[AntonioLao]

Global Circulation Models (GCMs) are 3D computer simulation programs predicting future climate changes. These include jet streams, deep ocean currents, solar reflectance off ice sheets and glaciers, vegetation dynamics, cloud movements, and many other climate related factors. The programs work from first principles of thermodynamics, fluid mechanics, carbon cycle, water cycle, etc. Based on the mathematics of finite element analysis http://en.wikipedia.org/wiki/Finite_element_analysis, the elements are 3D grid boxes covering a given contiguous quantized cubic unit of air. Land, and sea and the models step forward in time by 20 minutes or an hour increments, again and again for one complete cycle of the four seasons: spring, summer, fall, and winter. The idealized models are so big and complex that they can only be executed in supercomputers owned by mega-rich organizations: NASA, NCAR, NOAA, and Britain’s Hadley Center.

However, if these programs were to run backward in time by stepping 20 minutes or an hour into the past then they failed to match data of climate conditions that have already happened. The grid boxes would be found empty since negative time increments are not allowed in irreversible entropic nonlinear differential equations. Moreover, these equations could not possibly simulate chaotic and stochastic dynamic processes. These are then analogous to guessing the content of lunch boxes of school children before they open them during lunch recess.