```Date: Tue, 1 Jul 2008 14:45:06 -0400 Reply-To: Sigurd Hermansen Sender: "SAS(r) Discussion" From: Sigurd Hermansen Subject: Re: Cedar Rapids, Iowa, How's the clean up going? Comments: To: ajay ohri In-Reply-To: <4bc14e460807011044j3c32de25g2f99094c8049a24a@mail.gmail.com> Content-Type: text/plain; charset="us-ascii" Though I wouldn't presume to speak for all Bayesians and subjectivists, I'd say that one would have to state what we know about the relation between a 100 year flood and a 200 year flood. Specifically, how does an instance of a 100 year flood 15 years ago affect the probability of a 200 year flood during the current year? It seems likely that the concept of a x year flood really means that over the next n*x years one should expect n floods of some magnitude. It doesn't suggest anything about intervals between x year floods or between x and 2*x floods. It also assumes that the future will follow the same patterns as the past. S -----Original Message----- From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of ajay ohri Sent: Tuesday, July 01, 2008 1:44 PM To: SAS-L@LISTSERV.UGA.EDU Subject: Re: Cedar Rapids, Iowa, How's the clean up going? I wonder what are the odds of a 1 in 200 years event happening after 15 years of a 1 in 100 years event assuming the Bayes Theorem....Anyone here who can help calculate the probabilty ? On Tue, Jul 1, 2008 at 10:04 PM, Ed Heaton wrote: > Ben; > > It's interesting that the 1993 flood was considered a 100-year flood. > (I was there, filling sandbags beside the River Des Peres.) Now, only > 15 years later, we have a 200-year flood! > > Ed > > Ed Heaton > TB-286 > #4818 > > > > -----Original Message----- > From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of > ben.powell@CLA.CO.UK > Sent: Monday, June 30, 2008 5:02 AM > To: SAS-L@LISTSERV.UGA.EDU > Subject: Re: Cedar Rapids, Iowa, How's the clean up going? > > > On Fri, 27 Jun 2008 14:57:13 -0500, Paul A. Thompson > wrote: > > >I live in St Louis area. It is clear that the entire Mississippi > >value > > >system is not working. There are levees everywhere, and they are > >expensive to maintain. Yet, when the water rises, the levees have 3 > >effects, none of which are good: > > > >1) Exacerbate the overall effect of the rise. One geologist stated > >that > the > >Mississippi used to have a 4 mile wide flood basin, and now the basin > >is > 1.5 > >miles wide. When you shrink that basin to that extent, you are gonna > >see a rising river. Each levee built is a threat against all other > >levees. > Since > >1993, levees have been raised. Now, many are failing. > > > >2) Cost more when they fail. In the St Louis area in 1993, an area > >called the Chesterfield valley was under 10 ft water. Today, there > >is the biggest strip mall in the area there, as well as 20000 houses. > >These have been built in 15 years. If the levees fail, WHICH THEY > >WILL > > >EVENTUALLY, the effects of levee "belief" + basic stupidity will make > >the result a disaster of New Orleans magnitude. This area is unlikely > >to be flooded this time, since the Mississippi is 10 ft lower at St > >Louis than it was 15 years ago. However, I believe that eventually, > >we will have a big flood again. And it will be a terrible thing. > > > >3) Cause people to behave like idiots. When you build a levee, > >people act differently (see pt 2). Farmers believe that a field > >behind a levee is not going to be flooded, and plant that field and > >put expensive chemicals on > it, > >even when the waters are rising. > > > >Paul A. Thompson, Ph.D. > >Division of Biostatistics, Washington University School of Medicine > >660 > > >S. Euclid, St. Louis, MO 63110-1093 314-747-3793 > >paul@wubios.wustl.edu > > > > This is all true but the point is to make property insurable in the > short term. This almost certainly does mean that the next flood will > be worse, when it finally does happen, but if the engineering is sound > and it doesn't arrive for 100 years then 1 such flood every 100 years > is cheaper to insure than a smaller flood every 5 years. > > Rgds > ```

Back to: Top of message | Previous page | Main SAS-L page