You say:

"Hmm. An unusually warm winter -- wonder what caused that."

As you well know, the temperature change in any one year is, for all practical purposes, entirely stochastic, even in the presence of a long-term trend of about .008C warming per year.

Similarly, the change in GHG emissions / is any one year is, for all practical purposes, entirely stochastic, even in the presence of a long-trend of reduction.

Why do you (implicitly, at least) take one data point as evidence of a trend, but not the other?

Jim Manzi
(Planet Gore author and non-denier of AGW)

Good point. And welcome. I've been hoping to lure more conservatives onto this site. I hope you'll stick around and that you'll both offer and receive civil (but vigorous!) arguments.

grist.org

As you well know, the temperature change in any one year is, for all practical purposes, entirely stochastic, even in the presence of a long-term trend of about .008C warming per year.

What about for practical purposes like noting the presence of a persistent trend?  Is discerning the presence of such a trend NOT a "practical purpose?"

Save the world: Reduce greenhouse gas emissions 5% annually.

Joe didn't say anything about detecting a trend. He implied that global warming caused this latest warm winter in particular. Jim's point is that that attribution makes roughly as much/little sense as attributing a tiny one-year drop in carbon emissions to Bush policies.

Of course the trends are what we care about, and temperature and carbon emissions are both trending up, but still.

grist.org

Implies 0.08 C last ten years, no?  

Implies 0.8 C the next 100 years, no again.

DR:  I may be wrong, but I believe that Manzi was trying to state that Romm's musing was specious because it postulated that the weather in a given winter showed the fingerprint of global heating.

The classic truism is that the weather on any given day is random; I think Manzi has extended that much too far by saying that the weather in any given YEAR is purely stochastic.

Perhaps the truism once held.  It appears not to be the case any more.  As the statistical process control gurus would say, we appear to have a special cause of variation at work.  The expected randomness is not found anymore.

Let's try this:  anyone who thinks the the average global temperature in any year is still purely stochastic should have no problem taking the "colder" side this bet:

I'll wager $5, payable as a donation to Grist, that the average global temperature for 2007 will be HIGHER than the average global temperature for the past decade -- a decade which, after all, includes many of the hottest years on record!  And I'll wager $10, again payable as a donation to Grist, that the average global average temperature for 2007 will be HIGHER than the average global temperature for the past 20 years.

Surely the colder side of that bet seems pretty safe if the average in any YEAR is "purely stochastic," yes?

Save the world: Reduce greenhouse gas emissions 5% annually.

Surely there are more important things to discuss than almost pure semantics. The original comment was merely a throwaway comment, an extra sentence added to the post. Yes, blaming one warm winter on global warming is as ridiculous as crediting one year of fallen emissions to the U.S. president. Can't we move on?

If I share initials with 'Global Warming', is that a sign?