🧭 Conceptual Overview
In advanced epidemiological modeling, the assumption of a single, well-mixed population is often an oversimplification. The Multi-Group SEIR Model explicitly acknowledges population heterogeneity by dividing individuals into distinct groups based on age, behavior, occupation, or risk profile. Each group exhibits unique contact patterns, biological susceptibility, and disease progression characteristics. This framework enables precise identification of core groups, disproportionate drivers of transmission, and differential impacts of interventions, making it indispensable for realistic epidemic assessment and policy design.
🏗️ Compartmental Structure and Flow
The population is divided into n distinct groups. For each group i, individuals progress through four epidemiological states:
Susceptible (Sᵢ)
Individuals in group i who have not yet been infected and are at risk.
Exposed (Eᵢ)
Individuals in group i who have been infected but are not yet infectious (latent period).
Infectious (Iᵢ)
Individuals in group i who can transmit the pathogen to others.
Recovered (Rᵢ)
Individuals in group i who have cleared the infection and acquired immunity.
Flow of the System
Sᵢ → Eᵢ → Iᵢ → Rᵢ
The defining feature is that infection of susceptibles in group i depends on infectious individuals from all groups j, weighted by a mixing (contact) matrix that quantifies inter-group interactions.
🧮 Mathematical Formulation
The model is governed by a system of 4n coupled ordinary differential equations. For each group i:
Susceptible dynamics
dSᵢ/dt = − Sᵢ · Σⱼ ( βᵢⱼ · Iⱼ / Nⱼ )
Exposed dynamics
dEᵢ/dt = Sᵢ · Σⱼ ( βᵢⱼ · Iⱼ / Nⱼ ) − σᵢ · Eᵢ
Infectious dynamics
dIᵢ/dt = σᵢ · Eᵢ − γᵢ · Iᵢ
Recovered dynamics
dRᵢ/dt = γᵢ · Iᵢ
Here, Nⱼ = Sⱼ + Eⱼ + Iⱼ + Rⱼ represents the total population of group j.
📐 Table 1. Parameter Definitions
| Parameter | Definition |
|---|---|
| Sᵢ | Susceptible individuals in group i |
| Eᵢ | Exposed individuals in group i |
| Iᵢ | Infectious individuals in group i |
| Rᵢ | Recovered individuals in group i |
| βᵢⱼ | Transmission coefficient from group j to group i |
| σᵢ | Progression rate from exposed to infectious in group i |
| γᵢ | Recovery rate in group i |
| Nᵢ | Total population size of group i |
🌤️ Climatic Variable Integration: Weather-Driven Transmission
Seasonality and environmental forcing are incorporated by allowing the transmission matrix to vary with climate. Absolute humidity and temperature influence viral stability and host susceptibility across all groups.
A commonly used formulation is:
βᵢⱼ(W) = β₀,ᵢⱼ · [ 1 + ε · cos(2π · t / 365 ) ] · exp( − α · AH(t) )
Where:
• β₀,ᵢⱼ is the baseline contact matrix
• ε represents seasonal forcing due to behavior or climate
• AH(t) is time-varying absolute humidity
• α measures sensitivity to moisture
This approach captures synchronized or asynchronous seasonal waves across population groups.
📊 Table 2. Realistic Parameter Ranges for a General Viral Disease
| Parameter | Typical Range | Interpretation |
|---|---|---|
| βᵢⱼ | 0.2 – 0.8 day⁻¹ | Group-specific transmission intensity |
| σᵢ | 0.14 – 0.5 day⁻¹ | Latent period of 2–7 days |
| γᵢ | 0.1 – 0.33 day⁻¹ | Infectious period of 3–10 days |
| R₀ | 1.5 – 5.0 | Dominant eigenvalue of next-generation matrix |
🎯 Applicability
• Age-structured epidemic modeling
• School closure and age-targeted vaccination strategies
• Sexually transmitted infections with core-group dynamics
• Occupational risk analysis (e.g., healthcare workers)
• Differential impact assessment of interventions
⚠️ Limitations and Key Assumptions
• Group membership is fixed over time
• Requires detailed and reliable contact matrices
• Parameter space grows quadratically with number of groups
• Assumes homogeneous mixing within each group
• Does not capture individual-level network effects
🧠 Why This Model Matters
The Multi-Group SEIR Model represents a critical step beyond uniform population assumptions. It explains why epidemics are often driven by specific subpopulations, why interventions succeed or fail unevenly, and how targeted strategies can outperform blanket policies. For applied mathematics, it forms the backbone of modern, data-informed epidemic modeling.
📚 References
Hethcote, H. W. (2000). The mathematics of infectious diseases. SIAM Review.
van den Driessche, P., & Watmough, J. (2002). Reproduction numbers and sub-threshold endemic equilibria for compartmental models of disease transmission. Mathematical Biosciences.
Diekmann, O., & Heesterbeek, J. A. P. (2000). Mathematical Epidemiology of Infectious Diseases: Model Building, Analysis and Interpretation. John Wiley & Sons.
Wallinga, J., et al. (2006). Using data on social contacts to estimate age-specific transmission parameters for airborne-spread diseases. American Journal of Epidemiology.
Schenzle, D. (1984). An age-structured model of pre- and post-vaccination measles transmission. IMA Journal of Mathematics Applied in Medicine and Biology.