Multimedia models are used in a number of applications. For a specified region, these models facilitate the prediction of a chemical�fs distribution, concentration in each media and persistence. In addition to the consideration of agglomerated releases, the contribution to exposure attributable to individual emissions can be estimated. However, the number of chemicals that can be considered using multimedia models is limited by the availability of degradation half-life data, particularly for soils and bed sediments.

Typically the degradation rate of a chemical in water and in air can be estimated. However there are currently no consensus on a suitable approach for the prediction of these rates in soils and bed sediments. A number of steady-state approaches are presented that facilitate the determination of persistence and concentration estimates using available multimedia models given only aquatic and atmospheric degradation rate data. The approaches are derived scientifically, taking into consideration typical cross-media trends in degradation data. 

Given the mass distribution of a chemical and an incomplete set of degradation data, the steady-state persistence and concentrations in a specified region can be estimated. Given the aquatic and atmospheric degradation data, the increase in uncertainty is typically minimal. Depending on the approach used, conservative predictions are generated for screening.

Mass fractions can be readily calculated independently of degradation rates by assuming intermedia transport controls the distribution. This assumption is used in equilibrium partitioning models, for example. However, the assumption that intermedia transport is the controlling process in the distribution of a chemical can result in significant error. If these predictions are used to estimate persistence and exposure concentrations then the uncertainty is increased and there is no approach to ensure the predictions are conservative. However, the steady-state mass distribution of a chemical can be predicted given an incomplete set of degradation half-life data. The methodologies are compatible with those for persistence and concentration. It is demonstrated that the overall uncertainty is reduced and conservative estimates can be generated.