Saturday, August 14, 2010

Hottest year? 2010?

A frivolous topic, to be sure, but it will probably be discussed in coming months. Will 2010 set any kind of hottest year record? It doesn't look particularly likely, but I'm sure people will speculate. So I've produced a tracking plot. It shows the cumulative sum of anomalies for each index, for 2010, and the hottest year of each index to date. For all but GISS that's 1998; for GISS it is 2005.
Update - thanks to a commenter who noted that for NOAA also, 2005 was the hottest year. The plot is amended, with 1998 lightly dotted for GISS and NOAA.

The hottest year is shown as a dashed line. If 2010 is ahead by the end, it will be the new hottest year. Most indices lag 1998 now, but 1998 cooled at the end. GISS is ahead of its target, but 2005 didn't cool. Each index is offset by 1.5°C from the next, for visibility.

I'll update this plot regularly at the ongoing temperature site.
Update - Lucia also has a new post on record prospects.


  1. I think its going to come down to methodology.

    In trb/GSOD land only, if you use weighted latitudinal zoning but no land mask, it will likely be the hottest year. If you use simple weighted cells, it won't. This was for land stations and without a land mask (ie ... included islands).

  2. 2010 will cool at the end too. So ... UAH, RSS and HadCRUT no way.

    Maybe GISS, maybe NOAA. Probably pretty much a
    statistical tie in those too cases.

  3. Deep, I think they'll all be "statistical ties" (and the whole diagram is statistical fuzz) - that's why I think it's a frivolous topic.

    But I agree with your assessment.

  4. NOAA's warmest year is also 2005.

  5. Anon,
    I'd forgotten that. I'll re-do the plot. Since NOAA has been tracking 1998 so closely, I'll keep it in the plot as well, as a caution - if 2010 follows 1998 to the end, 2005 will catch it. Same for GISS.