1996 paper finds UHI bias in Jones et al IPCC South African temperature trends

This paper by Dr Robert Balling Jr and myself examines a series of rural and small town South African temperature data and compared 30 year trends with available Jones 1994 grid box trends.
Rural vs IPCC T trends
We found clear evidence that warming trends were much higher in the Jones data which of course includes many large urban centres. The trend of our 19 station series from non-urban sites is shown red in this extract from our Fig 2.

The large difference between the red trace and the black trace and 10 year smoothing from the Jones data is very obvious.
But not obvious for the IPCC or for the paper to be cited. The large ticks on the right represent 0.5 degree C anomaly graduations. See the linked pdf file for full diagram and larger Fig 2 version where I have added the red.

This little study came about because in 1991 I wrote to the South African Weather Bureau and by chance made contact with a helpful person who mailed me back some diskettes with max and min T data from a wide range of sites for the period 1960-1990. I had limited success getting longer term data and often wondered what scientific treasures might be stored in some vault in Pretoria.

8 thoughts on “1996 paper finds UHI bias in Jones et al IPCC South African temperature trends”

  1. Yes Simone82, the Jones in my blog title is Dr P D Jones of Climate Research Unit, University of Norwich, UK.
    www.cru.uea.ac.uk/cru/pubs/byauthor/jones_pd.htm
    Here are refs to two recent works with the Dr Michael Mann you mention.
    # Jones, P.D., Mann, M.E., “Climate Over Past Millennia”, Reviews of Geophysics, 42, RG2002, doi: 10.1029/2003RG000143, 2004.
    # Mann, M.E., Jones, P.D., “Global surface temperature over the past two millennia”, Geophysical Research Letters, 30 (15), 1820, doi: 10.1029/2003GL017814, 2003.

  2. “The trend of our 19 station series from non-urban sites is shown red ”

    That trend line appears to me to be as high or even higher than the urban sites….

  3. Our Fig 3 shows the large difference in trends over the 1960 to 1990 period between our 19 station “Small town” series and the Jones data. The trend difference is 0.2 per decade Matt. Huge, thats a rate of 2 degrees per century or ~3 times “IPCC global warming. That’s the extent of IPCC errors here. See the title of my Blog, “Errors in IPCC climate science”.

    Fig 2 is showing anomalies Matt, not actual temperatures. If the ten year trend on the red line could be shown, it would show higher anomalies than the black trend in the 1960-75 period then it would cross over the black trend and show lower red anomalies in the 1975-90 period, thus exhibiting a lesser warming trend, as Fig 3 spells out.

  4. HUGE NEWS in the world of surface records: michellemalkin.com/2007/08/09/hot-news-nasa-fixes-flawed-temperature-data-1998-was-not-the-warmest-year-in-the-millenium/

    Steve McIntyre finds major mistakes in the algorithm NASA uses to calculate suface temps and results in big corrections that are almost earth shattering.

    The boys at NASA fudge the data, he figured out how, and called them on it. GOOD JOB STEVE! It proves the 1990 were NOT the warmest decade in teh last 1000 yeras, adn 1998 was NOT the warmest year. Completely contrary to Mann/IPCC claims of being “99% certain”.

    This SHOULD be huge news. But I bet it gets completely swept under the rug.

    Warwick, does this deserve it’s own entry?

  5. Turboblocke; Perhaps you should take another read. Note Fig 2 and 3, the main thrust of the paper was comparing mean temperature trends for the different databases – Jones gridded data vs our data from various station population groups. Fig 3 sums up our main finding.

Leave a Reply

Your email address will not be published.