A long-term, station-based dataset of rice phenology dynamics in Middle and Lower Reaches of the Yangtze River Plain of China from 1981 to 2024

Derived from rigorously quality-controlled records of 95 national agrometeorological stations, this dataset provides two unique products: 1) a curated collection of 2,432 high-resolution visualizations generated through kernel density estimation and linear regression, depicting the probability distribution and trends of 11 phenological stages and 3 growth phase lengths; and 2) a systematically organized, machine-readable Excel spreadsheet that provides the complete statistical results of the linear trend analysis (including province, station ID, station name        , crop types, phenology period, slope (variation trends), intercept, R², model mean square, F-value, and p-value) for each station and variable.

This dataset establishes a critical, and previously unavailable, benchmark for detecting and attributing climate impacts on rice systems, with the machine-readable statistical data offering a transformative resource for reuse. It is specifically designed for immediate reuse, providing ready-to-use inputs for crop model validation, climate-crop meta-analyses, and the development of localized adaptation strategies.