Friday, May 8, 2015

TempLS April cooler by 0.1°C

As foreshadowed from the daily NCEP reanalysis, April surface temperatures were down, according to TempLS, from 0.721°C to 0.623°C (as at 8/5). Data came in early, though I don't think we have all of Canada. Warm in central Russia, West N America, an odd pattern in Africa, which may change with more data. I'll just paste the report below.

Update. I notice Roy Spencer's map has the same major features as the one below.


  1. It's shaping up to be nowhere close to as bad as I thought throughout most of the month, and now May is looking to be very warm.

  2. There are still some major areas remaining, Greenland, Mexico, Sudan, South Africa, etc, but they will probably not do very much. My wild guess, GHCN-based indices only 0.04 down, went wrong. I underestimated the land cooling, it was more than -0.20 (more likely -0.30 to -0.40), so it did not help that oceans warmed by 0.04.

    In hindsight I adjust my recipe for an early guess. Take two thirds ERSST anomaly change (+0.04) and one third UAH land (-0.30), which makes -0.073 in April, and applied on GISS an April value of 0.76-0.77

  3. Replies
    1. Well, SST anomalies are truly interesting. Since 2003 they have entered a pronounced seasonal regime, driven by the NH. At the same time NH and SH also start to diverge in annually averaged temps.

  4. GISS in at .75C. Still hot.

    1. Yes, it's actually the second hottest April ever. It beats April 2014 (.73) and we have a new running 12-month record again. And 2015 so far, is 0.11 above the 2014 average.

      April was -0.10 compared with March. Here is a rank of the forecast skill of different reanalyses:
      -0.12 Nick Stokes
      -0.14 Weatherbell
      -0.18 David Appell/CCI*
      -0.25 Karsten Haustein (both his versions gave the same result)

      * David Appell uses a conversion factor of 1.95. If he had gone straight with the CCI reanalysis the value had been -0.09, I assume