are the result of curve fitting mostly without reference to any physical mechanisms. You refer to a rebound from the LIA, but you admit this has no physical origin that you can point to. The same can be said for the 60 year oscillation. Without any physical basis, this a doubtful method to say the least.

The only physical effect that you allude to is CO2, and you give an estimate of 0.5C for doubling of CO2 produced by Lindzen.
That is a lowball estimate compared to the consensus of 3C which is 6 times larger.

For some reason, I "am not authorized to view any of your graphs". That is what it says at the bottom of your post. What is up with that?

[Mike Davis]

Q:
Is there or is there not a 60 year cycle apparent in the historic climate record?
I have provided links and graphs showing different cycles that can be found in historic temperature records and now you say there is no physical basis for claiming a 60 year cycle. Make up your mind because in the past you have agreed the cycles existed or did not exist depending on the point you were trying to make.

[Philip]

Your projections are the result of curve fitting mostly without reference to any physical mechanisms. You refer to a rebound from the LIA, but you admit this has no physical origin that you can point to.

I agree that the physical origins of the LIA rebound and the 60-year cycle are not well understood, but there must nonetheless still be physical origins. One of the outstanding challenges for climate science is to provide the explanation. This thread is also intended as a challenge. I know you broadly support the AGW hypothesis, but this still leaves a lot of room for differences over future temperatures, and I'm interested to know what you think is most likely. Will temperatures rise out of control in a runaway greenhouse? Or will temperature rises be limited to 3 K? What do you think will be the influence of natural and other anthropogenic effects?

The only physical effect that you allude to is CO2, and you give an estimate of 0.5C for doubling of CO2 produced by Lindzen.
That is a lowball estimate compared to the consensus of 3C which is 6 times larger.

The reason I used Lindzen's value for CO2 doubling is because it is supported by the satellite measurements of outgoing LW, and because it provides the best fit with the temperature record once cyclic effects have been factored out.

For some reason, I "am not authorized to view any of your graphs".

I'm sorry you can't see my graph, has anyone else had a problem? See if you have more luck with this alternative link, download/file.php?mode=view&id=421

[Mike Davis]

Philip:
I have no problem seeing your graph. The problem with that graph is that it is an estimate of past global temperatures based on faulty data and faulty methods. In other words a guess that best fit a desired outcome. If you have not already you should read the Harry read me file from UEA as that is about this data set. HadCRU refers to Hadley and UEA CRU. Hadley center is responsible for the sea surface temperature data which was based on faulty assumptions and little better than grape harvest times in France or price of wheat during the LIA. There were insufficient measurement locations globally before 1957 to provide a global temperature and the loss after 1990 distorted what ever results any group is claiming.

[Philip]

Mike, I agree there are many reasons to doubt the accuracy of the temperature record, but then again many of the climate datasets are disputed in one way or another. I think the fact that the CRU record shows the 60-year cycles so clearly argues that it is roughly right, even if the increases in the second half of the 20th century are exaggerated. Nonetheless, even by taking it at face value (as in the graph), it still contains no convincing reasons to accept AGW alarmism.

[Gropes]

Mike, I agree there are many reasons to doubt the accuracy of the temperature record, but then again many of the climate datasets are disputed in one way or another. I think the fact that the CRU record shows the 60-year cycles so clearly argues that it is roughly right, even if the increases in the second half of the 20th century are exaggerated. Nonetheless, even by taking it at face value (as in the graph), it still contains no convincing reasons to accept AGW alarmism.

'Even if', is this a stealthy attempt to inject doubt in to scientific understanding? Isn't that a common way FOX news manages to get it's point across. I'm sorry, but have you evidence the increases in the second half have been exagerrated?