by gary_malcolm » Fri 14 Oct 2005, 15:02:19
$this->bbcode_second_pass_quote('PenultimateManStanding', ' ')And how about the facts as revealed by the cores taken in Greenland regarding CO2 and its relation to climate change?.
IANACANAY (I am not a Climatologist and neither are you!).
However I
do have a few points to hang on the side...
1. Chaotic events are highly predictable, just not without a proper point of reference to the key points in the causality tree. That's where the problem lies... at what point in the observation of a Chaotic event can we say that a given input into the larger unknowable equation will have a predictive effect. In other words, can I show that the butterfly's flapping wing was an additive event that enabled the monsoon? That's the point where critics have problems with some models. Another event might not be so troubling... as in does a huge warm front moving into the southern states cause a likelyhood of rain when it interacts with an existing cold air mass.
Both are chaotic events with mathematically unknowable inputs (out of sheer size) however the likely output of the latter is inarguable.
2. Some scientific models of global warming are trying to show that a delta of a given input can have knowable outcomes on the larger model. So given a base line variability as in the Greenland cores can we show a variation in the expected rate of change at a given data point... or better yet an UNEXPECTED change that can be related to limited causation. A less sophisticated audience might suggest that all variance is equivalent... but rate of change is the key factor especially when considered from the standpoint of cumulative effect.
3. When discussing complex systems we simply CAN NOT delink individual components. We can talk until we're all living in the Sahara about CO2 effect and it's measurability on one output (temperature) but we are highly unlikely to have anything approaching strict 1 to 1 causality (more like 89 to 78 causality <smirk>). Simple is for simpletons. What we can do is look for expected variation based on delta input.
G