Keywords

Environmental Sciences, modeling, epistemology

Start Date

25-6-2018 2:00 PM

End Date

25-6-2018 3:20 PM

Abstract

Designing the models and their use in computer simulation in the environmental sciences inherently involves many epistemological questions. Although practical considerations often overruled the problems of epistemology sometimes it is necessary to make basic epistemological choices, especially in modeling. In sciences and accordingly in environmental ones for many reasons modelers ignore non-linearity of phenomena and processes. If we decide to linearize “the object of modeling” then we use linear equations where the variables and their derivatives must always appear as a simple first power. The theory for solving linear equations is rather well developed one because linear equations are simple enough to be solvable. The shortcoming of this approach is that it neglects that many aspects and phenomena, even important ones, which remain hidden. However, if we decide to follow as much as possible the existing nonlinearities in the object that we model we have to consider the following key points: (1) Model choice; (2) continuous-time versus discrete-time in building the model and (3) predictability (Lyapunov time). Our considerations will be supported by some examples in the modeling of environmental sciences processes.

Stream and Session

D1: Environmental Fluid Mechanics – Theoretical, Modelling, and Experimental Approaches

COinS
 
Jun 25th, 2:00 PM Jun 25th, 3:20 PM

Models and modeling in the environmental sciences: Between epistemology and practice

Designing the models and their use in computer simulation in the environmental sciences inherently involves many epistemological questions. Although practical considerations often overruled the problems of epistemology sometimes it is necessary to make basic epistemological choices, especially in modeling. In sciences and accordingly in environmental ones for many reasons modelers ignore non-linearity of phenomena and processes. If we decide to linearize “the object of modeling” then we use linear equations where the variables and their derivatives must always appear as a simple first power. The theory for solving linear equations is rather well developed one because linear equations are simple enough to be solvable. The shortcoming of this approach is that it neglects that many aspects and phenomena, even important ones, which remain hidden. However, if we decide to follow as much as possible the existing nonlinearities in the object that we model we have to consider the following key points: (1) Model choice; (2) continuous-time versus discrete-time in building the model and (3) predictability (Lyapunov time). Our considerations will be supported by some examples in the modeling of environmental sciences processes.