Evaporation pan with floating covers under infrared lamps (MAISSER-PE): data and code for a Bayesian heat–mass transfer model and state-space inference

Data, code and model outputs of two modelling studies of an instrumented Class A evaporation pan heated by halogen infrared lamps, ventilated by a wind tunnel and partly covered by 3-D-printed hexagonal floating plates (project MAISSER-PE, Departamento de Energia Nuclear, Universidade Federal de Pernambuco, Recife, Brazil; campaigns C1, August–September 2026, and C2, September–October 2026). (1) A one-dimensional heat and mass transfer model calibrated in a Bayesian framework (tempered AR(1) likelihood, differential evolution and an affine-invariant ensemble sampler), with Shapley and Sobol analyses of evaporation suppression. (2) A state-space analysis (augmented Kalman filter with diffuse likelihood) of the level records, a hierarchical model of suppression and a Bayesian experimental design. Contents: raw one-minute CR1000 records of campaign C2 (the C1 records are copied from doi:10.5281/zenodo.22844940), Python code of both studies, MCMC chains, posterior modes and summaries, scenario results, tables and figures. Code under MIT; data and outputs under CC BY 4.0. Funded by FACEPE (APQ-1428-3.01/25).