Monitoring chemicals in the environment is an expensive exercise. Samples need to be collected and analyzed in a proper way, which is often costly and time consuming and requires complex organization and laboratory facilities.
Although several monitoring programs exist globally for certain chemicals, data on the majority of substances are still sparse, heterogeneous, and difficult to compare. A few organizations operate globally in the collection and reporting of chemical monitoring data. However, chemical monitoring is still far from being a routinely, unified process.
Under these circumstances, model development entails necessarily a phase of data collection and analysis in order to check model performance, hence the potentials for policy support applications. There are essentially three levels of comparison of models and monitoring data:
At the first level, the aim is simply to verify that a model reproduces correctly the order of magnitude of monitored chemicals. When this happens, it is possible that emissions and chemical removal rates have the correct order of magnitude. However, this is still a necessary but not sufficient condition: for instance, one might be wrong both in the order of magnitude of emissions and in that of removal rates. The model is by chance correct but would not be reliable, for example, ...