The first 30 m annual plastic-mulched farmland dataset and its dynamics in Xinjiang, China from 2000 to 2022

In this study, we developed a streamlined and effective algorithm to generate annual PMF maps for Xinjiang, China, from 2000 to 2022. Using Landsat imagery, we identified PMF-sensitive spectral bands and critical phenological windows through spectral-temporal analysis, subsequently devising a Phenology-based Plastic-Mulched Farmland Index (PPMFI) to enhance differentiation of PMF from other croplands. We then established a ground-reference PMF dataset using Landsat data, delineating county-specific mulching periods. Integrating the PPMFI, reference dataset, and Landsat time-series archives, we constructed an algorithm to monitor PMF spatiotemporal dynamics across 2000-2022. Validation over four years confirmed robust performance of the proposed algorithm, with overall accuracies exceeding 86%, Kappa coefficients above 0.72, and F1-scores surpassing 0.87. Mapped PMF areas strongly correlated with provincial statistics (R² = 0.93, p < 0.001). Results revealed a 126% increase in PMF coverage in Xinjiang, rising from 1.49 million ha in 2000 to 3.37 million ha in 2022, accompanied by significant spatial expansion.