Dynamical model 2:
Two-Strain Dengue Model

1. Model Overview and Objectives

The model considers dengue transmission between provinces (patches), accounting for human mobility between them.

Key aspects include:
– Primary and secondary infections through mosquito bites in different provinces.
– Integration of rainfall and temperature data (f(T, R) and f(T, Racc)) to model seasonality.
– Human mobility patterns affecting local transmission.

Objectives (2024-2025):
– Calibrate the model using Bangkok dengue data from 2009 onwards.
– Estimate key epidemiological parameters.
– Study intervention strategies to reduce both primary and secondary infections.

Long-term goals include calibrating the model for all of Thailand, establishing an early warning system, and optimizing intervention strategies linked with mobility.

2. Model Dynamics and Equations

The model uses a compartmental structure considering two dengue strains and mobility across patches. Infections are modeled based on both local exposure and exposure due to mobility.

The following diagram illustrates the dynamics and mathematical formulations of the model:

The force of infection is defined for both primary and secondary infections, adjusted for mobility between patches. Temperature and rainfall functions f(T, R) and f(T, Racc) modulate transmission intensity.

3. Model Fitting and Prediction

Data from Bangkok and Chiang Mai were used for model fitting across three intervals: 1968-1984, 1985-2000, and 2001-2018. Each interval includes a validation period.

The model predicts dengue cases for different time horizons (0, 1, 3, 6, and 12 months) and provides:
– Peak month (when the highest number of cases occur)
– Peak number of cases
– Total seasonal cases

These predictions are generated using both raw temperature and rainfall data, as well as accumulated rainfall (Racc).

Further predictions extend to multiple provinces (e.g., Nakhon Ratchasima, Chai Nat, Chiang Rai, Ubon Ratchathani), supporting the model’s scalability for national application.