tag:blogger.com,1999:blog-12339127.post3446668836499980886..comments2024-01-13T01:17:55.325-06:00Comments on Tom Nelson: Settled science: "The ability to predict retrospectively this slowdown not only strengthens our confidence in the robustness of our climate models, but also enhances the socio-economic relevance of operational decadal climate predictions."Tomhttp://www.blogger.com/profile/08119241500221931600noreply@blogger.comBlogger4125tag:blogger.com,1999:blog-12339127.post-38654797525735769282013-05-17T11:32:33.589-05:002013-05-17T11:32:33.589-05:00Predicting the past isn't as easy as it sounds...Predicting the past isn't as easy as it sounds lolpinroothttps://www.blogger.com/profile/05584997377097902088noreply@blogger.comtag:blogger.com,1999:blog-12339127.post-11042716433657302722013-05-15T07:15:07.831-05:002013-05-15T07:15:07.831-05:00I agree with Anthony, if a model fails to predict ...I agree with Anthony, if a model fails to predict the past it's no more than digital palm reading when applied to the future. Referring to an undiscovered, ethereal feedback mechanism to explain failure is ridiculous. Betting all your nuts on CO2 confirms you are a squirrel.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-12339127.post-74009439849389170482013-05-14T14:32:35.621-05:002013-05-14T14:32:35.621-05:00Sure a retrospective model is very valuable in pro...Sure a retrospective model is very valuable in proving the correctness of the model, and thus, its value for predicting the future. The big problem for the 'warmists' is that there retrospective tests all fail! Lindzen has compared the calculations for the period 1998-2013 using the formulae developed prior to 1998 with actual resuts for this 1998-2013 period, showing a wide divergence between the measured and the caculated results.artcohnhttps://www.blogger.com/profile/07419382569294687644noreply@blogger.comtag:blogger.com,1999:blog-12339127.post-15092090029035333992013-05-14T14:13:52.800-05:002013-05-14T14:13:52.800-05:00For me to believe they have this correct, the mode...For me to believe they have this correct, the model must predict the past without any modifying of the model. Once a test case is made to work it can no longer be used to 'prove' the model predicts correctly. A new situation must be used. After the model can predict the past to an acceptable level then I might believe it can predict the future.Anonymousnoreply@blogger.com