The Magic Theory
Catastrophic Anthropogenic Climate Change is the magic theory -- every bit of evidence proves it. More rain, less rain, harder rain, drought, floods, more tornadoes, fewer tornadoes, hotter weather, colder weather, more hurricanes, fewer hurricane -- they all prove the theory. It is the theory that it is impossible not to confirm. Example
It will take climate scientists many months to complete studies into whether manmade global warming made the Boulder flood more likely to occur, but the amount by which this event has exceeded past events suggests that manmade warming may have played some role by making the event worse than it otherwise would have been...
An increase in the frequency and intensity of extreme precipitation events is expected to take place even though annual precipitation amounts are projected to decrease in the Southwest. Colorado sits right along the dividing line between the areas where average annual precipitation is expected to increase, and the region that is expected to become drier as a result of climate change.
That may translate into more frequent, sharp swings between drought and flood, as has recently been the case. Last year, after all, was Colorado's second-driest on record, with the warmest spring and warmest summer on record, leading to an intense drought that is only just easing.
Generally one wants to point to a data trend to prove a theory, but look at that last paragraph. Global warming is truly unique because it can be verified by there being no trend.
I hate to make this point for the five millionth time, but here goes: It is virtually impossible (and takes far more data, by orders of magnitude, than we posses) to prove a shift in the mean of any phenomenon simply by highlighting occasional tail-of-the-distribution events. The best way to prove a mean shift is to actually, you know, track the mean. The problem is that the trend data lines for all these phenomenon -- droughts, wet weather, tornadoes, hurricanes -- show no trend, so the only tool supporters of the theory have at their disposal is to scream "global warming" as loud as they can every time there is a tail-of-the-distribution event.
Let's do some math: They claim this flood was a one in one thousand year event. That strikes me as false precision, because we have only been observing this phenomenon with any reliability for 100 years, but I will accept their figure for now. Let's say this was indeed a one in 1000 year flood that it occurred over, say, half the area of Colorado (again a generous assumption, it was actually less that that).
Colorado is about 270,000 KM^2 so half would be 135,000 KM^2. The land area of the world (we really should include oceans for this but we will give these folks every break) is about 150,000,000 km^2. That means that half of Colorado is a bit less than 1/1000 of the world land area.
Our intuition tells us that a 1 in 1000 year storm is so rare that to have one means something weird or unusual or even unnatural must be going on. But by the math above, since this storm covered 1/1000 of the land surface of the Earth, we should see one such storm on average every year somewhere in the world. This is not some "biblical" unprecedented event - it is freaking expected, somewhere, every year. Over the same area we should also see a 1 in 1000 year drought, a 1 in 1000 year temperature high, and a one in one thousand year temperature low -- every single damn year. Good news if you are a newspaper and feed off of this stuff, but bad news for anyone trying to draw conclusions about the shifts in means and averages from such events.
fearandloathing:
Please keep making the point a trillion more times if you can please. The people on the other side of the debate will not remain silent. Even after they have passed laws on absolute quackery and destroyed our economy with said quackery, they will not stop talking about it. If good men stay silent evil endures.
September 16, 2013, 6:17 ammorganovich:
a perusal of the dsm 5 would be instructive here.
there are names for people who see the same picture in every signle inkblot, and they apply in a very real sense to the agw crowd.
September 16, 2013, 7:17 amesoxlucius:
"More rain, less rain, harder rain, drought, floods, more tornadoes, fewer tornadoes, hotter weather, colder weather, more hurricanes, fewer hurricane -- they all prove the theory."
New Word: Ecochondriac
September 16, 2013, 7:35 amDon:
I find it humorous that the same people that make fun of creationists for applying religion to science don't see they are doing the exact same thing. It's all religion, and a cultish one at that.
September 16, 2013, 7:55 amdavesmith001:
This a a well stated refutation of the "one in a thousand years" event stuff.
September 16, 2013, 8:21 amMatthew Slyfield:
It's worse than that, they aren't applying religion to science, they have turned science into a religion.
September 16, 2013, 8:22 amRuss R.:
"It is virtually impossible (and takes far more data, by orders of magnitude, than we posses) to prove a shift in the mean of any phenomenon simply by highlighting occasional tail-of-the-distribution events. The best way to prove a mean shift is to actually, you know, track the mean. The problem is that the trend data lines for all these phenomenon -- droughts, wet weather, tornadoes, hurricanes -- show no trend, so the only tool supporters of the theory have at their disposal is to scream "global warming" as loud as they can every time there is a tail-of-the-distribution event."
Not quite... they acknowledge that there may be no evidence of change to the mean, so instead they allege either A) that the variance has grown... i.e. there is more deviation from the mean, or B) the skewness has changed... i.e. there is a change in the directionality of variance, obviously in whatever direction would be for the worse. Occasionally I've even heard C) the tails are getting fatter (a.k.a. the kurtosis has changed), i.e. there are now disproportionately more outliers events than expected.
The problem is that this requires even more data than simply establishing a change to the mean (a first-order calculation). Variance is a second-order calculation, skewness a third-order calculation, and kurtosis a fourth-order calculation. With each higher order, the calculation has greater uncertainty. More observations are required to calculate these with any reasonable degree of confidence.
http://en.wikipedia.org/wiki/Central_moment
September 16, 2013, 8:36 amrandom_eddie:
As sympathetic as I am to your overall message, there is a huge flaw in your specific argument. Your math only holds true if we assume that the events taking place at each subunit of the globe are independent, i.e. that the temperature and rainfall and storm intensity (and seismic activity and particulate emissions and human activity and whatever else you want to measure) in each 1/1000 cell is randomly determined every day and is uninfluenced by the randomly chosen values from the neighboring cells.
I hope it's obvious that these are not independently-chosen values.
Yes, it's true that the media is statistically ignorant and overly-sensationalist when it comes to tail-of-the-distribution events. But your refutation here goes too far.
September 16, 2013, 8:56 ammarque2:
Almost nothing is completely independent - not even.slot machines - but with weather there is some relationship but it is less the further away you get ( distance wise) from an event.
I think it is fair to say we should see a wetness event of 1000 year proportion on the planet every year.
Your point is well taken but I think Warren's analysis deserves more credit than you are giving it.
September 16, 2013, 12:55 pmdesconhecido:
As already noted, all these 1000 year events are not independent. This is the same sort of analysis (neglecting the dependence of events) which greatly contributed to the financial "collapse". Housing markets in different areas were considered to be independent when the truth was that the same forces which could cause widespread mortgage default in one area would also cause widespread mortgage default in other areas. Mortgage backed securities, swaps, etc., were much more risky than acknowledged. See, fore example, this.
Also, Michael Lewis wrote about this in "The Big Short."
September 16, 2013, 1:54 pmJoey:
Whether the events are independent or not doesn't entirely matter. It wouldn't change the expected amount of 1/1000 events from happening; it would only change their variance.
September 16, 2013, 6:19 pm