Lets talk about New Zealand temperature data and trends.

Global temperature (T) datasets are organized into gridded "rectangles" defined by latitude and longitude, commonly these are 5 degrees by 5 degrees. 
A look at the NZ map shows that most of the country is covered by four grid cells bounded by 35 to 45 South and 170 to 180 East.    I sometimes refer to these as Auckland, East Cape, Hokitika and Masterton.
Much of our discussion re trends will be related to these four grid cells, sometimes only one of them.



In our press release of  2 August we said.

"It is obvious that the science cannot be settled when two important authorities cannot agree more closely on such a basic measurement as the recent New Zealand temperature trend, in a region where data must be among the best in the world," said Mr Hughes.

He said that New Zealand temperature trends as measured by the Jones et al group at the University of Norwich for the IPCC (United Nations Intergovernmental Panel on Climate Change) show warming of 0.63 degrees C over the 27 years 1979 to 2005 which is 0.38 degrees more than the seven station NIWA anomaly trend which warms at 0.25 degrees C.
"When we have the IPCC believing that New Zealand is warming at more than double the rate that NIWA publishes, then this is a sign that the science is far from settled," said Mr Hughes.
In their reply, NIWA said that we were not comparing "apples with apples". NIWA seems to be uncritically accepting that the sea around NZ has warmed compared to land by more than 0.38 deg in 27 years, a rate of over 1.4 degrees per century, about twice the rate of IPCC Global Warming (GW).  I say "..more than 0.38 degrees in 27 years..", because the marine only data of the Hadley Centre must warm at a slightly higher rate compared to the merged land/sea data that the IPCC quote.
Personally I find that proposition somewhat unlikely and I would expect land and bordering seas to be in closer concert.
But each to his own.
Here is a range of global T datsets available and the trends they generate 1979-2005 for those 4 NZ region grids making the 10 degree by 10 degree area 35 to 45 south  and 170-180 east.

Database	Trend 79-05	Comments
HadSST2	??	The marine component of HadCRUT2 and HadCRUT3 Trying to access this
HadCRUT3	0.81	Update of HadCRUT2, normally certain to be adopted by IPCC.
HadCRUT2	0.63	IPCC land/sea combined
NCDCv2	0.54	USA based SST dataset by NOAA / NCDC
Reynolds	0.45	USA based SST's developed by R Reynolds of NOAA
MSU	0.28	Satellites reading lower troposphere, bottom 8kms of atmosphere
CRUT2	0.26	IPCC land only, now superceded
CRUT3	0.25	Land only latest update just out, likely will be adopted IPCC
NIWA	0.25	NIWA's seven station NZ anomaly series
MOHMAT	0.12	UKMO Hadley Centre marine night air readings from ships, buoys etc

We see a very large range in warming trends;  0.12 deg in 27 yrs = 0.44 per century and 0.81 in 27 yrs = 3.0 deg per century.
Politicians and policymakers  say the science is settled.  We do not think so.

But hang on !! It gets worse because it looks very likely that there are errors in the Hadley Centre SST data.  So maybe some of the fruit is in fact partly rotten.

To show these errors for readers I have made two graphs of monthly anomalies differences , first of HadCRUT3 minus the Reynolds SST's and then HadCRUT2 minus Reynolds for the grid cell 40 to 45 South, 175 to 180 East.
Note the Reynolds SST data only runs from 1982 so I have spliced on NCDC data from 1979-end 1981. 
Note also that the grid cell we are talking about here, incl. Masterton, does not have any land data in CRU data, so the numbers probably refact just SST's.  We do not understand either how CRU can not find Masterton data in 20 years.

You can see the prominent spurious readings in HadCRUT3 in 2003 and also the elevated area from 2001 to 2004 where most reading are skewed to be greater than Reynolds without counterbalancing negative differences.
In the HadCRUT2 comparison it seems that for most of the 1982-2005 period, HadCRUT2 is significantly warmer than Reynolds but in that 2001 to 2004 period again there is a predominance of readings where Hadley Centre SST's are all warmer than Reynolds.

We have told Hadley Centre about these spurious readings, they say, "..you are not the first person to point out the large differences between land and sea temperatures around NZ.." and go on to say,  "We are investigating this, and other, anomalies by comparison of ship, buoy and satellite SSTs, but it will be a while before we are satisfied we understand the reasons for the difference."

Back to Review comments on  climate papers by  key "IPCC Supportive"  scientists main page
Back to warwickhughes.com
30 August 2006