Long-term global climate prediction; rogue science?

Abstract

warming be at the same time responsible for cooling? Climate is normally defined in terms of six variables; temperature, rainfall/ precipitation, humidity, solar radiation, air pressure and wind. Air composition or quality is also increasingly considered as an element of climate. When there is erratic or excessive change in the core variables, then scientists have been quick to point at changing climate. In predicting climate, historical data is analyzed and developed into models for predicting the future. The data can be applied under the architecture of a mechanistic model that uses the underlying thermophysical processes to relate input and output. The term energy and mass balance may be used to describe such mechanistic models. Computational fluid dynamics packages based on these energy and mass balances are also being used nowadays. Alternatively, the inputoutput behavior of the system may be related without considering the underlying thermo-physical processes in the system. Such data-based models are commonly referred to as black-box models. Large sets of input-output data can easily be developed into prediction models through ‘system identification’ which is essentially regression models for large data sets.