Home > Deconstruction > Jo Nova: The trend repeats?

Jo Nova: The trend repeats?

2010 August 19

Introduction

Over at Deltoid, a commentator by the handle of cohenite makes the following point:

The hockey stick purports to do 2 things; show the current temperature rate of increase is unprecedented and that current temperatures are unprecedented; the first is rebutted by this:
http://jonova.s3.amazonaws.com/graphs/hadley/Hadley-global-temps-1850-2010-web.jpg

The second here, Figure 3 is good:
http://dahuang.dhxy.info/ClimateChange/940.pdf

The Chinese monsoon paper is actually pretty interesting, but jackerman had the same instinct that I did: verify Jo Nova’s graph. He used “wood for trees” and came up with different figures than Nova reports. So did my first pass using CRU’s data set. But careful reading of Jo Nova’s graph (presented in a post titled The BOM & CSIRO report: It’s what they don’t say that matters ) shows she used the monthly data set from UK MET. So I thought I would give that a go …

Hadley-global-temps-1850-2010-web.jpg

So this first graph is using the same range of years as indicated in the Nova’s legend. I assume that the years begin in January and end in December.

nova-hadobs-trends-1.png

Since the trends are obviously not replicated, I thought maybe Jo had been a little loose in her labeling and maybe 1860-1880 really meant 1860-1879 (inclusive).

nova-hadobs-trends-2.png

Still no go. So I thought I would fish around a little to see if I could pump up the numbers in her favor (but using whole years). This replicates the red arrows a little better, but not the numbers for the trends.

nova-hadobs-trends-3.png

I don’t know how Nova arrived at the numbers for her trends, but they don’t seem to reflect the current HadCRUT3 monthly anomalies. Remember: Trust but Verify

Code: hadobs-trends.R

Advertisements
  1. 2010 August 19 at 5:00 am

    If you check Jo Nova’s graph you will see that the trends are credited to Phil Jones. They appear to be the figures (rounded to two places, but omitting the 1975-1998 figure in favour of the 1975-2009 figure) given by Phil Jones at http://news.bbc.co.uk/2/hi/8511670.stm , the BBC News Q&A:

    Here are the trends and significances for each period:
    Period Length Trend
    (Degrees C per decade) Significance
    1860-1880 21 0.163 Yes
    1910-1940 31 0.15 Yes
    1975-1998 24 0.166 Yes
    1975-2009 35 0.161 Yes

    B – Do you agree that from 1995 to the present there has been no statistically-significant global warming

    Yes, but only just. I also calculated the trend for the period 1995 to 2009. This trend (0.12C per decade) is positive, but not significant at the 95% significance level. The positive trend is quite close to the significance level. Achieving statistical significance in scientific terms is much more likely for longer periods, and much less likely for shorter periods.

    rb: edit to set the quote.

  2. 2010 August 19 at 5:48 am

    She used the URL supplied in the lower left hand of her chart to create a graph on to which she superimposed Dr Jones’ newspaper interview answers without calculating the trends herself? That wraps up that puzzle. I appreciate the clue-in, Peter.

  3. PolyisTCOandbanned
    2010 August 19 at 8:15 pm

    Who’s JoNova and how do I class her into the climatosphere? I mean, how stupid is she?

  4. PolyisTCOandbanned
    2010 August 19 at 8:18 pm

    And how biased?

  5. 2010 August 20 at 8:15 am

    Note that the baseline for the trend moves.

  6. 2010 August 20 at 12:29 pm

    Eli, your meaning isn’t quite clear.

    Are you referring to the idea that data sets are constantly changing? That there are adjustments to the older data as well as the current data?

    Might be an issue, but a bigger issue (in debugging the Nova HadCRU/Jones mismatch) is that Jones probably was using HadCRUT3vgl annual anomalies while I was using the data set Nova listed in the lower left hand corner. ‘spose I should run the HadCRUT3vgl numbers to close that loop.

    TCO, I don’t see Nova’s stuff unless someone else points to it, but Nova might be described as the love-child of Tim Lambert and Steven Goddard. More popular than both of them put together. Computer geek with a relatively big following but some pretty strange ideas about science which occassionally falls off the rails. Similar use of a snark/snide/mocking attitude to Watts, McIntyre, Morano.

  7. carrot eater
    2010 August 23 at 2:20 pm

    I think Eli simply meant that that it’s warmer each time. Not exactly what you’d expect if there were something purely cyclical going on, which I take it is the implication.

    The look around here has changed. This is upsetting to me. Makes me nervous.

  8. barry
    2010 August 23 at 3:40 pm

    carrot eater, I think the implication is meant to be that it’s natural warming for each period. Climatically significant periods (20+ yrs), similar rates of warming, last 30 years rate not so remarkable after all.

    Point out that each period is itself warmer than the last, and the rejoinder is ‘rebound’ from LIA, or 900 year (450 X 2) climate cycles… the underlying trend is part of a larger cycle or shift, not CO2-driven.

    Whatever it takes to knock CO2 out of the ring.

  9. 2010 August 23 at 4:40 pm

    The look around here has changed. This is upsetting to me. Makes me nervous.

    Pastels are supposed to be calming… 😉

    I like the old theme well-enough for the posts – but didn’t much like the comments fields. And thought the overall look was too dark. Just trying on a new dress…

  10. 2010 August 25 at 5:15 am

    Computer geek with a relatively big following but some pretty strange ideas about science which occassionally falls off the rails. Similar use of a snark/snide/mocking attitude to Watts,

  1. No trackbacks yet.
Comments are closed.