Code and generated data for a 5 arc-min agroclimatic suitability model of rainfed peanut under CMIP6 scenarios (Mato Grosso do Sul and São Paulo, Brazil)

This package contains the code and the generated data of an agroclimatic suitability model for rainfed peanut (Arachis hypogaea L.) in the Brazilian states of Mato Grosso do Sul and São Paulo, at a resolution of 5 arc-min. Suitability classes (Suitable, Restricted, Unsuitable) are derived from the water-requirement satisfaction index of the critical phase and a thermal criterion, for the 1983–2012 baseline (CRU TS 4.09 downscaled with WorldClim 2.1) and for nine CMIP6 runs that combine three scenarios (SSP1-2.6, SSP2-4.5 and SSP5-8.5) with three periods (2021–2040, 2041–2060 and 2061–2080), using a weighted mean of seven global climate models. Each run is evaluated for three uniform hypothetical soil scenarios and six sowing months (September to February). Included: the complete processing chain (R, Python and PowerShell scripts), the generated rasters and vectors, area tables by class, soil, sowing month, run and state, complementary analyses (signal and dispersion among models, robustness, sensitivity), the agreement with the official Agricultural Climate Risk Zoning (ZARC), the master table of results, the figures and their scripts, and the logs of the run. Third-party inputs (WorldClim, IBGE boundaries and the ZARC table) are not redistributed; the scripts download or read them, and their sources are listed in the package. The package supports a master's dissertation and the journal articles derived from it; their references will be added as related works after publication. Code: MIT License. Generated data, tables and figures: CC BY-NC 4.0, consistent with the WorldClim terms of use. See README.md for the reproduction steps and the correspondence between files and results.