This dataset contains a log-linear model based on site-specific average annual temperature, average annual precipitation, average tidal amplitude, sediment organic carbon, aboveground biomass, and reactive iron in the catchment, developed to predict mangrove pyrite stocks. Constants were calculated through an iterative least-squares process using Microsoft Excel. The resulting model output was used to estimate global pyrite stocks in mangroves. The global model inputs included temperature (Fick & Hijmans, 2017), precipitation (Fick & Hijmans, 2017), average tidal amplitude (Vestbo et al., 2018), sediment organic carbon (Sanderman et al., 2018), aboveground biomass (Simard et al., 2019), and reactive iron (Rossel et al., 2016).