Land vs Ocean temperatures

Every month, GISS publishes the latest horror in the continuing drama of the earth going to hell, or at least becoming Gehenna, if not hell itself. Yes, soon it will be so hot on earth that we will all be able to be sun burned merely by standing in the shade. Don't invest in Coppertone, it won't do you any good in this future world, which is really a return to the past. Herodotus, five centuries before the current era, thought that Africans were black because living so close to the sun burned them. Oh well so much for that piece of pseudo-science. Let's look at the prevalent pseudo science of today.

We are going to compare land, sea and satellite data to see how badly we measure the temperature. The data for the graphs below can be found atthis location.



What I have plotted above is the monthly global anomalies from Dec 1978 to the present. "What an odd time frame," you might say. Yes, it is, but it is the time over which the satellites data has been measured. Besides that, the chart is of the proper scale to see the details.

And what interesting details we see. The land is heating quicker than the sea is. [sarcastic mode on] I guess CO2 must only work over land, or work twice as efficiently over land than over the sea. [sarcastic mode off] In reality this is a clear sign that we are not measuring the land temperature correctly, and that its more rapid rise is not entirely due to CO2 but to urban heating.

But, lets play their ridiculous game of taking short periods of measurements and then projecting ridiculously far into the future while at the same time ruling out or ignoring natural cyclicity and only allowing monotonically increasing extrapolations. The oceans, according to the time period under study, are warming at 0.11 deg C per decade. The land is warming at 0.29 deg C per decade--approximately 3 times as fast as the oceans. By the logic of the hysteriacs, by the end of this century, the oceans will have warmed by 1.1 deg C and the land by 3 degrees. This will eventually make the lands warm enough that the winds will always blow offshore and we can use the winds generated by global warming to run our ecologically friendly wind turbines to generate our electricity. See what benefits we can derive from such a simple observation? See what benefit there is from warming the earth? :-)

The second detail to pay attention to is the rapid changes in land temperature. There are months in which the anomaly rises as much as 1.25 degrees over the entire land area of the globe and then drops as much as 1.37 degrees over the entire globe. What causes that? Does the land suddenly absorb more energy than it did last year at that time? Does the sun stop shining so that the land cools rapidly? This is more likely due to the turning on and off of heaters and air conditioners stationed nearby the global thermometer system on land. By comparison the oceans rise and fall on a monthly anomaly basis by only about 1/10th of a degree.

The third detail is the standard deviation of the land (.41 deg C) and the ocean (.13 deg C). The noisiness of the land data is 3 times that of the seas. This means that for a 95% confidence in a conclusion that the land has warmed, we must see more than 1.2 deg C warming. The trend line shows that since Dec 1978 we have warmed the land by .9 deg C--this is as much as the earth is said to have warmed over the past century. Clearly this is a problem and just as clearly the land's temperature rise is driving the rise in global temperature as measured by the thermometers. And clearly the fact that the land is rising 3x faster than the seas, most of the land's temperature rise is not due to CO2.

The astute reader will say that with a rapidly rising curve the standard deviation from the average might not be the best measurement of the noise and that the standard deviation from the trendline is what should be done. I did that and it makes the comparison worse. If I subtract the trendline from the land and ocean temperature series respectively, we find that the standard deviation away from the trend line is 0.08 deg C for the oceans and 0.32 deg C for the land. There is almost 4x more noise in the land data than in the seas when viewed in this fashion. Remember that the world is said to have warmed by .84 deg C over a century, yet the data for this period of time says that a 95% confidence bound is .96 deg C when the deviation is measured from the trend line.

Now lets compare the GISS with the satellite data. In the picture below you can see that the reported anomalies are almost always higher for the thermometer record.



Now, to compare these two different measurement systems, whose anomalies are differently based, I subtract from each series, the average value of that series and then plot the two together. That is shown below and at first glance one would think this is not a bad fit.



Now, in this form I will subtract the two series to see how closely the satellite and thermometer system can measure changes in temperature over this time period.



What you can see is that there is a lot of difference in the measurment of global temperature by these two methods. While the average temperature difference is 0.1 deg C, some months vary by as much as .43 deg C, or half the amount of purported warming. Clearly we are not very good at measuring the globe's temperature, yet we are going use this data to try to save the planet.