Raw Data for "STR-ConvLSTM: A Spatiotemporal Residual Learning Framework for Multivariate Salinity Forecasting in the Yangtze River Estuary"

Dataset for "STR-ConvLSTM: A Spatiotemporal Residual Learning Framework for Multivariate Salinity Forecasting in the Yangtze River Estuary" (PeerJ submission). It includes spatiotemporal data (29.5°N–33.5°N, 121.0°E–123.0°E) for the Yangtze River Estuary: Sea Surface Salinity (SSS, primary predictand) and covariates (current velocities, SST, SSH). Sourced from CMEMS GLOBAL_MULTIYEAR_PHY_001_030 (1/12° spatial, daily temporal resolution), these raw/preprocessed data support STR-ConvLSTM model training/validation.