Abstract:

This study constructed a dataset characterizing the historical variations in eight phenology stages (sowing, emergence, fifth true leaf, bud emergence, bolting, flowering, green maturity and maturity) and their lengths for winter rape, based on long−term phenological observation data (1981−2022) from 13 national agrometeorological observation stations in Jiangxi province, China. Utilizing standardized quality control protocols, kernel density estimation and linear tendency analysis, the dataset comprehensively documented the kernel density distribution patterns of Ordinal day from Jan.1 (DOY) for 8 phenology stages and its 4 lengths across all 13 stations, along with historical trend equations and statistical significance test results. The dataset comprised 256 JPG figures (345MB) organized into 14 sub−folders. This data resource provided critical scientific support for estimating yield potential, optimizing field management practices, refining regional cultivation zoning, mitigating agrometeorological risks and breeding improved cultivars. It is significant practical value for advancing the national strategic objective of "expanding oilseed production capacity" and offering essential data−driven insights to enhance the climate resilience of winter rape production systems in Jiangxi province. The dataset is publicly accessible via the ScienceDB platform at https://doi.org/10.57760/sciencedb.22236.

Key words: Winter rape, Phenology, Kernel density estimation, Linear propensity estimation