Die längste Temperaturmessreihe mit Themometer weltweit gemessen:
CET - Central England Temperature record

Diese Messreihe zeigt einen Anstieg von 2/1000stel Grad pro Jahr, 2/100stel pro Jahrzehnt und 2/10tel pro Jahrhundert.

Einige Behauptungen dazu:

1. Weil die Messungen nur aus einer Gegend kommen, sagen sie überhaupt nichts aus. global ist die Temperatur doch viel anders.

Die Antwort dazu: Falls Zentral-England nicht auf einem anderen Planeten liegt, wird sich langfristig der Temperaturtrend ausmitteln, auch wenn sich die Temperatur über mehrere Jahrzehnte hin verschiebt, wie es zum Beispiel durch die Ozeanzyklen geschehen könnte. Auch mit vielen extremen Auf und Ab stellt sich doch wieder ein Trend ein, der der Globaltemperatur entspricht. Der Vorteil ist dabei die lange Zeitreihe, die die auch große Temperaturschwankungen ausmittelt.

Zum Beispiel der Temperaturanstieg um ca. 4°C von 1677 bis 2014. Das entspricht 1.1°C pro Jahrzehnt.

Der CET entspricht übrigens ziemlich genau anderen nordeuropäischen Temperaturaufzeichnungen.

2. Die Temperaturaufzeichnungen waren damals nicht so zuverlässig wie heute.

Die Antwort dazu:
Auch wenn es heute präzisere Instrumente gibt, kann man statistisch davon ausgehen, dass Messungenauigkeiten nach oben und unten recht gleichmäßig verteilt sind und daher der Trend nicht wesentlich verändert wird. Auch dass andere europäische Temperaturreihen recht ähnlich sind, spricht für die relative Zuverlässigkeit der CET.
Aber auch die Meeresbojen und die Satelliten arbeiten mit einer Messgenauighkeit von nur 1°C plus/minus.

Dass die Klimaerwärmung seit der Beginn des erhöhten CO2-Ausstoßes seit den 60er Jahren nicht dramatisch ist, im Langzeitkontext gesehen, sieht man daran, dass eine Erhöhung der Temperatur um 0,2 bis 0,3 pro Jahrzehnt nichts dramatisches ist.

Trying to get unbiased data about the last 18 years - and longer

For a layman it is not easy to get unbiased data. As we know, land surface temperature (actually air temperature about two meters above the surface) often is polluted by the heat island effect. Which means growing cities, traffic and additional heat sources add temperature which doesn't happen in the countryside.

We have already the RSS satellite data, which shows the the air temperature from 0 up to about 10,000 meters. It is referred as the (only) graph with the longest zero trend in temperature rise. One source alone seems to be not so much to get prove about a certain matter. So I thought how to find other sources.

The ocean has much (I think about 1000 times) more heat content as the atmosphere and is more stable. And the temperature is actually measured from the surface water, not from the air. Most of the ocean water is in the southern half of the globe, and not so much disturbance from industrial and from other human sources heated water is there to be expected. So I choose the HADSHSST2 data to compare it with the RSS.

http://woodfortrees.org/plot/hadsst2sh/from:1995/plot/hadsst2sh/from:1996.4/trend/plot/rss/from:1995/plot/rss/from:1996.6/trend

I was somehow surprised to see the southern sea surface temperatures with an about two months longer zero trend than the Satellite measurements.

Okay, they have stopped to give out now data for the HADSHSST2 for some months, but you can see something very interesting: The sea surface temperature is giving the lead in the zigzag game. And then the air temperature follows, added with some additional heat from solar radiation, creating higher bounces of the graph.

No I was curious: are there other graphs that have similar lengths of zero trend? And especially above land? And avoiding the heat island effect? Maybe to measure the air a bit higher above the warming cities? So I came to the RSS again, but now to the land-only graph.

http://woodfortrees.org/plot/hadsst2sh/from:1995/mean:13/plot/hadsst2sh/from:1996.4/trend/plot/rss/from:1995/mean:13/plot/rss/from:1996.6/trend/plot/rss-land/from:1995/mean:13/plot/rss-land/from:1996.6/trend

A 13 months running average was added to make the graphs more visible.

And still the same zero trend, or even a bit longer for the air above the land.

Some more things are visible:

• The air temperature above land has much higher spikes as the global satellite temperature, especially during the 1998 El-Nino.
• When sea and global temperatures are decreasing after a strong spike, The land temperatures are steadily rising, even for some years (see 2001 to 2008).

Which means:

• global temperature depends on the ocean temperature
• land surface is amplifying the temperature
• with using RSS you can get a somehow unbiased land temperature
• The southern oceans, bordering the antarctic continent, are not getting hotter.

Just being curious: What about the "improved" HADSHSST3 graph? Important changes?

And what about the northern sea surface temperatures compared to the others?

Just lets check:

http://woodfortrees.org/plot/hadsst2sh/from:1995/mean:13/plot/hadsst2sh/from:1996.4/trend/plot/hadsst3sh/from:1995/mean:13/plot/hadsst3sh/from:1996.6/trend/plot/hadsst2nh/from:1995/mean:13/plot/hadsst2nh/from:1996.6/trend

So we see: Carefully selected data have given the HADSHSST3 some "boost".

And the northern Sea actually seems to be warming a bit quicker. Ocean cycles? Biased data? The land mass of the northern hemisphere influencing the ocean?

And now a small guess: If we want to look back into former times, the HADSHSST2 could be a somehow unbiased graph for this purpose.

http://woodfortrees.org/plot/hadsst2sh/from:1978/mean:13/plot/hadsst2sh/from:1978/trend/plot/rss/mean:13/plot/rss/trend/plot/hadsst2gl/from:1978/mean:13/plot/hadsst2gl/from:1978/trend

It didn't work as expected. The southern ocean temperature rise during the complete satellite data from 1997 on was even lower. But global ocean temperature HADSST2 fitted nicely.

Which shows us also, that the recent warmer northern ocean was possibly due to an ocean cycle.

Just if my guess is right and the ocean temperature has the same way or trend than the RSS and can be compared, how has it been during the last 150 years or so?

http://woodfortrees.org/plot/hadsst2sh/mean:60/plot/hadsst2sh/trend/plot/hadsst2gl/mean:60/plot/hadsst2gl/trend

A five year running mean is added for easier viewing.

The total difference for HADSST Global and Southern hemisphere ocean trend is 0.05 °C in 165 Years. Could be worse.

The trend of global warming for this time span is 0.7 °C. Per decade this is about 0.04 °C.

Could also be worse. To reach the 2 °C goal we need centuries...