Land Temperature Anomalies.


Figure 1: First release of GHCN (1990) compared to Jones et al. 1988.

Bookmark the permalink.

7 Responses to p2

  1. Broadlands says:

    What are the temperatures for the respective Northern and Southern Hemisphere LAND surfaces?

  2. Clive Best says:

    The early data is dominated by the Northern Hemisphere.

    • Broadlands says:

      Perhaps you misunderstood my query? If there is a global LAND value one assumes that it is derived from LAND values on both hemispheres. NOAA says that the 20th century global LAND surface is 8.5°C (47.3°F. What are the 20th century hemisphere LAND values that make up that global value? Jones et al., 1999 said they were 14.6C and 13.4C, leading to 14.0°C for the globe.

      • Clive Best says:

        The absolute average temperature on land is not actually measured. It is only estimated. If you use the measured station data then you get a systematic trend which only depends on where the particular stations are at any particular time. That is why they cheat and use anomalies instead. There is an underlying assumptions that DT is the same for all nearby stations. If they find there is a difference they correct it. Perhaps it isn’t true they need to be the same.

        • Broadlands says:

          Only estimated? But not estimated for the hemispheres? Phil Jones explained to me why it is only estimated…

          “ The 1999 paper is only the second attempt to calculate the number directly from observations. It is within about 0.5 deg C from the only other study in the 1969/70 period – by Crutcher and Meserve and Taljaard. The reason for the differences (and why the accuracy is only expected to be within +/- 0.5 deg C) is because of areas without data in the base period – see the maps in the paper. These areas include large areas of the Southern Oceans, parts of the Antarctic, the central Arctic, central Greenland etc. To get values for these areas you have to make estimates and these introduce errors. Lapse rate estimates have to be used for mountainous regions such as Tibet and Greenland. So the bottom line is that 14 C number for the globe is probably only accurate to +/- 0.5.”

          Now they have been re-estimated and reduced? If found somewhere they should add up to 8.5°C which is more than five degrees colder than the LAND and SEA value of 13.9° Plus or minus 0.5°C? And we are worrying about fractional values? Am I missing something?

        • Clive Best says:

          Of course you can calculate the average temperature but it will be biased by the distribution of stations that you use. I did this for Crutem4 stations and got the following

          The absolute values agree with Jones but the trend is purely dependent on changing location of stations.

          • Broadlands says:

            Well, I guess I must be “out to lunch”? The 20th century global LAND temperatures are five degrees colder than the global Land & Sea temperatures, and it’s because of station selection? And, the southern hemisphere is warmer than the northern hemisphere? Or, do you have the labels reversed? The LAND surfaces do show that the SH is warmer than the NH…at least they did in the 1930s.

            Sorry to have bothered you, but do appreciate your attempt to explain all this.

Leave a Reply to Broadlands Cancel reply