tag:blogger.com,1999:blog-19706647.post3060026579655168811..comments2019-10-19T02:02:28.787-04:00Comments on Hurricane & Tornado Climate: Comparing hurricane return levels using historical and geological recordsJames Elsnerhttp://www.blogger.com/profile/02586522288616135639noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-19706647.post-70808560765838095742010-01-19T02:16:38.745-05:002010-01-19T02:16:38.745-05:00Amiable post and this post helped me alot in my co...Amiable post and this post helped me alot in my college assignement. Thanks you for your information.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-19706647.post-62497651257098675332007-01-09T13:42:00.000-05:002007-01-09T13:42:00.000-05:00Anon,
In the context of hurricane climate studies,...Anon,<br />In the context of hurricane climate studies, it is perhaps best to consider the machinery of generalized linear models (glm). For example, when considering hurricane frequency, although the response is a count variable the covariates are continuous variables (e.g., ENSO as characterized by sea-surface temperature over the tropical eastern Pacific). The R project for statistical James Elsnerhttps://www.blogger.com/profile/02586522288616135639noreply@blogger.comtag:blogger.com,1999:blog-19706647.post-18674190347168077882007-01-09T12:36:00.000-05:002007-01-09T12:36:00.000-05:00Oh, shoot. As long as I'm here, I was going to ask...Oh, shoot. As long as I'm here, I was going to ask, can you suggest a reference describing techniques to apply "student-t like" hypothesis tests to data that are Poisson distributed?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-19706647.post-81470472703911796912007-01-09T12:33:00.000-05:002007-01-09T12:33:00.000-05:00Jim
Thanks. True. I wasn't thinking in as Poisson ...Jim<br />Thanks. True. I wasn't thinking in as Poisson processes. <br /><br />As it happens, I also don't think the snippet buried in my dicussion of Holland in any way "means" there is a 200-300 year oscillation. Only that if one were to use the "eyeball" method, and "think" of a sine wave, you could speculate that between 1/2 and 3/4 of a period has based. This would obviously be pure Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-19706647.post-34514538021826494132007-01-08T23:13:00.000-05:002007-01-08T23:13:00.000-05:00Anon,
Be careful. A return period is the expected...Anon,<br />Be careful. A return period is the expected (or average) time between events. It does not imply an oscillation or a preferred time period. In fact the use of return periods for describing rare events is strictly valid only for a random set of events over time. I hope this helps.<br />JimJames Elsnerhttps://www.blogger.com/profile/02586522288616135639noreply@blogger.comtag:blogger.com,1999:blog-19706647.post-27601286626879743192007-01-08T21:53:00.000-05:002007-01-08T21:53:00.000-05:00I was commenting on Holland and Webster at my blo...I was commenting on Holland and Webster at my <a href="http://www.truthortruthiness.com/blog">blog</a>, made what I thought was a joke that maybe I had "discovered' a 200-300 periods oscillation in hurricanes.<br /><br />Still, it was a coincidence to come here and read the time period you found.Anonymousnoreply@blogger.com