Malcolm Hill alerted me to Cohenite’s comments that are worthy of a new thread:
I’m just a middle man connecting the points first raised by John McLean and Thomas Quirk in their paper, ‘ Australian Temperature Variations – An Alternative View:’
And Bob Tisdales work with Pacific Decadal Oscillation (PDO)
I agree with Malcolm that this is a crucial issue because if there has been no temperature increase then AGW is shot to bits.
A starting point would be a graph of PDO phase shifts over the 20th Century;
There were 3 PDO’s during the 20thC; a warm and dry +ve PDO from 1905-46; a cool and wet -ve PDO from 1946-76; and another +ve PDO from 1976-2006.
A typical temperature chart of the 20th Century is as follows;
The 2 upward trends at the beginning and end of the 20th Century are typical because they are similar +ve PDO’s with similar temperature regimes; if one looks at the slope and amplitude for the temperature increase at the beginning, it is identical to the one from 1977 onwards; the only difference is that the one at the beginning of the 20th Century starts from a lower base. The reason for this is not because the temperatures were lower, but because of base period bias. HadCrut uses a base period of 1961-90. This period covers the end of the middle -ve PDO and the beginning of the 2nd +PDO; an average of the 30 years of this base will cause temperatures in the 2nd +PDO period to be anomalously higher because these temperatures will not have the impact of the cooler temperatures of the base period dragging them down as occurred in the averaging process; conversely, the temperatures in the -ve PDO from 1946 onwards will be anomalously cooler because they do not have the averaging benefit of the +ve PDO temperatures; there will be, therefore a step-up in temperature after 1977 and a step-down before 1946. The base period weighting for Hadcrut is 0.15C, which would drag the temperatures of the 2nd +ve PDO back down slightly; but the weighting doesn’t prevent the step-up at 1977 or the step-down at 1946.
What Bob Tisdale has done is to remove the base period bias; he does this by the simple method of annual variance; Tn+1-Tn over the full range of the HadCrut data; the result is this;
This shows only variance within the PDO climate; if there was a seperate anthropogenic signal based on increasing CO2 increases it would show as an increasing trend; there is no seperate upward temperature trend, so there is no CO2 caused temperature increase; a comparison between the 2 temperature histories is here;
It is interesting that Lucia has undertaken something similar, but from an opposite direction, when she removed the ENSO signal from all temperature indices in the post 2000 period;
That Lucia shows a cooling trend would tend to suggest that if there is an anthropogenic signal, it is a cooling one.