model_name: WNV-R0
model_version: "1.0"
model_type: Process-based epidemiological model
primary_task: Thermal suitability Estimation
domain: West Nile Virus (WNV) Transmission Potential

organization: HeiPlanet
date_created: "2026-01-12"
contact: julian.heidecke@iwr.uni-heidelberg.de

description: >
    WNV-R0 estimates a relative version of the basic reproduction number (R₀) 
    for WNV transmission by Culex pipiens using ambient temperature data. 
    The model is intended for seasonal and regional temperature suitability 
    assessments rather than real-time forecasting of WNV outbreaks.

intended_use:
  in_scope:
    - Temperature-driven seasonal WNV risk assessment
    - Regional comparison of thermal suitability
    - Historical warming impact assessments on transmission potential
    - Climate scenario analysis
    - Public health preparedness planning

  out_of_scope:
    - Individual risk prediction
    - Real-time outbreak forecasting
    - Automated policy decisions

architecture:
  type: Process-based epidemiological model
  characteristics:
    - Ross-Macdonald type relative reproduction number
    - Derived from deterministic compartmental vector–host model
    - Temperature-dependent mosquito-pathogen parameters for Culex pipiens

input_features:
  temperature:
    description: Monthly mean ambient temperature
    units: °C

outputs:
  r0_estimate:
    description: Estimated relative reproduction number
    units: Dimensionless

training_data:
  sources:
    - Laboratory experimental data on mosquito-pathogen traits
    - Temperature response curves were fitted using Bayesian hierarchical models

validation:
  testing_data:
    description: Historical WNV outbreak records in Europe
  metrics:
    - Seasonal and geographical alignment between R0 and cases
    - Rank correlation coefficients
    - Overlap between lab-based and field-observed "optimal" temperature for transmission

assumptions:
    - R₀ is a measure of long-term average transmission under constant temperatures
    - Homogeneous spatial mixing
    - No explicit modeling of interventions
    - No intra-species mosquito variability in temperature sensitivity

limitations:
    - Deterministic structure
    - Neglects host community composition and immunity dynamics
    - Only focused on temperature-driven effects on transmission potential via mosquito-pathogen traits
    - Accurate risk predictions need to account for additional climatic factors
    - Cannot predict number of cases 
    - Cannot account for potential adaptation of mosquito populations to increasing temperatures
    - Relative R₀ cannot be interpreted as a threshold parameter like absolute R₀


bias_and_fairness:
    - Based on laboratory data
    - Only validated against WNV observations in Europe (and to some extent in the USA)

potential_harms:
    - Misuse without expert review
    - Misinterpretation of outputs

risk_mitigation:
  - Expert review
  - Communication of limitations and uncertainty

maintenance:
  update_policy: Updates as new data become available
  development_approach: Continued methodological updates and model extensions

datasets:
    - Laboratory experimental data on mosquito-pathogen traits compiled through 
    systematic literature review
    - Copernicus ERA5-Land Climate Data
    - ECDC human West Nile Neuroinvasive disease cases data for validation

publications: 
    - Heidecke, J., Wallin, J., Fransson, P., Singh, P., Sjödin, H., Stiles, P. C., ... & Rocklöv, J. (2025). Uncovering temperature sensitivity of West Nile virus transmission: Novel computational approaches to mosquito-pathogen trait responses. PLOS Computational Biology, 21(3), e1012866.
    - Heidecke, J., Fransson, P., Wallin, J., & Rocklöv, J. Thermal Biology-Informed Reproduction Number Explains Spatiotemporal Patterns of West Nile Incidence in Europe. Available at SSRN 5597581.