Integrated use of monitoring and modeling in water resources research

Despite the many advances in sensors and recording techniques, monitoring programs can still be relatively expensive. In practice, this often limits the density of monitoring programs. Yet, large amounts of data are monitored and filed without proper analysis of their information contents. The combined use of monitoring and simulation models can reduce the costs and facilitate rigorous analyses of monitored data. Physically based simulation models provide the best means of interpolating between measurement points (in space and time). The models can also aid the effective design of monitoring programs. Field data can be used to improve the quality of simulation models. For real time monitoring, information can be fed back into the simulation models through automatic update routines. These combined techniques, long used for hydraulic data, are now also developed for water quality data. Whenever possible, the integration between monitoring and modeling should be designed from the outset to obtain full benefit. New techniques are developed for linking the two methodologies including data mining, data validation, and data assimilation techniques. The paper describes some of the recent developments in this field, giving examples of practical applications.