# Computational simulations

Computational simulations are an established and widely used method for tackling hydraulic engineering questions that can help to obtain results about natural processes. They also permit predictions of natural developments and the impact of anthropogenic intervention in natural systems. The simulation methods are supplemented by methods of data processing and grid generation (preprocessing) as well as visualising and analysing the calculation results (postprocessing).

Before using a computational simulation method, the natural system has to be analysed with regard to the processes involved. With a view to the system in question, e.g. river or estuary, and the issue at stake, a choice has to be made in terms of the processes that the model has to consider (model functionality). This is the basis for formulating the conceptual model, usually in the form of ordinary or partial differential equations. As a rule, it is not possible to solve these exactly when natural systems are involved. Individual processes such as bed friction can only be considered in a simplified form (parameter configuration). In addition, the unknown variables (water level, current velocity, etc.) can only be calculated at a finitely large number of places and dates (spatial and temporal discretisation). The fundamental equations and assumptions for discretization and parameter configuration lead to the algorithmic implementation of the model using numerical mathematical methods. This is then followed by software implementation that is verified using basic examples. The result is a mathematical method with defined validity that can be used to simulate various site-specific models (river Rhine near Bingen, tidal Elbe, etc.). Different computational simulation methods can be used at the BAW for a wide range of natural processes.

Preprocessing entails generating initial and boundary values for the entire simulation period as well as the computation grid for the corresponding site-specific model. Later on, the consistent computational results of a simulation are compared with observation data (calibration, validation), visualized in graphic terms, turned into animations to clarify dynamic processes or subjected to further analysis for calculating characteristic parameters (e.g. high-tide, maximum current velocity) and their change (prediction).

9