Abstract:
As arable lands are highly limited and intensively exploited in China, it has become important to increase crop yield in low-to-medium yield farmlands in order to improve grain output. Forecasting crop yield few months before harvest using remote sensing technique has often been used in most crop yield estimation. In view of the above research condition, the potential productivities of main crops were estimated in low-to-medium yield farmlands in the Lowland Plains of Hebei, where is the area of scientific and technological demonstration of the Bohai Granary project. Using MODIS/NDVI remote sensing data (with 250 m resolution) and crop statistical data from 2000 to 2013 of different regions of the area, the spatiotemporal characteristics of the production areas of winter wheat and summer maize were estimated. A crop yield model was fitted using the least squares theory based on yield statistical data, accumulative annual NDVI and accumulative maxium NDVI for the period of 14 years. The crop yield estimation model used to estimate the potential yields of wheat and maize was a quadratic function. The results showed that: 1) the productivity of winter wheat was higher in Handan and Hengshui regions and lower in other regions of the study area. Even under improved production practices, productivity in the latter regions was always less than in the former regions. The productivity of maize was high in most of the study area even where potential productivity was difficult to attain. 2) The highest yield potential for both winter wheat and maize was in Handan region. The yield gap (the difference between potential yield and actual yield) for winter wheat was generally less than 10%, with an average of 356 kghm
-2 (5.87%). Then the yield gap for maize was generally more than 10%, with an average of 798 kghm
-2 (12.33%). Yield gaps were different for different regions in the study area. The ranked order for winter wheat in terms of yield gap by region was Langfang > Baoding > Cangzhou > Handan > Xingtai > Hengshui. Then that for maize was Xingtai > Handan > Baoding > Cangzhou > Hengshui > Langfang. 3) Based on the 14-year maximum accumulative NDVI, the maximum production gap (the difference of potential productivity and actual productivity) for both wheat and maize occurred in Cangzhou. In the low-to-medium yield farmlands in the Lowland Plains of Hebei, yield-increasing potential was 3.90 × 10
8 kg for winter wheat and 9.62 × 10
8 kg for maize. The total yield-increasing potential of both wheat and maize (13.52 × 10
8 kg) was approximately 1/5 of the theoretical production gap of the two crops. In general, yield gap and production gap were both lower for winter wheat than for summer maize. Thus maize was the most important crop in terms of increasing future grain production in the study area. The method used in this study was applicable to various other crops at different scales, whose results were useful for decision-making and management policies.