SD–GIS-based temporal–spatial simulation of water quality in sudden water pollution accidents

System dynamics (SD) is well suited for studying dynamic nonlinear complex systems. In this paper, SD is applied to a rapid-onset water pollution accident using a 1-D water quality model and a conceptual GIS–SD framework is constructed to simulate the temporal–spatial changes of pollutant concentration. Based on the component GIS and the SD model, a prototype system of water quality simulation in water pollution accidents is developed. The data collected on the spot in the Songhua River water pollution accident in November 2005 were used for model parameter calibration and model validation. The results showed that: (1) the constructed model could simulate the changes of nitrobenzene concentration with time in the Songhua River water pollution accident, especially during the peak concentration and at the arrival time of peak concentration, and that the simulated values and the on-the-spot monitored values corresponded with each other well; (2) the scenario simulation could be made by adjusting parameters u (longitudinal current velocity), E (longitudinal diffusion coefficient), and k (decay rate coefficient). Such a model can provide decision makers with quantitative information to optimize related emergency response measures.