A friend showed me this video, and I thought I would share it here: "In an average year, weather related damage totals about $6 Billion (based on data from the past 40 years). So far in 2011, weather related damage has totaled $32 Billion." And it's only June/July. [ame="http://www.youtube.com/watch?v=kXYL3pF-obo"]YouTube - ‪2011: A year of violent weather‬‏[/ame] Some people may allege global warming is a myth and hurricanes and other weather just happened to occur multiple times in some years by random chance. However, insurance $$$$$$ claims can be quantified statistically. Of course, we need to take into account inflation and growth of civilization - which should both be able to be calculated. However, the growth of civilization, due to the current economic situation hasn't been very rapid over the past several years. If the average year over the past X time period was $6 billion per year, and it's suddenly skyrocketed to $32B in 6 months, I think we can all agree that is statistically significant. However, this is only one year; 2011, and one datum itself is not statistically significant. However, since this is a La Nina year, and more severe weather does occur during La Nina years, we can compare the number of large weather events and magnitude of the events to past La Nina years. Anyhow .... interesting. This video alone says nothing about global warming - which would require a comparison of frequency of major weather events, and data from a longer period of time. I don't intend this video to be in proponency or opponency of global warming - it's just an interesting observation.
Nope. The way to determine if something is statistically significant is to do statistical analysis, not do a popular opinion poll (certainly not proclaim general consensus given no data at all). It isn't like we don't have those tools available to us.
I agree with you that I should have actually used a different phrase than "statistically significant" in that sentence since this refers to one data point and not several data points being entered into a p-test, t-test, and so forth with a confidence interval. I suppose what I meant to say in that sentence is that the data point of $32B in six months is significantly greater than the statistical average of $6B per year over the past forty years. It's been a while since I took statistics in my quantitative chemical analysis class back in college, but I believe there is a test that will determine that a given data point is "statistically different" from the other data points. If you know, please refresh my memory. Thanks.
The idea of statistical significance is that the correlation is unlikely to have occurred through other variation. For that you need to not look at the mean, but the deviations from the mean. From this a confidence can be calculated. There is over 95% confidence that global temperatures are rising. The standard deviation can be calculated and compared to this years deviation, but the shape of the curves need to be known to understand the confidence. In the last 40 years there is extreme variation. Hurricane Andrew in the early 1990s comes to mind, and should have by itself caused more inflation adjusted damage. But to here we should look to experimental design to decide what the statistics mean. Was Andrew the worst Hurricane? No it wasn't but it hit Miami where it could do the most damage. The biggest winds of a hurricane at landfall recorded in the united states happened in the 1930s. This should tell you that the experimental design is flawed. Big weather disasters occur rarely, individual ones can not be determined by warming, and the damage is also determined by a random fashion. The dust storm was cool to watch, but it needs to be looked at in the perspective of the storms in the dust bowl which happened more than 40 years ago. Dollars of damage and using only the united states are also a bad way to determine this. The amount of expensive property and people being built in harms way of likely weather events greatly outpaces inflation. The big bad weather event this year is the Japanese earthquake and tsunami. The most costly to life was arguably the bhola cyclone in 1970 that may have killed up to half a million people. As far as I know there are no peer reviewed studies that show statistically significant more extreme weather events caused by global warming. The study showing this in the IPCC was done by an insurance company, was not peer reviewed, and the company has not provided the data to allow researchers to determine whether it is higher property values or harsher weather. For now it is only a disputed theory.
Have to disagree and agree with both of you to extend. First with insurance companies; all of them believe that AGW is real, it causes more extreme weather, more damage, etc. Their rates reflect it. The rates on our property increased by 166%, while the cost of property only increased by 80%. This is ~x1.5 faster then due to inflation. And this is not 1-off example, it is just an illustration of industry as whole. Austin, the idea that insurance companies will submit their research to peer review is ludicrous at best. This could open the door to very very expensive litigations, and to tell the truth it is surprising that there was a company brave enough to share their findings with IPCC, most will not, period. For very good reason too. Now with respect to current losses in US they are linked to La Nina condition over the winter, which caused extensive snowfall in NW and Rockies. THere was some research done showing that previous Ice Age and LIA had not significantly altered the El Nino/La Nina cycles.. Unfortunately studies are not thorough enough to tell if the global cooling/warming had impact on El Nino/La Nina intensity.
Austin, thanks. I recall in my memory now that if a data point exceeds the standard deviation, then it is statistically different from the other data points. Although, we frequently just used average deviation for exclusion/inclusion purposes. I posted that OP at 12:20am - I must have been tired. Now that you bring it to my attention, you are right that there are dozens and dozens of variables that need to be taken into factor. A perfect example is a category 5 hurricane that crosses no major cities versus a category 2 that goes straight over a metropolis. Our human perception of the damage from the category 2 is going to be much greater, but the NASA satellites are going to show the category 5 was a much greater natural/global storm. Instead of focusing on hurricanes, dust clouds, and other "storms" as CBS news did in this video, I think NASA has the correct approach here to measure ocean temperatures and plot them through the 1900s to present and place a best-fit line onto the data of the ocean temperatures. Cyclopathic, I think we've pretty much determined that insurance $$$$, although an entertaining data trend to observe, is not a very good way to track global warming (take, as you mention, business $$$$ effects upon the numbers). I think the 1.5x rates are because the insurance companies are taking advantage of the global warming hype, in addition to recovering the costs of these huge pay outs they've had to make in recent years. I don't think the insurance companies are specifically increasing rates based on calculated global warming (which would probably be less than a percent). I think what you say makes perfect sense that the insurance companies would never peer review. But, that won't keep academics and the U.S. Govt. from conducting their own studies to determine if the insurance company studies even make sense. El Nino and La Nina is a normal cycle; however was this La Nina in 2011 more extreme than past La Nina years? It seems that way, but I have not studied it. Do you happen to have a link for info regarding the El Nino/La Nina activity during the last Ice Age? (I am not questioning your reference, lol, I'm just curious to read about the subject myself. thanks)
IMHO insurance companies do not specifically single out GW or make assumptions about "A" part of the GW. They just calculate the risks. Part of the rate hike is that the industry in whole in the "recovery" phase, they are trying to recover losses they suffered in 90s and early 2000, but part of it that the weather became more extreme, AGW or not. Yes the assumption that companies in capitalistic society may be greedy is plausible, however their calculations are based off Lloyd risk tables, and it is not just one or some companies, all of them went through the same rate hike. It would require world wide conspiracy which very unlikely if not improbable. sorry do not remember specifically; just remember reading as a side note. As I recall they looked at geological markers, but it did not sound like the study was systematic or wide spread, they referred to some samples from Bolivia(?) Problem is that as with any indirect studies while you see the signs of something going it is hard to project the scale out of a few random samples. There could be many reasons why there will be a no impact in one area and there will be large impact in another one.
First since it was not peer reviewed, the IPCC should not have published it according to their own rules. I don't think it is brave at all putting out stories that help justify your prices, its seems like business. +1 It appears some insurance companies do using warming models to adjust their rates. This typically has to do with flooding, but is beside the point. We should not rely on non-peer reviewed data. As Rybold said, if their is really a link, I would expect some good peer reviewed papers. There have been peer reviewed papers on specific weather events like hurricanes that there is not correlation between much more extreme or frequent hurricanes and warming. NOAA has put out new items finding they do not have corelation between weather leading to the formation of tornados and warming.
austin; in ideal world yes, and I agree with you that non-reviewed data are suspect. But you do realize where they are coming from? They might have real scientists working for them, but at the end it is a business decision, and noone would risk loosing job, violating non-disclosure, getting tangled with litigation and ruining life for pure science, esp when you are not really saving the world. You remember what happened to anonymous guy who was bypassing chain of command and sending out to media all the Katrina predictions which turned to be true? He got fired.
