Zonal Statistics of Relative Humidity derived from ERA5-Land for Brazilian Municipalities, 1950-2025

Climate indicators are widely used in statistical models across multiple research fields and are particularly important for modelling the incidence of Climate-Sensitive Diseases (CSDs). These models typically adopt a lattice structure, in which data are aggregated over administrative units (e.g., municipal disease incidence). In contrast, climate indicators are commonly provided as continuous variables on regular spatial grids. To reconcile these differing spatial representations, zonal statistics can be applied. Zonal statistics are descriptive measures computed from raster cells that intersect a given geographic boundary spatially. For each spatial unit, summary statistics—such as the mean, maximum, minimum, standard deviation, and sum—are calculated to represent the values of the underlying grid within that boundary. This dataset provides daily mean zonal statistics of relative humidity (percent) for Brazilian municipalities from 1950 to 2025. The estimates are derived from Copernicus ERA5-Land daily aggregates of dew point temperature and mean 2-m air temperature, using the following formula: e_actual <- exp((17.625 * dew_point) / (243.04 + dew_point)) e_saturation <- exp((17.625 * temp_mean) / (243.04 + temp_mean)) rh <- round(100 * (e_actual / e_saturation), 2) The full code is available here.