Tuesday, November 17, 2015

GISS October 1.04°C, record month anomaly.

GISS is out, a bit late. But it is at or above expectations. At 1.04°C, up by 0.24°C from September 0.8°C. That is well clear of the previous highest, 0.97°C in Jan 2007, which was itself something of an outlier.

Update. Needless to say, 2015 is pulling away in the progress to hottest year. Sou has the story here, with her updated chart. Looks like 2015 will be at least 0.1°C hotter than any previous year. It will be hard to gin up uncertainty about that.

The rise is almost the same as TempLS mesh, 0.235°C. Here is a plot of the last 20 years, monthly, with annual average overlaid:


Graphs are below the jump.

Here is the GISS map of anomalies



And here is the TempLS Spherical Harmonics map for comparison.



13 comments:

  1. On the "no relief in sight" front, I see November is making up some ground.

    ReplyDelete
    Replies
    1. Well above 1.0C with one day to go. Warmest November?

      Delete
    2. I think that for NCEP/NCAR will be far hotter than any month except Oct 2015. That month beat the previous record by 0.15°C, and Nov looks like being about 0.06 behind.

      Delete
  2. The Antarctic matters. JMA is only slightly up (0.02). But then again JMA had a sharp rise in August and September when Antarctic was cold. So October JMA is still a crushing record. Guess we will not see the same rise in NOAA and Hadcrut, but in spite of that big records.

    ReplyDelete
    Replies
    1. Yes, I think so too. There is a group that rose about 0.2C - NCEP/NCAR, TempLS mesh, and GISS. These weight Antarctica according to its area, by interpolation etc. They were held back in Sept (and earlier) by a very cold Antarctica. Then there are grid-based TempLS GRID and JMA. They had gone ahead recently, I believe because they were less held back by cold Antarctica, where few cells have data, and are less warmed by the change to warm there. I think HADCRUT and NOAA are like the second group - very warm in September, but with less upside. Still probably a record.

      Delete
  3. Agree on the Antarctic. This biggest source of disagreement between the records is probably coverage. Among global coverage records, the biggest source of disagreement is the Antarctic - it's a big source of uncertainty. There isn't much agreement at all, although Berkeley shows better agreement with the reanalyses than the other records, which might be telling us something.

    Note that GISS adjusted most of the rest of the year *downwards* by 0.01C this month, slightly offsetting the gains due to the hot October.

    ReplyDelete
    Replies
    1. Kevin,
      The down move means my guess of a .1 excess over 2014 is probably too high. It's currently about 0.08. I'd expect that to increase, but we'd need two more months like October to bring it to 0.1. And November is looking not quite that warm.

      Delete
  4. Well, Gistemp should rise even more when all countries are in. Brazil should have lots of red on the map (+2-4). Greenland is also missing, and the station Nord in the far Northeast has an anomaly that should be above 4 C. GHCN must have some data ingestion problems, because these countries have been available for while..
    It will be interesting to see the likely jump in TempLS when those countries arrive in GHCN. (e g Sweden and Iran are also missing, but I believe that the infill does a proper job there)

    Nick, I was wondering about the TempLS map. The three northernmost Russian stations (Krenkelja, Ostrov Vize, Fedorova) have an unadjusted average anomaly of +6.4 C in Oct (1951-1980 base). Why isn't the area in the deep red color (>4) on the map?
    After GHCN adjustment the anomaly of those three russian stations decrease to 4.1 C, but that still render some deep red on the Gistemp Map. ( data source KNMI Climate explorer)

    ReplyDelete
    Replies
    1. Olof,
      Check with this map. Unfortunately, the map key seems broken, but you can show stations, magnify with right button drag (vertically), and then shift click for details. They agree with your numbers. The reason it isn't all red is that there are a lot of SST values nearby with moderate anomaly. Thee aren't cold enough to indicate frozen, and get most of the weight.

      Delete
  5. Here are my prediction of the GISS anomaly in Oct.
    http://moyhu.blogspot.de/2015/11/temperature-records-broken-in-templs.html?showComment=1447321998636#c5776508546622872922
    The model with the most parameters got it right. It had the lowest residuals but I thought that it was overfitted. Maybe I should throw away the mean of the models and that without fitting the residual annual cycle and only use the seemingly overfitted model?!

    ReplyDelete
  6. Did you notice that TempLSmesh just went up by 0.017 to 0.911 C with the arrival of Brazil data, and a red spot appeared over the country on the map.
    If Giss had waited one day with publishing, the anomaly might have been 1.06 C. However, we will know for certain by next monthly update, when all data are in for October.

    ReplyDelete
    Replies
    1. I just saw that Greenland data also arrived, so it may also have contributed to the uptick in TempLSmesh

      Delete
  7. For Antarctica, GISS shows warm anomalies extending over sea ice while GFS is colder for Weddell and Ross seas. Both are warm on land. The area uncovered by GISS on october (as well as previous months), Marie Byrd Land, shows a warm anomaly for GFS.
    Maybe than would mean it was no so warm over Weddell and Ross but even warmer on land with Byrd ? Considering both observations, maybe that would mean Antarctica was a little bit colder that what GISS showed. Those swings in Antarctica are hard to follow...

    ReplyDelete