Abstract

Dynamical systems are the paradigm for the representation of complex systems. The fixed encoding in a closed set of equations, however, contrasts with the openness of biogeochemical systems. Parameter identification is a major problem in biogeochemical systems and calibration of parameters converts models into 'fitting machines'. Openness, self-modification, and historicity of biogeochemical systems make non-trivial predictions of future outcomes impossible. Notwithstanding, simulation models serve as instruments of synthesis and have heuristic value to challenge existing data and theories. The modeling process itself, as a learning and communication process, can be a mode of coping with different types of complexity