I find it interesting that after years and years of development of computer modeling in aircraft design, engineers still also use wind tunnels. Arguably, the amount of variables involved in aircraft (or auto) development are many fewer and less complex than those involved in weather and climate on a planetary scale, yet there is still enough uncertainty that an actual physical model system remains indispensable. Yet, climate scientists use computer modeling exclusively in their predictions and have the hubris to say it's ‘settled science’..
[[Yet, climate scientists use computer modeling exclusively in their predictions and have the hubris to say it’s settled science..]]
When the facts refute the claims, scream ‘DENIERS!” and claim that it’s “Settled Science” to shut down any debate and to try to make your opponents look like ignorant clueless ‘science deniers’, ‘science haters’- that’s how science works these days
Interesting thought. A model that complex would require a Cray at the very least, and they ain’t cheap.
My largest concern with any predictive model is, of course, the data being fed into it. I believe that there are those who wish to produce an accurate model and get scientific results that can be scrutinized and seen as competent. But without hundreds of thousands of reporting stations, physically separated from artificial heating and cooling, the extracted data is tainted if not discardable.
Stations need to accurately report: temperature, humidity, barometric pressure, wind speed and direction. All of those factors need reported at the surface, 5k’, 10k’, 30k’ and 100k’ in order to garner an accurate representation of conditions on the planet.
After that, we need external influencers such as the Sun and its derivatives such as CME’s and sun spots. Then, we need to account for gravitational perterbations and the moon’s influence as well as that of the oceans and large lakes.
Most of this is conveniently overlooked in climate models.
Engineers use models to solve complex math problems and then check to see if the models make sense. The models for aero need wind tunnel verification before committing loads of cash to an aircraft design. (The models help, the the engineers believe the wind tunnel data more, even though the wind tunnel is only a small scale structure of the actual thing being designed.)
I worked with a modeler who tried to predict the explosive effect of a high pressure tank being shot with a 50 cal round. To test the model, (which predicted only low level over pressure we built a small “room” to study the effects of firing the bullet into the tank. When the tank exploded, it knocked down all four walls and all of our instrumentation was pushed out of range. We went back to the modeler and he said “Whoops” I must have made some wrong assumptions.
This is basically why I do not trust models.
KC, systems engineer (retired)
Engineers use models to solve complex math problems and then check to see if the models make sense. The models for aero need wind tunnel verification before committing loads of cash to an aircraft design. (The models help, the the engineers believe the wind tunnel data more, even though the wind tunnel is only a small scale structure of the actual thing being designed.)
I worked with a modeler who tried to predict the explosive effect of a high pressure tank being shot with a 50 cal round. To test the model, (which predicted only low level over pressure we built a small “room” to study the effects of firing the bullet into the tank. When the tank exploded, it knocked down all four walls and all of our instrumentation was pushed out of range. We went back to the modeler and he said “Whoops” I must have made some wrong assumptions.
This is basically why I do not trust models.
KC, systems engineer (retired)
Leaving “subsidence” off your list?