Publications

Liu, RX; Zhang, FZ; Gao, YH; Zhang, JB; Liu, ZX; Li, ZH; Yang, JT (2025). Winter wheat maturity date prediction using MODIS/ECMWF data: Accuracy evaluation and spatiotemporal variation analysis. EUROPEAN JOURNAL OF AGRONOMY, 167, 127581.

Abstract
Winter wheat is a staple food crop, and accurate monitoring the maturity date of winter wheat is crucial for increasing grain yield and achieving a bountiful harvest throughout the year. There are problems with crop growth models and remote sensing methods, such as excessive input parameters for the model, inability to achieve real-time prediction, and expansion beyond the experimental area. In response to the above issues, this study proposes a new method for predicting the maturity date of winter wheat in large-scale regions by combining remote sensing data with Growing Degree Days (GDD). Taking Shandong Province as the research area, firstly, MODIS Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), and Kernel Normalized Difference Vegetation Index (kNDVI) time series data were calculated based on the Google Earth Engine (GEE) platform. The Savitzky-Golay (S-G) filter was used to smooth the time series curves, and the vegetation index (VI) maximum method was used to extract the heading date of winter wheat. Secondly, by combining the phenological observation data of winter wheat provided by Shandong Agricultural Meteorological Stations with European Centre for Medium-Range Weather Forecasts (ECMWF) meteorological data, the average Accumulated Growing Degree Days (AGDDave) was calculated. Finally, by comparing and analyzing the accuracy of the heading date extracted from the three vegetation indices (VIs), the corresponding heading date was selected as the starting point for the GDD prediction model to calculate Accumulated Growing Degree Days (AGDD). When AGDD >= AGDDave, winter wheat was considered mature. The research results indicate that compared with the phenological observation data of meteorological stations, the GDD prediction model established based on kNDVI has high accuracy (RMSE = 2.69 d, R2 = 0.63). From the perspective of spatial distribution characteristics, both phenological stages of winter wheat in Shandong Province show a gradually delayed distribution from south to north and from west to east. For every 1 degrees increase in latitude, the heading date is delayed by about 2.38 d, and the maturity date is delayed by about 2.44 d, while the longitude difference in phenological stages is not significant compared with the latitude difference. From the perspective of time distribution characteristics, the interannual variation trend of winter wheat heading date is mainly advanced, with the interannual variation rate of advanced trend concentrated between- 14 d/10a and- 5 d/10a. The interannual variation trend of winter wheat maturity date in Shandong Province is mainly delayed, with the variation rate concentrated between 5 d/10a and 14 d/10a. This method fully considers the important impact of temperature on crop growth and development, and can achieve large-scale prediction of winter wheat maturity date, providing effective assistance for agricultural managers in field crop management and yield prediction.

DOI:
10.1016/j.eja.2025.127581

ISSN:
1873-7331