Extracting the Spatiotemporal Pattern of Cropping Systems From NDVI Time Series Using a Combination of the Spline and HANTS Algorithms: A Case Study for Shandong Province

Multiple cropping systems are beneficial for grain yields. Monitoring the spatial and temporal variations in cropping systems is important for evaluating food production and making scientific decisions for agricultural development. Satellite-derived NDVI time series can describe crop metrics (e.g., seeding, jointing, and harvesting) and can be used to obtain cropping information. In this study, rather than statistical data, MODIS NDVI time-series data at a resolution of 250 m for 2000, 2005, and 2010 were used to extract the cropping index in Shandong province. A method combining the spline interpolation and HANTS algorithms was proposed to reconstruct high-quality NDVI time series. The results showed that the proposed method can compensate for the weaknesses of the individual algorithms and effectively reduce the noise in NDVI time series. The NDVI-derived cropping index using the proposed method showed a high degree of consistency with statistical data and field sample data. The errors in the cropping index were due to land use changes and mixed crop types within a pixel. Croplands with 2 crops per year were mainly located mainly in the western plain, whereas croplands with 1 crop were found throughout Shandong province, most predominantly in the north, the middle mountainous regions, and the eastern hilly regions. The cropping index was largest in 2000 and smallest in 2010. The interannual variations are caused mainly by social and economic factors such as economic profit and crop rotation.