Supplementary data and analysis code for: Long-Term Assessment of Salinity Gradient Power Potential in the Lake Urmia–Simineh River System

This dataset contains the complete data compilation, analysis code and processing documentation underlying the article "Long-Term Assessment of Salinity Gradient Power Potential in the Lake Urmia-Simineh River System" (Results in Engineering, manuscript RINENG-D-26-10333). The study evaluates the salinity gradient power (SGP) potential of the Lake Urmia-Simineh River system in northwestern Iran over a 28-year period (1993-2020). Theoretical potential is estimated from the Gibbs free energy of mixing using a NaCl-equivalent limiting-ion formulation, and the technical output of pressure retarded osmosis (PRO) and reverse electrodialysis (RED) is estimated using literature-derived efficiency models. Contents are organised in eight folders: 1. Lake Urmia data - published sodium and chloride measurements (1967-2020) compiled from the literature, unit standardisation from mEq/L to g/L, the interpolation script and the resulting annual series. 2. Simineh River data - annual water chemistry (1993-2023) and monthly discharge (2003-2023), the linear-interpolation and KNN imputation scripts, holdout validation results and methodology documents. 3. Theoretical potential - input dataset, the Forgacs-equation calculation script, descriptive statistics and correlation analysis. 4. Seasonal analysis - seasonal input data and the decomposition script for 2003-2013. 5. Efficiency analysis - the PRO and RED efficiency models and the resulting power-output series. 6. Statistical analyses - Monte Carlo simulation, analytic error propagation and sensitivity analysis. 7. Economic viability analysis - revenue and break-even calculations for ten membrane scenarios. 8. Figures - the code used to generate every figure in the article, with a figure list documenting the source, code file and revision status of each. All scripts are Jupyter notebooks written in Python 3 (pandas, NumPy, scikit-learn, SciPy, matplotlib). Every stochastic procedure uses a fixed random seed and the imputation routines are deterministic, so all values reported in the article are exactly reproducible from the material supplied. Data sources are the published studies cited in the article, together with monthly hydrometric records obtained from the Iranian Ministry of Energy - Azerbaijan Water Regional Company. No restricted, personal or third-party licensed data are included.