Processed DEM data and MATLAB code for Nu River cascade reservoir terrain correction and climate-resilience assessment

This repository contains processed data and MATLAB R2024b code supporting the manuscript “Future cascade reservoirs could enhance climate resilience in China’s last large free-flowing transboundary river”, submitted to Earth's Future. The archive includes four main components: (1) corrected GeoTIFF DEM files for the 13 planned Nu River cascade reservoirs after terrain correction, (2) original uncorrected GeoTIFF DEM files used as the baseline terrain data, (3) MATLAB R2024b code for LSTM-based runoff simulation, and (4) MATLAB R2024b code for DDPG-based cascade reservoir operation modelling. The corrected DEM files were generated using the terrain-correction workflow described in the manuscript and were used to reconstruct reservoir elevation-area-volume relationships and evaluate storage-capacity differences. The LSTM code was used to simulate runoff using meteorological forcing, and the DDPG code was used to assess potential cascade-operation effects under future hydroclimatic conditions. The code was developed and tested in MATLAB R2024b. Required toolboxes include Deep Learning Toolbox and Reinforcement Learning Toolbox. Users should refer to the README file for file descriptions, software requirements, input-data organization, and reproduction instructions. Restricted engineering-design documents and observed hydrological records are not redistributed in this archive because of data-owner restrictions. Instead, this record provides the processed non-proprietary DEM products, model code, and derived materials needed to support peer review and reproduce the main analytical workflow.