2046 wrote:Quod erat demonstrandum, chief . . . after all, aren't you guys making a claim based on a specifically limited set of data? Yes, you are, because you're talking about the past 100 years. Yet 50 of those (one large gap of 40 and then the most recent 10) don't work for you.
Not really, no. The claim being made
vis-a-vis carbon as a driver is that it accounts for:
-About a third of the variation prior to the period of manmade global warming
-Significantly more of the variation of recent years
The hypothesis is that a fundamental shift has been made in where the changes in climate are coming from. Normally dominated by solar and orbital cycles, it is currently being dominated by CO2.
The element that is significant is that there is a change in the speed of variation in carbon. This is
so far supported by
the ice core data, which - while not especially close to the CO2 year data - clearly sufficient to resolve the decade-level shifts we've seen lately.
The closest you could find was not in the same order of magnitude - and that was considered a "significant" event in the record. Carbon
generally changes slowly, it's
generally a lagging indicator, and the rare short-term massive carbon emissions tend to be point sources tied to particulate emissions (volcanic eruptions). There have been
very dramatic shifts in climate in single-year events that had nothing to do with carbon (see the "Year With No Summer", for example).
You guys reduced the planet's history to the last 100 years, and I reduced it by a factor of two, and all the sudden I'm logically invalid somehow? I think not.
Not by a factor of two, precisely.
Present data is scarcely better. It pretty much all comes from the GHCN one way or another, and that's sad. In the US, for instance, only about 10% of surface monitoring stations meet the specs for a good station (i.e. errors of less than a degree Celsius). And worldwide, about half the stations got decommissioned circa 1988-1991, and the remainder showed a much higher average, by weight.
On the good side, we have satellite data, but these are calibrated and corrected off of ground sites and radiosondes, which themselves are subject to error and calibration by ground station. So they're really severely weighting things toward the ground stations, of which there are fewer and of which many are of poor quality (i.e. errors of one or more degrees Celsius).
Which is why meteorology has gone down the tubes in the past two decades?
I think not. These same instruments are being used by meteorologists, not climatologists, who have gotten a little bit better during this period. And it's
much better than the tree ring and ice core data we have for the pre-instrumental period, or the geographically limited distribution of the "pre-modern" instrument period, which is what we're comparing it to.
Considering they are claiming accuracy of tenths of a degree and factoring that against the real error potential of their input, I'd say they have pretty bad data.
Actually, if I have ten thousand sensors with independent errors of ten degrees, I can claim an accuracy on the order of a tenth a degree from their aggregate. That's basic statistics there - error goes down with square root of the number of measurements. The main cause for worry is if there are trends
within the data, i.e., urban heat island effects.
Moreover, with what they have and the amount of noise they perceive in the system, they
aren't making forecasts claiming to be accurate to tenths of a degree. They are claiming post-facto measurements aggrating to
within a couple tenths of a degree over the last century - reasonable, all things considered, since any unmeasured bias in the instrumentation over time should be less than that - but the forecasts have margins of errors of whole degrees.
I don't recall denying the possibility of a planetary greenhouse effect, merely its plausibility in regards to Earth's short-term temperature trend as claimed by AGW folks. Hell, Mars is losing its southern icecaps as of 2005 data . . . insert your own joke there about rovers-as-SUVs.
Well, here's the deal. We know straight up from the physics how much of an effect to expect for each gas. Straight-up black-body radiation terms combined with atmospheric abosrption-emission spectra is pretty much what the 19th century model I pointed out used. It makes similarly dramatic predictions.
The modern models are much more complex - but the carbon greenhouse element is pretty well understood, in and of itself. Solar irradiance is pretty well understood, in and of itself. Again, that's essentially a black-body problem, so it's easy to model. Cloud formation? Not so easy. Oceanic heat transport? Not so easy.
What we have as a modeling problem
isn't figuring out what carbon dioxide does; we know that, just like we know what causes the Earth to zip around the Sun and what causes a nuclear meltdown. It's not figuring out where it comes from; we know that, just like we know approximately how much steel has been manufactured, how much gold has been mined, to a pretty good degree from the economic records.
DING DING DING!!! My point exactly . . . have I not been saying that the whole AGW claim is based on NOW? This gives them free reign (in their mind) to ignore the past. The problem I've pointed out is that they can't do that, and that even if we allow it their claim doesn't hold water for the time period they claim is extra-special, because half of that time period fails their claim!
Half the period doesn't fail. There are other effects in play - solar irradiance, particulates - and carbon isn't supposedly causing all the variation in temperature. Simply most of it.
Because generally, large shifts in CO2 are triggered by temperature shifts. Thus, CO2 does not - normally - trigger changes, merely make them a little more dramatic.
The catastrophic thermohaline climate shifts are part of the "alarmist" models you're complaining about, though. That's why there's worry that if too much ice melts at once, it might trigger sharp cooling in Europe even as the globe
overall is warming.
Tree rings are generally recent (when reliable . . . Mann and company had to cherry-pick the cores they wanted to use and even then couldn't make them work past 1960, though they still pretended perfect reliability prior to that).
Some ice core data I'm finding does have superior resolution at least over recent millenia, though, as good as 5 year and maybe better. See
here, but note that none of the links work at present so I'm kinda guessing by title. These are the ones listed as having data over the first century CE. Other ice core data I've found (where the links are also not working now, implying a problem as of this writing) relating to Vostok cores had resolution for the past 250,000 years as low as 80 years at times, as I recall, but seldom better than that.
However, over geological time (which, of course, comprises the vast majority of Earth's history), we don't have even that kind of resolution, meaning significant events (e.g. the Industrial Age) could be completely missed.
Could be missed, yes. But the ice core record gives us a lot more data to work with.
Yes, "could be". CO2 could, in isolation or as an atmosphere all its own, function as a pretty powerful greenhouse gas. However, in the real atmosphere, where it represents a mere 0.038%, it seems to be of limited effect, even in spite of recent contrary claims.
The fact that it only accounts for a few hundred PPM is accounted for. We know quite well how much radiation of what kind CO2 will absorb and emit.
It
is a fact that we've seen less warming than we would see were CO2 levels the only things driving shifts in climate. Hansen, quite recently but not too recently (2000, to be precise) was suggesting that since aerosol and excess carbon emissions are generally tied together, the aerosols have been providing a pretty good dampening effect on the carbon.
No crap. However, the pretense that it is a danger is based on (1) ignoring climate history of what forces what, (2) considering its effect in theoretical isolation, outside the actual climate system, and (3) ignoring the temperature and CO2 history of the planet over the past few hundred million years.
Except that (1) doesn't apply - we know what CO2 does, and it
normally acts as an amplifier for warming effects - unfortunately, we've introduced a new magic source for CO2 outside of the model.
If the models are correct about where carbon comes from and goes to, then we
won't find any period of geologic history where CO2 is a primary driver of climate.
Then stop.
I'm not the one trying to isolate this decade or 1940-1980 from the rest of the data.
That statement is incorrect . . . it is claimed to be the primary driver of climate change by mainstream AGW sources. Hansen says it, the AGW apologists of RealClimate
make fun of those who deny it, the EPA's recent BS said it, and so on.
How many times is this going to take before you get what I - and they - are actually saying? Not that CO2
is always, or even
is usually, the primary driver of climate change, but
only very recently has it gained that status.
And that
in geologic past, CO2, while having a significant effect on climate, has served the role of an amplifier - since warmer temperatures release CO2 from the ocean, which warms the planet a little further. Similarly, cool temperatures lead to ice formation, which increases albedo, leading to further cooling. This is what helps the Earth find multiple short-term equilibria - ice ages, warm periods, et cetera.
This is why I agree with a general correlation over time, but not this "lockstep" of yours, because during some historical periods the correlation looks non-existent, provided you're looking at the right timeslice. Picking the 1980-2000 timeslice and saying "we're all gonna die!" is unconvincing to me.
The only slices I'm interested in looking at separately are those in which something demonstrably different is happening. The only person wanting to divorce 2000-2009 from earlier time measurements in this thread is
you.
Let me put it this way . . . if I ever wanted to lie with a graph, I would use that method. In the real world, correlation implies a temporal component, otherwise they are disconnected events. We do not say things correlate when they bear no relationship to one another in or over time.
In short, removing the temporal component to prove correlation doesn't, and picking a particular period of known correlation and ignoring periods where correlation fails doesn't either.
Not really. There are a number of identified short-term cyclic effects that do nothing but introduce noise into the relationships.
I could plot monthly data, and this mysterious 12-month cycle would strongly permeate the data. The annual carbon cycle and the annual temperature cycles would dominate the entire relationship. I could plot by year, and I have the El Nino/La Nina cycles going bonkers, plus an 11-year sunspot cycle throwing in a little extra noise. I have volcanic eruptions at odd times. I have ice freezing and melting, and seasonal movements of water vapor from one area to another.
But CO2? CO2 plays a very immediate role, one that has no relationship with the past CO2 levels. Tomorrow, the sun will rise, and it will shine down light outside my window. Some of it is absorbed and re-radiated down in the infrared spectrum; some of that is reabsorbed and re-radiated back down by carbon dioxide. If I'm living under 1000 ppm, it'll be a little warmer today than if I'm living under 100 ppm - and in particular, tomorrow
night will be a good bit warmer.
And unless I'm keeping track of the serious long-term carbon and heat reservoirs and accounting for their behavior in my model, all I'll get is a blurrier picture putting them on a timeline. I
want to put them on a timeline - but then, I want to have all those other effects in there, too. So if I'm sitting there and I want to isolate carbon as a variable, knowing it has a pretty immediate impact - I may as well desort by time.
It's actually a very useful trick when you have data with a lot of cyclic noise - take time out. Plot X to Y alone, look for trends in
that data. And here, you can
Unprecedented warming . . . e.g. a rate, not a temperature. Or are you suggesting that rapid warming of the type you believe to be now occurring has happened before? And praytell, did it have CO2 as a primary driver, as is claimed for modern times? Or was CO2 either lagging behind or perhaps even increasing along with temperature decrease, as is observed in the actual record?
Rapid temperature shifts can occur for any of a wide number of possible reasons. If the Sun were to enter a true flare period, for example, as other stars are wont to do, or a major impact or eruption occurs, or any of a number of other causes, we could
easily see even sharper temperature shifts than we see now.
It is unprecedented to have rapid warming
triggered by a CO2 increase. It is a remarkable warming within the record we have. It is certainly one with the potential for economic catastrophe and significant costs; the amount of property too close to the beach, and thus moving into erosion/storm surge/underwater areas is disproportionately high. For
According to some, but not others. One flaw in ice core and tree ring data is that it's strictly
local, and as you no doubt noticed when you were linking to D-O events, it's quite possible for the Antarctic to warm while the Arctic cools, or vice versa. Some of the more "alarmist" models have had Europe entering a new ice age while most of the rest of the planet fries. Climate change is not simply a uniform heating or cooling.
Even the sea level does not rise evenly everywhere.