Thursday, February 5, 2015

Homogenisation makes little difference to global average

There has been much blog chatter (reviewed here) recently about homogenisation of temperature, and adjustment in general. A few individual stations have been picked out and pored over. But homogenisation is a general effort to reduce bias prior to computing a global average, and the logical place to look for its effect is in that average.

This was something some bloggers were interested in doing back in 2010. Zeke writes on that here. His own investigation was mainly US. At that time, I started the TempLS code, and I've been using it for monthly reporting for over three years. It uses by default unadjusted GHCN land data, with ERSST for ocean. There general experience, noted back in 2010, was that it made little difference. TempLS compares well with the major indices.

In this post, I'll try to quantify that more, using the current improved graphics. I can compare directly the variants of TempLS with and without adjustment, with an active map below. The results are a little surprising, but the end effect is still small. A typical result is for TempLS mesh, where the trend 1910-2014 is 0.711°C/Cen, or 0.759°C/Cen after adjustment. But that is actually a high point of adjustment effect, and over more recent periods, adjustment actually has a cooling effect. Claims that AGW is a creature of adjustments are way off.

My previous post reviewed the analysis of station trends. They tend to be dominated by relatively short periods, even with a cut-off. It is fairly easy for a short term to produce a high trend, but it's the longer terms that contribute most to the average.
Update. I should mention a 2012 benchmarking paper by Victor Venema and many others.

The map below is a variant of the one on the latest data page. I've restricted it to global surface measures, and included the adjusted TempLS, marked as TLS_m_a (mesh) and TLS_g_a (grid). The clearest showing of the effect is with "trendback" mode. This shows the trend from arbitrary start year to present. Start year is shown on the x-axis. I've set it up showing the TempLS mesh unadjusted and adjusted, but you can make other choices. The grid comparison is interesting. For any graph, you can press on Data and it will show a new tab with the numbers. The Trendback button toggles between Trendback and timeseries modes (common anomaly base 1981-2010). Operation details are here.


Looking just at the TempLS mesh curves, from about 1960 back to 1900, the adjusted trend is higher. That is the "cooling the past", but it isn't much. Max about 0.04°C/Cen. After about 1970, the unadjusted trend is higher. If you switch to timeseries most (click Trendback), it seems that adjustment has relatively cooled the global temp in just the last few years, which affects the short term trends. But again, it isn't much. It looks a lot in trend, but short term trends are volatile. The grid version of TempLS doesn't really show this. It could be an Arctic effect.

Another way of looking at it is that the difference between adjusted and unadjusted is about the same as that between GISS and HADCRUT.

Remember that you can drag the plot, change the scale etc. Also, of course, change datasets, regress and smooth.






11 comments:

  1. Because land is only 1/4 of the total area being integrated over, it's not surprising to see that bias in land measurements are getting diluted in the global average.

    Maybe look at land only?

    I'd be very surprised if there isn't a significant effect from adjustments and especially homogenization in the surface temperature stations.

    [To be clear, I've always argued that adjustments are necessary. The issue is whether the adjustments are reducing the bias in the global mean average, and I believe that is generally true, even though it is likely that it is making the bias worse in specific stations, for which the homogenization algorithm would then be failing.]

    ReplyDelete
    Replies
    1. Carrick,
      I think dilution is a legitimate part of my point. But yes, land only might be interesting. Not all that easy - my mesh generator does the sphere, so I have to intervene to subset land triangles, with some realignment. And grid is a bit inaccurate without a land mask. But I could see the different effect on urban/rural etc.

      Delete
    2. In GHCNv3 the trend in the raw land surface temperature since 1880 is 0.6°C per century. After removal of non-climatic changes by the pairwise homogenization algorithm, this trend is 0.8°C per century.

      Lawrimore, J.H., Menne, M.J., Gleason, B.E., Williams, C.N., Wuertz, D.B., Vose, R.S. and Rennie, J. 2011: An overview of the Global Historical Climatology Network monthly mean temperature data set, version 3. J. Of Geophysical Research-Atmospheres, 116(D19121): doi:10.1029/2011jd016187.

      Delete
    3. If you want to compare data with and without removal of non-climatic effects, you should also use raw data for the Sea Surface Temperature. The SST trend is reduced after the removal of non-climatic changes. Something people mainly reading WUWT & Co. might not know.

      Delete
    4. UHI correction of urban stations leads to a reduction in trend too.

      By the way, there is a very good article by Zeke Hausfather and Matthew Menne on RealClimate.

      It does a pretty good job of looking at the magnitude of the various corrections. (I think Victor has done some of this too, but I'll leave it to him to provide the URLs if he so desires).

      Shub might want to look at Zeke & Mathew's article and see how much of his concerns are allayed by it.

      Delete
    5. Thanks, Victor. The more direct comparison might be 1901-2010 (Table 4), which is 0.70 unadjusted, 0.91 adjusted, difference of 0.21. I get 0.17 for 1900-2014, just using the 3.4 factor. By a better method I'm using for the breakdown post, it comes to 0.20.

      I agree about SST, though defining "raw" might be a challenge. But I don't know where to get the data you mention.

      Delete
    6. Zeke Hausfather has been working on this, he hopes to write something soon:

      Ironically enough, the net effect of adjustments on global temperatures (not land-only) is to reduce the trend bias, as the pre-1940 upward adjustments to SST are larger than the (mostly pre-1940s) downward adjustment of land temperatures from homogenization. I really need to get around to putting together an adjusted vs. raw land/ocean figure at some point…

      Delete
  2. Once you start breaking up the trend into land and ocean separately, the pattern of the warming clearly argues against the argument being advanced by Watts and others that AGW is just an artifact of the adjustment process associated with station moves, TOBS, UHI, etc.

    I got interested in zonal means (averages over bands of latitudes) as a way of reducing the noise, but still exploring positional dependence of the warming. In particular, I generated this a number of years ago as a counter argument:

    https://dl.dropboxusercontent.com/u/4520911/Climate/temperature_trends.jpg

    (Your software is randomly kicking out links by the way. It’s not restricted to https urls either.)

    What is seen here is that the land temperature trend increases as you go towards the poles, and peaks in the Arctic.

    I have argued that this pattern is not consistent with what you would find, were the main source of warming to be artifactual in nature.

    Anyway, this figure is nearly five years old now. It’d be interesting to see what this looks like with other methods for analyzing the data.

    By the way, isn’t it the case that there are tessellation algorithms that can handle preset boundaries?

    ReplyDelete
    Replies
    1. Carrick,
      Thinking about the linearity, I think the land issue is fairly simple. The difference between the plots shown (timeseries or trend) is the same weighted sum of the difference of data - ie the adjustments. To get that for land only, just multiply by the area ratio globe/land (~3.4). So from 1900, about 0.16°C/Cen for land only.

      I'm going to do a breakdown analysis - the trend difference due to adjustments by continent (and CONUS), urban/rural, maybe decades. I can do this without mesh tinkering.

      Thanks for the plot about zonal trends. Yes, it would be interesting to update.

      Sorry about the software problem (it's Google, not me). Iv'e always used the html href version, without problems so far. But again, Blogger seems more tolerant of people with IDs.

      Delete
  3. Could someone explain to me the anomaly that exists in the IPCC's first report in 1990 that has the mean global temperature at 15 deg and in their 2007 report it is 14 deg c ?

    ReplyDelete
    Replies
    1. Temperatures (but not anomalies) are very inhomogeneous. That means that a sample average will be very dependent on how representative the sample is of all the factors that influence temperature (eg altitude). Jones, in 1998, published a new estimate (after a lot of work). That was 14°C. With a lot of uncertainty.

      Delete