The software's used to create weather and climate forecast are significantly different. Weather forecast software is well development and ever improving particularly for periods of one to fourteen days. Supercomputers processing speeds, accuracy and availability of global seeding data and the sophistication and data output accuracy are constantly improving.
Climate models however are still, with few exceptions in the computing stone age with conflicting basic premises on methods and assumptions. The ¨climate sensitivity number¨, in particular the all important, Equilibrium Climate Sensitivity (ECS) is under review.
It seems a basic assumption on climate models is the clouds structure and reflectivity have remained the same for, 10´s , 100´s 1,000´s and millions of years. Since the dinosaurs didn´t keep good satellite imagery (in fact good imagery only started 50 years ago) the Climate modelling is, understandably, seen by many as floored.
2023 was the hottest year on record. Arctic ice was at its minimum and sea temperatures at an all time high since records began. Severe weather event frequency is at an all time high. This does not match with ECS numbers presently in use.
Here is the exceptionally well researched presentation by the ever excellent Sabine Hossenfelder summarising present modelling methods including detailing the significant deficiencies in present Climate Modelling and how the UK Met office is leading the new climate model thinking to get more meaningful and verified output.
Plus 5 degrees Celsius looks likely! The consequences are depressing. Denial and lack of understanding prevails. (Climate videos are generically the most disliked on you tube).
Comments