Thursday, September 10, 2015

TempLS up 0.07°C in August

GHCN was late this month. I think the US Labor day weekend may have been the cause. However, today a lot of new data appeared, and Canada seems to be the only major one now missing. The TempLS mesh anomaly goes from 0.607°C in July to 0.679°C in August. That is a smaller rise than NCEP/NCAR (0.14°C), and does not make August as warm as Feb/Mar of this year. But it is still warm. A 0.07 rise would bring GISS to 0.82°C. Still noting NCEP/NCAR, I think GISS may rise more, though.
The smoothed map plot is here:


Warm spots in MidEast/Sudan, Central Europe, S America, Mongolia. Cool in E US. The map does not show it well but Antarctica was cold (even more than usual). The contribution plot from the report is here:

Much as in July, with SST continuing to rise, and being the main influence. Antarctica has been cool for some months.

Update. A commenter asked, which other recent months have been a record. Here is a table of years in which each month had the max value. I have colored cells where that month is the most recent with data:

HadCRUTGISSloNOAAloTempLSgridTempLSmesh
Jan 20072007200720072007
Feb 19981998201519982015
Mar 20022010201520152010
Apr 20102010201020142010
May 20152014201520152014
Jun 20152015201520152015
Jul 20152015201520152015
Aug 20142014201420152015
Sep 20142014201420142014
Oct 20142014201420142014
Nov 20132013201320132013
Dec 20062006201420142014


The rowss Jan-Apr are relatively colorless. I think the reason is that there records tend to be at the El Nino peak, which we haven't yet reached. Maybe next year.














19 comments:

  1. El Nino is showing only in the SST up to now. The land has not warmed so much. But there is also a strange observation. The pacific has a strong El Nino, but only the central and northern parts are warm, the southern parts are cool. It looks strange. Does this happened in past El Ninos? How unusual is that asymmetry?

    ReplyDelete
    Replies
    1. I think the NH warmth is unusual; it preceded the main El Nino, but may be related to the Kelvin Wave that looked like the El Nino starting up in early 2014.

      You can see movies of various past ENSO years here. In the table on the right, go to the second selection box, choose a year, and then click the radio button next to it. Or you can go here to see stills of the SST at various dates in the last 20 years, including El Ninos.

      Here is the movie of the last year of SST; you can see the P Pacific warmth early.

      Delete
    2. To make sense of El Nino behavior, you have to really start with the standing wave mode known as the Southern Oscillation. This is a hydrodynamic activity that can't change instantaneously due to the huge inertia of the ocean's volume. What it possibly can do though, is reverse phase as it crosses a zero-point node. There is zero net inertia at this point so it can be sensitive to noise in the external forcing. Something like this possibly happened in 1980 as the 1982 El Nino had a different character than the 1998 El Nino. This doesn't happen too often but it could be happening again.

      IMO, all the talk of Kelvin waves is "just-so" stories that mixes correlation with causation. I tend to think that the standing wave will eventually be determined by a regular periodic forcing only obscured by the number of factors involves, i.e. different tidal frequencies, mixed together with the yearly solar cycle and the earth''s wobble.

      Delete
  2. You can click your way back through time and see the way El Nino and PDO episodes unfold.

    http://www.ospo.noaa.gov/Products/ocean/sst/anomaly/

    ReplyDelete
  3. FWIW, if your estimated 0.07 rise is seen in GisTemp, it will be the hottest August on record....How many months would that be the case in the last 12 months?

    ReplyDelete
    Replies
    1. Anon,
      I've added an update with a table of the years in which the hottest month to date occurred. The last 12 months are prominent.

      Delete
  4. Splendid, well at least your effort is, not so much what the data is telling us,,,,

    Thank you.

    ReplyDelete
  5. OK, I'll make a forecast, by May next year all warmest months in all surface datasets will be from the latest 12 months.
    I expect nothing from satellite data, I believe more and more that they are flawed, with a trend being 0.18C/dec lower than global troposphere radiosonde data (RATPAC) since year 2000.

    Data from Canada are in now, nothing happened with TempLSmesh. The only areas left are Sudan and Spain, hence no larger changes in anomaly is expected.
    ERSSTv4 is upp 0.02 from July and NCEP/NCAR land is upp 0.24. Thus, there is a potential for GISS Aug to rise about 0.09-0.10 C.

    ReplyDelete
    Replies
    1. I forgot to mention that ERSST Aug 2015 is the warmest monthly anomaly ever, and also the warmest absolute temperature. August is the warmest month in global SST climatology. (Data from Knmi climate explorer)

      Delete
    2. Olof: I expect nothing from satellite data, I believe more and more that they are flawed, with a trend being 0.18C/dec lower than global troposphere radiosonde data (RATPAC) since year 2000.


      Well, they measure something besides SAT, so you shouldn't expect them to match up. That said I agree there are unique issues have to be overcome before the satellite data can be used.

      The fact The fact we've seen the blips in previous ENSO events (the 1998 one really stands out with the satelltie reconstruction for example), but not in this warming blip in the satellite data tells us something meaningful. So I don't think the satellite reconstructions are so poor that you'd just miss the blip in the upper atmosphere. If they aren't seeing a blip of this magnitude, it's because it's not a feature in the 3-D temperature field

      I think one shouldn't just throw away data sets because they don't tell the story one wants to hear. Perhaps the problem is with the interpretation of the data instead.

      Delete
    3. Carrick,
      You can check here to get a detailed plot of TLT vs surface plots in the run-up to the 1997/8 El Nino, for example. the data button gives the numbers. In Sept 1997, UAH was 0.05, RSS 0.19. GISS was already up to 0.55. UAH from Jan to Apr '98 went 0.49, 0.67, 0.48, 0.74. RSS was 0.55, 0.74,0.58,0.86.

      If that event is any guide, the TLT measures should start rising Nov-Dec.

      Delete
    4. Carrick, UAH and RSS are useful for global temperature patterns and changes month by month, but are IMO not reliable for long term trends etc.
      Compared to other datasets covering the troposphere ( e.g 850-300 hPa layers from RATPAC, NCEP/NCAR, or ERA-interim) the satellites series start to diverge at the turn of the century. It looks like this: https://drive.google.com/file/d/0B_dL1shkWewaNDVmS0t1bjZjQXM/view?usp=docslist_api
      I don't know if the transition from MSU to AMSU is to blame, or something else changing at that time..

      Delete
    5. Nick, I generated a comparison of RSS to GISTEMP using woodfortrees:

      figure

      In the last decade, the relationship of warming in the surface temperature index versus lower troposphere seems quite varied.

      Note the spike in 2001 in GISTEMP that is absent in RSS

      There's a delay in warming in RSS by about 2 months at the end of 2003

      There's a spike in early 2008 in GISTEMP that's not seen in RSS. In fact there's a general cooling of RSS during that period.

      There's a prolonged spike at the end of 2010 in RSS that isn't seen in the GISTEMP.

      There's another spike at the end of 2012 in GISTEMP, not see in RSS.

      Then there's the recent anomalously long warming in GISTEMP from the start of 2014, again it does not appear to be reflected by RSS.

      So I think it's basically a crap-shoot whether we'll see significant warming in the TLT from this warming period. I think your scenario is the more likely one, so if I were a betting man, that's the one I'd pick.

      Delete
    6. Since 2011 there has been a dominance of La Nina and ENSO neutral. GISS show significant warming. UAH 5.6, presumably, and RSS don't show it. UAH version 5.5 does pick it up a bit, but version 5.6 probably would not as UAH indicated version 5.6 is more like RSS.

      a prolonged period of La Nina and ENSO neutral conditions since 2011 have confused the satellites

      Delete
    7. Olof, I did say on a previous thread that I don't think the long-term (secular) trend in the satellite data is robust.

      That said, I think it's plausible that the difference in longer period fluctuations could yield responses for near surface and at higher elevation out of phase with each other. Remember that 10-years is not particularly a long period of time compared to the timescale of natural quasi-periodic variability seen in the atmosphere.

      I wouldn’t be surprised to see the pattern you are seeing reversed at some point, with TLT warming more rapidly than the surface temperature. Or possibly even have a “step-function” like adjustment in temperature like we’ve seen in the past at higher elevations.

      Delete
    8. High frequency differences between GISS and UAH/RSS will happen due to differences in regional dominance of the global signal. In surface records land temperature changes dominate, particularly the highly sensitive extratropical Eurasian continent and Antarctica. While UAH/RSS show warming/cooling spikes for Eurasia concurrently their amplitude is much smaller - by 2 to 3 times - so influence on the global average is reduced.

      UAH/RSS variance is typically dominated by SSTs, particularly in the tropical/sub-tropical regions.

      Interestingly, the closeness of the fit for the 2004 cooling appears to be partly coincidental - In GISS the July "peak" is dominated by severe Antarctic cooling, which is barely recorded by the satellite datasets.

      Delete
  6. Olof,
    I think it is likely that by May all warmest months will be 2015 or later (surface). Probably by about May, some 2016 months could start being cooler than 2015.

    ReplyDelete
  7. Hi Nick,
    The contribution of El Nino to global average surface temperature usually peaks about three months after the peak in Nino3.4, but the fall off from the peak is slower than the rise, so it seems likely temperatures will continue to be influenced until at least the fall of 2016. If past El Ninos are any guide, the strongest influence on temperature will be between 30N and 30S, with warming in that latitude range three months after the Nino3.4 peak of about 18% of the Nino3.4 peak value. Influence north and south of 30 degrees seems to be mainly in rainfall changes, and not much in temperature.

    Steve Fitzpatrick

    ReplyDelete
    Replies
    1. Steve,
      In Southern Australia, 30-40°S, we weren't much affected directly by the 1998 El Nino, though the following decade was droughty. But the 1982/3 El Nino brought both drought and heat. So far this time, there is nothing evidently abnormal.

      Delete