Groundwater modelling in relation to the system's response time using Kalman filtering

To simulate the effects of human interventions in a geohydrological system, often stationary groundwater models are used. An important question arises: Knowing that a groundwater system is in principle dynamic, can a stationary model be applied? In this paper this question is discussed. First, a one dimensional example is used to consider the influence of the characteristic response time of the geohydrological system on the (reliability of) model results. Next, the use of a Kalman filter algorithm for assessing the reliability of a deterministic groundwater model is described. To integrate measurements with stationary model predictions, a stationary algorithm is used. When it is applied to fast reacting groundwater systems the same results as the (transient) Kalman filter algorithm can be obtained at only a fraction of the calculation time. Finally, the results are further evaluated using a 'real world' case of network analysis and design.