Steve,

But the “worse than we thought” comment that they used implies just that: CO2 is a much worse problem than we even imagined it was. I was just pointing out the flaw in their logic.

Steve: “They” didn’t use ‘worse than we thought’ in this context. I used it sarcastically to comment on the methodology. And until we understand the instability in the SST estimates from parameter selection, it’s pointless speculating on its downstream impact and I don’t want to spend time or energy on such speculations at this time or on this thread.

I see little purpose in trying to convert using CO2 impact as a metric for every discussion.

What we’ve learned here is something quite different – that seemingly innocuous parameter selections in ERSST have a large impact on SSTs before 1880. Why is this? What is it exactly in the interaction of data and parameter selection and methodology that leads to this enormous and unexpected impact? And what are the lessons of this for related applications of the methodology (which include studies that we’ve spent a lot of time – Steig et al 2009, Mann et al 2008.).

OOPS, in my last post the rates of change should read 0.075°C/decade and 0.083°C/decade… Off by a factor of ten…

Jason,

You’re on the right track here… but the key is that this is non-CO2 sensitivity. If we take it at face value (and I’ll leave it in the capable statistical hands of Steve, Ryan and Jeff to show how bad an assumption that is), then there’s a huge warming that cannot be attributed to CO2.

Eyeballing figure 2, from 1890 to 1950, the anomaly went from -0.6 to -0.15, or 0.75°C/decade. From 1970-2000 it went from -0.2 to 0.05, or about a 0.83°C/decade.

So fine, say it was cooler back in the late 1800s, but please explain how something that was completely natural back then is now obviously the result of man.

‘The adjustments do not just cool the late 19th century. They also increase the amount of warming since then. More warming implies greater climate sensitivity. Thus, lower temperatures in the 19th century result in higher temperature predictions for the 21st century.’

This is what I am wrestling with (#10). Would this work if valid imply increased climate sensitivity to human emissions? If so, why would the putative extra warming be essentially prior to 1880, well before our emissions rose substantially? Wouldn’t the new work even if valid only imply a higher estimate of natural variation in the late 19th century?