Wednesday, November 23, 2016

Update check on Hansen's 1988 projections

Hansen's famous 1988 paper used runs of an early GISS GCM to forecast temperatures for the next thirty years. These forecasts are now often checked against observations. I wrote about them here. That post had an active plotter which allowed you to superimpose various observation data on Hansen's original model results.

I did an update in 2015 here, and a lot of text from there is repeated here.. I think Hansen's projections had stood up well, but they ran ahead of warming during the "pause" of around 2006-13. That pause is now over, so the interest is in whether Hansen's projection is still running warm.

I've updated to Oct 2016, or latest available. Hansen's original plot matched to GISS Ts (met stations only), and used a baseline of 1951-80. I have used that base where possible, but for the satellite measures UAH and RSS I have matched to GISS Ts (Hansen's original index) in the 1981-2010 mean. But there is also a text window where you can enter your own offset if you have some other idea.

A reminder that Hansen did his calculations subject to three scenarios, A,B,C. GCM models do not predict the future of GHG gas levels, etc - that must be supplied as input. People like to argue about what these scenarios meant, and which is to be preferred. The only test that matters is what actually occurred. And the test of that are the actual GHG concentrations that he used, relative to what we now measure. The actual numbers are in files here. Scenario A, highest emissions, has 410 ppm in 2015. Scen B has 406, and Scen C has 369.5. The differences between A and B mainly lie elsewhere - B allowed for a volcano (much like Pinatubo), and of course there are other gases, including CFC's, which were still being emitted in 1988, but not much now. Measured CO2 fell a little short of Scenarios A and B, and methane fell quite a lot short, as did CFCs. So overall, the actual scenario that unfolded was between B and C.

Remember, Hansen was not just predicting for the 2010-16 period. In fact, his GISS Ts index tracked Scenario B quite well untill 2010, then his model warmed while the Earth didn't. But then the model stabilised while lately the Earth has warmed, so once again the Scenario B projections are close. Since the projections actually cool now to 2017, surface air observation series for now are warmer than Scen B (Giss). GISS Ts corresponds to the actual air measure that his model provided. Land/ocean indices include SST, which was not the practice in 1988. Hansen himself has expressed the view that the right measure of his projection now lies between Ts and Ts+SST.

So in the graphic below, you can choose with radio buttons which indices to plot. You can enter a prior offset if you wish. It's hard to erase on a HTML canvas, so there is a clear all button to let you start again. The data is annual average; 2016 is average to date. You can check the earlier post for more detail.




20 comments:

  1. I have the smallest MacBook, and I have to reduce the text quite a bit to be able to see the radio buttons. After that it works great. Hansen's most excellent model can help the mankind dullards avoid slipping below 350 ppm.

    ReplyDelete
  2. The file with the Hanson scenario GHG concentrations was missing. When I looked at this using 2014 data, GHG forcing was just below Scenario C with methane and CFCs more than compensating for CO2. Big unknown is aerosals which could swing the results quite a bit. Overall results make the point that was discussed recently on the chaos threads: long-term model projections are less uncertain than short-term due to the steadily increasing forcing from GHG which swamps variability in the long-term.

    Chubbs

    ReplyDelete
    Replies
    1. When I looked at this using 2014 data, GHG forcing was just below Scenario C

      I don't think that can be right. Have you accounted for difference in baseline? The Hansen 1988 zero forcing year is 1958, whereas most modern forcing estimates use 1750.

      Delete
    2. Oops you are right. I misread the spreadsheet. Here is the GHG forcing that I estimated for 2014 (W/m2): scenario C - 1.52, scenario b - 2.00, 2014 actual 1.91.

      Chubbs

      Delete
  3. The GHG concentrations you quote above do not match the ones stored at Real Climate (at end of article): http://www.realclimate.org/index.php/archives/2007/05/hansens-1988-projections/?wpmp_switcher=desktop

    Reading up about these Climate Audit numbers they apparently derive from 'verbal descriptions of Hansen et al 1988'. Seem to have been removed now anyway.

    ReplyDelete
    Replies
    1. Thanks for the RC links. I have been posting too quickly today. The numbers I gave above were for CO2 only. Totals for all GHG are (relative to 1750 W/M2): Scenario B - 3.61, Scenario C - 2.74, 2014 actual - 2.82. The Real Climate links give the the net effective forcing from GHG and all other factors including aerosals. Here is how the net forcing in the scenarios compares to that used in CMIP5 (W/m2 relative to 1880): Scenario B - 2.33, Scenario C - 1.41, CMIP5 - 2.31. So in summary GHG are tracking closer to Scenario C, but total forcing including aerosals is tracking close to Scenario B. Hope this isn't too confusing

      Chubbs

      Delete
    2. Yes, there always was a mismatch. Sorry about the lapsed link; I reprinted from last year without checking.

      Delete
  4. All of a sudden the graphic works perfectly on my old ipad..

    I am wondering about "Estimated temperatures during Altithermal and Eemian times", 0.5-1 C above baseline. Altithermal is OK, but in Hansen 1981 Eemian temperatures were +1-2 C, which I believe is more "consensus".

    Also, I have found that the observational index used in Hansen 1981 and 1988, Gistemp dTs, has a near perfect fit with the CCSM4 model through history. The dTs is almost indistinguishable from the ensemble members (except for the colour).

    ReplyDelete
    Replies
    1. Do you mean that it now shows without having to clear? Yes, I find that sometimes now on PC, but not always. It never used to.

      Hansen gives no reason in the paper for the apparent reduction in Eemian, so I don't know the reason.

      dTs is a good fit there. I have seen studies which say that part of the reason for discrepancy between GCMs and indices is that it is usually air temperatures from GCMs that are compared with land/ocean. I think Hansen thought that the proper comparison is something between because dTs does not have full ocean coverage, and so is still partly a land temperature.

      Delete
    2. Yes, it shows with background chart and everything on the Ipad. Yesterday it only worked on my PC after "clear all".

      Gistemp dTs runs a little hotter than CMIP5 average, but CCSM4 also does..

      I have have been wondering, maybe the best way to make a global observational 2m dataset, is to actually use SST as a proxy for SAT, and not as a direct replacement.
      This could be done, for instance, with Cowtan's hybrid method, where the SST field gets "anchored" in gridcells with co-located SST and met station measurements (coastal and island stations).
      However, I don't have the skills to do this in practice..

      Delete
    3. I've always thought Hansen was a ~2ºC Eemian guy, though maybe there's a distinction between peak Eemian warmth and Eemian average warmth. What's shown on the plot may be close to current wider beliefs about the Eemian though. AR5 Chapter 5 made this summary statement:

      'New temperature reconstructions and simulations of the warmest millennia of the last interglacial period (129,000 to 116,000
      years ago) show with medium confidence that global mean annual surface temperatures were never more than 2°C higher than pre-industrial.'

      So, even at absolute peak warmth the belief is that 2ºC is a hard upper limit. The chapter states:

      'From data synthesis, the LIG global mean annual surface temperature is estimated to be ~1°C to 2°C warmer than pre-industrial (medium confidence) (Turney and Jones, 2010; Otto-Bliesner et al., 2013)'

      But notes that 'proxy reconstructions may overestimate the global temperature change' due to biases in the proxy network.

      My reading of the section suggests general belief in probable global average Eemian warmth of about 1ºC +/-0.5 above pre-industrial, though I'm not sure it would be strong enough for the label "consensus".

      Delete
    4. Regarding the SST/SAT issue, things are bit complex with the current GISS L/O product after implementing ERSSTv4. ERSSTv4 uses ship-based air temperature data (NMAT) for bias adjustment between different types of ship-based SST observations. Theoretically that means SST observations are nudged towards SAT measurements over time and ERSSTv4 is essentially an ocean SAT product. Along with the GISS practice of extrapolating land SAT temps over sea ice areas, it's perhaps reasonable to view current GISS L/O as being fairly close to a global SAT product.

      There's another wrinkle though, which really complicates matters, because these bias adjustments are only applied to ship-based measurements, whereas SST observations have become increasingly dominated by buoys over the past couple of decades. So, over the past 30 years there's essentially been something of a transition from an SAT to an SST record.

      Delete
    5. "ERSSTv4 is essentially an ocean SAT product"
      I don't think that is really true. NMAT is used to make relative adjustments - ie to adjudicate when there is uncertainty and you have coincident SST and NMAT. But I don't think it is meant to shift mean SST. And the test for that is buoy readings, which as you say are unaffected. The recent ERSST V4 adjustment happened because of a 0.12°C difference between buoy and ship. That doesn't suggest ship SST had been moved far in the direction of SAT.

      Delete
    6. It's not the intention but is a probable side-effect of the adjustment method, which is noted in the write-up (Huang et al. 2015): 'The slight tendency (less than 0.08C century) of NMAT-SST indicates that NMAT increases faster than SST; bias adjustment may be slightly underestimated in the early period; and therefore the global averaged SST trend may have been slightly overestimated in ERSST.v4'

      Delete
    7. Does the ocean SAT increase faster than SST?
      With assistance by KNMI climate explorer, the CMIP5 average trend for ocean SAT 60N-60S, 1901-2000, is 0.44 C/century, and corresponding trend for SST is 0.36 C/ century, a difference of 0.08 C/century.

      The same exercise with ERSST4 gives 0.67 C/ century, and for HadSST3 0.58 C/century.
      Finally, corresponding ocean SAT trend by HadNMAT2 is 0.67 C/century, whereas Gistemp dTs masked to ocean 60N-60S has the trend 0.71 C/century.

      The difference between the Hadley center products is 0.09 C/ century, quit close to that of the models.
      It looks like the trend of ERSST4 is too similar to that of HadNMAT2, at least in the investigated area and period, thus supporting Paul S suggestions..

      Delete
  5. Hi Nick,

    Do you know what dataset the original 'observed' line on the graph comes from? None of the other datasets seem to match it, though some are closer that others.
    (The link to the original paper seems to be not working)

    ReplyDelete
    Replies
    1. Peter,
      Here is another link to the original. The index is from Hansen and Lebedeff, 1987. This subsequently became the GISS met-stations only index GISS dTs.

      The fit of that is, I think, very good, as shown here. The discrepancy is partly due to inclusion of new stations; GHCN did not exist in 1988. Part may be due to the famous adjustments, but this comparison to 1988 data shows that any net change made by adjustments is very small.

      Delete
    2. Thanks for that. That link does work.

      Delete
  6. It works!! Just a little mess across the right border: "recent comments" and "blogroll" truncated left AND rest still overlapping. But not fatal it seems. Is it weirdness of Chrome's rendering machine? I can't imagine this being your bug. (Lord give us back the horizontal scrollbar and Google programmers who appreciate logic and completing a project.)

    ReplyDelete
    Replies
    1. It's probably me. I give these gadgets high priority with screen space - they will overwrite side bars etc if the screen is too narrow.I don't think it was what the Blogger designers really wanted.

      Delete