Climate and Environmental Drivers of Dengue Expansion in São Paulo: Data and Code

This repository contains the curated datasets and the full R source code used in the manuscript "Climate and environmental drivers of dengue expansion in São Paulo, Brazil: An ecological niche modelling approach", submitted to Remote Sensing Applications: Society and Environment.

This package is designed to ensure full transparency and reproducibility of the Ecological Niche Modelling (ENM) analysis presented in the study.

Contents:

1. Processed Dataset (.csv) This file contains the fully pre-processed data used as input for the MaxEnt models. It aggregates epidemiological, climatic, and demographic variables for all 645 municipalities of São Paulo State for the analyzed epidemic years (2011, 2015, and 2019).

Spatial Resolution: Municipalities (aggregated from raster data as described in the manuscript).

Temporal Resolution: Semi-annual (January–June) averages for environmental variables; Annual incidence rates for dengue.

Variables included:

Municipality Code (IBGE)

Dengue Incidence Rate (cases per 100,000 inhabitants)

Precipitation (mm) - WorldClim v2.1

Maximum & Minimum Temperature (°C) - WorldClim v2.1

Elevation (m) - WorldClim v2.1

Normalized Difference Vegetation Index (NDVI) - NASA/MODIS

Population Density (people/km²) - WorldPop

2. Analysis Script (niche-modelling-script.R) The complete R script required to reproduce the results, figures, and tables. The workflow includes:

Data Pre-processing: Loading and cleaning raw data.

Multicollinearity Test: Variance Inflation Factor (VIF) analysis.

Model Training & Validation: MaxEnt modelling using dismo, incorporating spatially explicit cross-validation via blockCV to account for spatial autocorrelation.

Evaluation Metrics: Calculation of Area Under the Curve (AUC) and True Skill Statistic (TSS).

Mapping: Generation of Environmental Suitability Index maps and Population at Risk maps using tmap.

Instructions for Reproducibility:

Download both files to a local directory.

Ensure the necessary raw raster files (WorldClim, MODIS) are available in your working directory if running the raster extraction steps from scratch (details provided in code comments).

Run the script niche-modelling-script.R in R or RStudio.

Data Sources:

Dengue Cases: DATASUS (Ministry of Health, Brazil) - https://datasus.saude.gov.br

Climate Data: WorldClim v2.1 - https://www.worldclim.org

Remote Sensing: NASA LP DAAC (MODIS/NDVI) - https://lpdaac.usgs.gov

Population: WorldPop - https://www.worldpop.org