Guidance on the selection of efficient computational methods for multimedia fate models

Area of Science:

• Environmental Science
• Computational Chemistry
• Numerical Analysis

Background:

• Dynamic multimedia fate models (MFMs) simulate chemical behavior in the environment, facing challenges with temporal and spatial variations.
• Analytical solutions are often infeasible for complex MFMs, necessitating robust numerical methods for speed and reliability.

Purpose of the Study:

• To identify sources of stiffness in multimedia fate model equations for realistic chemical and scenario properties.
• To compare the performance of various numerical schemes for solving MFM equations.
• To evaluate the effectiveness of automatically adapted timesteps and implicit timestepping.

Main Methods:

• Pinpointing sources of stiffness in MFM equations using diverse chemical and scenario properties.
• Comparing numerical schemes representing three broad classes: explicit, implicit, and adaptive.
• Evaluating accuracy and computational effort for each numerical method.

Main Results:

• Automatically adapted timesteps significantly enhance solution quality and reduce computation time.
• Implicit timestepping methods demonstrate advantages in handling stiffness.
• A medium-order (around 5) implicit method emerged as the optimal general-purpose engine for MFMs.

Conclusions:

• Automatic error control and adapted timesteps are crucial for efficient and accurate MFM simulations.
• Implicit numerical methods, particularly medium-order ones, provide a superior computing engine for dynamic multimedia fate modeling.