Model simplification and loss of irreversibility

Area of Science:

• Thermodynamics
• Statistical Mechanics
• Chemical Kinetics

Background:

• Continuous-time Markov chains (CTMCs) are fundamental for modeling dynamic systems.
• Time-scale separation is a common simplification technique in complex systems.
• Irreversibility, quantified by entropy production, is a key thermodynamic property.

Purpose of the Study:

• To establish a general relationship between model simplification and irreversibility in CTMCs.
• To investigate how different simplification strategies affect thermodynamic properties.
• To understand the trade-off between dynamic and thermodynamic information during model reduction.

Main Methods:

• Utilized a two-time-scale continuous-time Markov chain model.
• Classified states into transient and recurrent based on topological structure.
• Investigated two simplification methods: transient state removal and recurrent state aggregation.
• Analyzed changes in entropy production rate and its adiabatic/non-adiabatic components.

Main Results:

• Demonstrated that model simplification in CTMCs leads to reduced irreversibility.
• Showed that both transient state removal and recurrent state aggregation decrease the entropy production rate and its adiabatic part.
• Confirmed that the non-adiabatic part of entropy production remains unchanged by these simplification methods.
• Quantified the loss of thermodynamic information as a trade-off for retaining dynamic information.

Conclusions:

• Model simplification of CTMCs is intrinsically linked to a reduction in irreversibility.
• Simplification strategies offer a way to manage complexity while preserving essential dynamic behaviors.
• There is an inherent trade-off: dynamic information is largely retained, but some thermodynamic information, specifically related to entropy production, is lost.