“Extremely likely”?

One of the most widely leaked claims in the AR5 SPM is this, in a box on page 12:

This evidence for human influence  has grown since AR4.  It is extremely likely that human influence has been the dominant cause of  the observed warming since the mid-20th century. {10.3–10.6, 10.9}

In the text below the box, this is expressed as “It is extremely likely that more than half of the observed increase in global average surface temperature from 1951 to 2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings together.”

In IPCC-speak, “extremely likely” means at least 95% certain.  We will have to wait until Monday to see what evidence the IPCC has in chapter 10 to support these statements.

Recall that the corresponding statement in AR4 was

Most of the observed increase in global average temperatures since the mid-20th century is very likely due to the observed increase in anthropogenic greenhouse gas concentrations

with “very likely” indicating 90% confidence, so there has been an increase in confidence levels between AR4 and AR5. Note however the change of wording from greenhouse gases in AR4 to the broader term “human influence” in AR5, which might include other effects such as deforestation or soot.  I think Roger Pielke Sr will welcome this;  he has argued for many years that the IPCC focus on greenhouse gases is too narrow.

So what could justify the increase in confidence from 90% to 95%, and it what sense has the evidence for human influence grown since AR4 was published in 2007?

Here are some things that have happened since 2007 that might have changed confidence:

  • We’ve had another six years with no warming, that climate scientists failed to predict (in fact they predicted there would be a resumption of warming).
  • Climategate showed the private doubts of climate scientists, plus journal-nobbling and data-withholding.
  • Several new papers from mainstream climate scientists have acknowledged that there is an increasing inconsistency between models and observations.

Yet confidence has increased and evidence grown?

Judith Curry has a good post questioning the 95% claim.  She draws attention to an IAC recommendation  that “Chapter Lead Authors should provide a traceable account of how they arrived at their ratings for level of scientific understanding and likelihood that an outcome will occur.” I look forward to seeing the IPCC’s account of how the 95% figure was derived.

Advertisements

3 thoughts on ““Extremely likely”?

  1. We’ve had another six years with no warming, that climate scientists failed to predict (in fact they predicted there would be a resumption of warming).

    Oddly, if we just take time series with a positive trend, mere lack of trend reversa to negative values l can increase or confidence the trend is positive. Of course, the best estimate of the magnitude of the trend declines. So a thorough discussion would mention the fact that the best estimate if for reduced warming rate relative to previous predictions. Plausibly, they should suggest that the increase in confidence it is warming comes at the expense greater uncertainty about its magnitude with observational evidence pointing to lower values.

    Of course they don’t say this because the models aren’t predicting any lower rates of warming. And they go with the models even if the observations are showing much less warming.

    Climategate showed the private doubts of climate scientists, plus journal-nobbling and data-withholding.

    For the AR5, this is a non-event.

    Climategate showed the private doubts of climate scientists, plus journal-nobbling and data-withholding.

    After the deadline. Heh! So instead it’s “volcanoes, sun, ocean, natural variability, maybe wrong forcings, maybe bad models. ” 🙂

  2. One of the underlying reasons for the increase in confidence in human caused warming is possibly the increased confidence in the measurement of radiative forcing components. Comparing Figure 2.4 of the AR4 Synthesis Report with Figure SPM.5 of last weeks report I note
    – The error range for CO2, CH4 and NOX has been doubled. But confidence in CO2 measure has gone up a notch to “very high confidence” and CH4 and NOX stay the same.
    – The forcing effect of CO2 has gone down 10% and CH4 doubled.
    – The forcing effect of halocarbons has been halved, and is well outside the 2007 error range. The error range is five times large. Yet the confidence on the figures has gone up a notch to “very high confidence”.
    – The direct aerosols have completely changed and includes the possibility that they may have a positive forcing effect. Scientists still have high confidence in their results.
    It would seem that scientists, despite having massively revised their estimates of 2007, have not become a little bit more circumspect about their work.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s