I wrote a post over a year ago about climate change where I gave a brief ontology of the climate change debate and my place in it. This post explains my position as a climate change slight-luke-warmer. I think the human release of carbon dioxide is leading to warming in the atmosphere. Still, I think the negative feedbacks must necessarily be stronger than the positive feedbacks, and so whatever the equilibrium temperature response is to a doubling of CO2 without feedback (estimates of 1-2ºC), the real-world response will be smaller.
We already have several examples of runaway positive feedbacks between CO2 and temperature
The lock-step ramp in temperatures and CO2 levels that lead to the Eemian and Holocene interglacials clearly derived from positive feedback between temperature and CO2 that got triggered by the Milankovitch cycles. The solubility of CO2 is less in warm water, so when water warms, it outgasses CO2, which increases radiative forcing of the atmosphere, which warms the water, etc. Note that the current CAGW scare is not predicated on positive feedback between CO2 and temperature, but rather between other [positive] feedbacks and temperature triggered by our release of carbon. Any theory for why our release of carbon is going to lead to runaway warming must explain why the more direct feedback between CO2 and temperatures capped out at the onset of the Eemian and the Holocene and did not continue leading to more warming and why those constraints no longer hold. Or if we think such a theory is doomed from the start, we can use our brains for other purposes. For instance, what happens on earth when the temperature reaches the levels of the Eemian and the Holocene? The Arctic Ocean melts in the summer. Pope’s climate theory is probably barking up the right tree. Incidentally, the ~800 year lag between CO2 and temperature that us “deniers” like calling attention to may be a result of the ~1000 year thermohaline circulation time.
The current CAGW scare is not restricted to CO2 processes
Because alarmism is not predicated on a positive feedback loop between CO2 and temperature but rather between other quantities and temperature, the possibility of run-away warming would be triggered by any spontaneous increase in global temperatures on the scale of feedback-less CO2 forcing (1-2ºC). This has not happened in 10,000 years. The conclusion is that there is no possibility of run-away warming between non-CO2 quantities and temperature.
CAGW is literally based on linear extrapolation of trends 1970-2000
Wattsupwiththat provides a very nice graphic of this:
Figure 1. Temperature series with IPCC and harmonic models.
Making a harmonic extrapolation may seem naïve, in that it means you don’t really know anything about the system except that it’s likely to continue doing in the future what it did in the past, but if it is naïve, certainly a linear extrapolation takes the cake because it’s worse in every respect. Well, a constant extrapolation would be worse, but that brings me to my next point.
The most naïve model possible performs better than alarmist models
The only thing more naïve than an extrapolation of the first derivative of a series (the extrapolation of a linear fit) is the extrapolation of the zeroth derivative (a constant value). In other words, if I’m a modeler in the late 90s, the least intellectual thing I could possibly do is extrapolate the current global temperature into perpetuity. Currently, this model is doing better than CMIPanything. If your fancy schmancy models can’t even beat the nullest of the null hypotheses, then maybe you don’t know what you’re talking about.
Meteorologists have been testing their models for far longer and they still can’t predict anything farther out than a dozen cycles or so of the primary oscillator
In meteorology, the primary oscillator is the day. In climate, the primary oscillator is the year. Meteorologists have tested their models thousands of times over the past few decades. Even with these thousand iterations, they still can’t predict farther than two weeks or so. Climatologists have tested their models (i.e. comparing them to the actual temperatures) less than once. Many climate scientists give every indication that they don’t think testing them is a even a worthwhile enterprise. We are still many decades, if not centuries, away from being able to predict climate evolution with any kind of certainty, if the comparison with meteorology is apt, and I think it is. Many climate scientists might disagree, but I think that’s because many of them can’t tell the difference between natural variability and noise.
It seems like the climate field often equivocates between natural variability and noise
Noise is random variation about a signal. Natural variation is an explanation for why temperature trends aren’t perfectly smooth. Therefore, natural variation is noise. /s
Natural variability, rather than being noise, is a highly structured signal in and of itself. Actually that’s wrong. Natural variability is many highly structured signals at many temporal frequencies with highly structured causes and dynamics. Bulldozing over them, aside from blatantly missing the point of climatology (they are >50% of the very point of climatology), makes any model you might make completely useless. Add in chaos, and I’ll want to fire you myself.
The models don’t account for correlations in the parameters
[WARNING: I may be talking out my a** on this point. Maybe some models do take these concerns into account.] Say we know that the natural range of some parameter A that gets plugged into a climate model is 3.2 to 7.8. Say we know that another parameter B has a range of 0.012 to 0.029. The climate model will than have as many runs as it needs to do a full factorial exploration of these parameters. However, say A and B are anticorrelated in nature. Then it will be that lower range of A and of B or the higher range of A and of B cannot both occur simultaneously in nature due to unknown causes. Consequently, any effect that these models have in the final averaging will be completely spurious. Models exploring impossible regions of the parameter space may be the ones making all the ridiculous catastrophic predictions. But climatologists wouldn’t know; they haven’t tested any of their models.
Climate science doesn’t know anything about the Arctic
In summer of 2013, I was an alarmist believer. I’m also an extreme news junkie, and so I eagerly awaited the record-smashing September 2013 arctic ice low. I was utterly underwhelmed. I started reading and learned that climatologists weren’t predicting an ice-free September until 2050 or beyond. But clearly, anyone looking at the trend would predict an ice-free September as soon as 2020. Either climatologists are completely wrong about the very late 2050 prediction, or there’s a cyclicity inherent in September minima, in which case they’re still completely wrong. I asked a Dr. Kathy Crane who did an AMA on Reddit a question along these lines, but I’m still waiting for an answer. I think Arctic ice is the canary in the coal mine for the CAGW movement, but that’ll be another post.