Slashdot Mirror
Emissions of Key Greenhouse Gas Stabilize
brian0918 writes "Multiple news sites are reporting that levels of the second most important greenhouse gas, methane, have stabilized". From Scientific American: "During the two decades of measurements, methane underwent double-digit growth as a constituent of our atmosphere, rising from 1,520 parts per billion by volume (ppbv) in 1978 to 1,767 ppbv in 1998. But the most recent measurements have revealed that methane levels are barely rising anymore — and it is unclear why." From NewScientist: "Although this is good news, it does not mean that methane levels will not rise again, and that carbon dioxide remains the 800-pound gorilla of climate change."
The real 800lb gorilla for methane is the Arctic. If the predictions are right then this is the calm before the storm. If the Arctic melts, which it is, it'll release vast amounts of methane. It's likely to dwarf all other greenhouse sources. Everyone seems to be ignoring the Arctic but all the CO2 sources combined can't compare so a melting Arctic should be our primary concern. If it's the canary then the canary isn't just dead but it has been reduced to a skeleton.
There is, of course, a vast difference between predicting weather - which is a local phenomena, with significant specificity - and predicting the climate trends - which is averaging general trends globally. Consider, for instance, that it is very hard to stand on a beach and predict the exact height and shape of the next wave and precisely where and when it will break. On the other hand predicting the approximate height and time of the next high tide is rather easier. GCMs are, indeed, currently rather poor at making predictions down to the level of day to day local weather. They have, however, been very accurate at predicting year on year global climate.
As noted above, contrary to your claim, the models have proved to be remarkably robust and accurate. They are also, contrary to popular perception in some circles, not just a big pattern matching machine that are "trained" on past data. They are models that are fed in physics. Yes, there are some tweakable parameters, as there should be in any model where there is some uncertainty. The greatest area of uncertainty in models currently is clouds, since they can be both a positive or negative feedback depending on the exact nature of their formation. Of course this problem is taken very seriously and there is a lot of study. The last IPCC report had considerable detail summarising that work. The simple reality, however, is that the models have worked pretty well, and have, in fact, made significant predictions that have since been observed.
Historical temperatures from ice-cores are determined by ratios of hydrogen or oxygen isotopes in the ice. The guts of the issue is that when combined in water the different isotopes, being different masses, fractionate out at slightly different temperatures, thus the exact isotope ratio is a function of many things, but a very signficant factor is the prevailing temperature at the time the water became vapout before precipitting out. Thus the ratio, while not an exact indication of specific temperatures (unless the many other factors are also accounted for), is a good indicator of general temperature trends over long time scales. For more detail see here. The result is that, using ice cores, we can plot temperature and carbon dioxide independently.
Furthermore, more recent temperature reconstructions (as in reconstructions of only the past 1000 years or so) rely not on ice cores but on a wide variety of sources including coral, tree rings, glaciers, and more. Usually many of these different methods are cross referenced with each other to create any single reconstruction. The results can be seen in this plot of 10 different reconstructions by different independent teams. The results, as you can see, while different, all show the same trend. If you're still uncertain, feel free to use the
Craft Beer Programming T-shirts