Intra-annual crop sequences are particularly important in double-cropping regions for understanding cropland productivity, farmers’ decision-making processes, and the constraints imposed by local water availability. However, the heterogeneous nature of croplands makes their characterization challenging, as neighboring pixels may represent either single-cropping or double-cropping systems, with cropping seasons varying from pixel to pixel. A phenology-parameterization approach can address this issue by explicitly modeling intra-annual crop phenology using representative functions such as skewed normal distributions or double-logistic functions. In this study, we demonstrate the usefulness of this approach by focusing on agricultural regions in Japan where single- and double-cropping systems coexist. The resulting dataset provides not only crop classifications but also representative phenological information for each crop. This approach offers a novel framework for both classifying cropping systems and quantifying crop phenology, thereby improving our understanding of agricultural dynamics in heterogeneous landscapes.