"... using dynamic models in static mode ..."
Because
ALL of your models
FAIL, WITHOUT EXCEPTION. I dont use garbage to predict what is coming.
Dynamic models and climate models use different algorithms ... any similarities are strictly superficial ... I've been following NWS copy for over 25 years and these dynamic models are actually quite good and mostly accurate out three days ... and the forecasters in my area are usually quite candid about the results ... if the several all give the same solution, then that's the forecast and almost always spot-on ... if the several are everyplace, the forecasters will say that and admit to guessing tomorrow's weather ... every spring, the National Hurricane Center postmortems all their forecasts the previous hurricane season ... and a fairly exhaustive error analysis is posted every year ... dynamic models don't "fail" at 72 hours out very much ... past 72 hours you'll want your money on the pass line ... I believe the models NOAA developed are available for download, you can run them yourself and read through the code; unfortunately, the better dynamic models are proprietary and only the results can be made public, and the error analysis I mentioned above ...
The problem is that forecasters only have six hours to run the several models, interpret the results and type out a dozen text documents ... then start over for the next forecast package ... and everything is changing all the time ... so in "static mode", nothing's changing which greatly improves accuracy ... and we can set a bigger, faster, better computer on the problem and let it run for a few weeks ... and most important, we use the exact same computer, the exact same code for the exact same time for
each year ... so whatever errors there are will be reflected through the
entire data set ...
Climate models are a different kittle of fish ... punch in data from 100 years ago and run the fool thing ... then go look and see if the oceans are boiling off ... No? ... keep that in mind when we punch in today's data and run it ... the ocean might not be boiling off 100 years from now either ... if ($∆T > 1ºC) {echo "The oceans will boil off";} ... else {echo "The oceans will <i>still</i> boil off";} ...
The National Enquirer pays good money for results like this ...
I'm blathering again aren't I? ... go look at the 5 day forecast for Hurricane Sandy ... five days before she made landfall on the New Jersey coast ... now explain why you think that's a failure? ...