Data and code from: Integrating microclimate to understand vector development and disease patterns: Challenges and lessons from plague in Madagascar's Central Highlands

This dataset contains code and data supporting the generation of
microclimate-driven estimates of plague vector development across the
Central Highlands of Madagascar. It comprises three zipped archives. The
first archive includes scripts, spatial input datasets, and intermediate
outputs required to construct a mechanistic microclimate model of
below-ground temperature at 0.15 m depth for summer 2023 using the
microclimf modelling framework. Input data include categorical soil class
rasters, a digital elevation model, MODIS-derived land-cover rasters, and
hourly ERA5 climate data, and the archive produces hourly microclimate
temperature rasters. The second archive contains scripts and microclimate
outputs used to convert hourly below-ground temperatures into annual
Vector Development Index (VDI) rasters for two flea species (Synopsyllus
fonquerniei and Xenopsylla cheopis). The third archive contains scripts
and data for calculating temporal slopes of VDI, plotting VDI gradients,
and conducting null analyses that compare observed pre-case VDI gradients
with temporally permuted null case series. Dummy human plague case
datasets (RDS and CSV formats) are provided to enable execution of the
workflow without releasing sensitive surveillance records. All analyses
are implemented in the R programming environment, with package
dependencies specified within scripts and installed automatically via the
pacman package. No personally identifiable information is included.
Original human plague surveillance data are curated by the Plague Unit of
the Institut Pasteur de Madagascar and may be requested separately. The
dataset supports reproduction of the full analytical workflow, adaptation
of the modelling framework to other regions or vector systems, and reuse
of scripts for microclimate modelling, vector development estimation, and
null-model based gradient testing. Data are released under a public domain
waiver.