All codes were created and run using R statistical software version 3.5.3 and Rstudio version 1.1.383.
This repository provides code to run climate-based mechanistic models of arboviral transmission. The models support of the manuscript "Climate predicts geographic and temporal variation in mosquito-borne disease dynamics" by Caldwell et al. Nature Communications (https://www.nature.com/articles/s41467-021-21496-7).
Temperature-, rainfall-, and humidity-dependent model: SEI-SEIR_model_THR.R
Temperature-, rainfall-, and humidity-dependent model with trait variation: SEI-SEIR_model_with_trait_variation.R
Code to set up simulations: SEI-SEIR_simulation_setup.R
Initial conditions (proportion of population in each compartment): LHS_inputs.csv
Posterior distribution of c, T0, Tmax for temperature-dependent traits: Random_sample_of_posterior_traits.csv
Epidemic characteristics datasets: (1) Duration_intensity_outbreaks.csv and (2) Number_outbreaks.csv
Time series of temperature, humidity, and rainfall for study sites: climate_data.RData
Time series of vector data for study sites: vector_data.csv. Only adult Ae. aegypti data was collected at Ecuador sites.
Time serires of arboviral data for study sites: arbovirus_data.csv. Data is only available for Kenya sites.
A single code is provided to run each model with appropriate data.
Compare model output with observational data: pairwise_correlation.R
Compare model output with different socio-ecological characteristics: socio_ecological_relationships.R
Compare predictions and observations for different epidemic characteristics: epidemic_characteristics.R