基于作物生长模型参数调整动态估测夏玉米生物量

针对如何利用作物生长模型定量解析区域夏玉米生物量动态变化的热点问题,该文在沿东海岸的江苏省盐城市大丰区设置大田夏玉米生物量估测试验,在构建夏玉米生物量过程模拟模型的基础上,对夏玉米多个生育阶段的生物量(指地上部生物量)及其变化特征进行分析,并结合试验实测数据探讨利用实测叶面积指数和生物量数据调整生物量模拟模型参数的可行性。结果表明:夏玉米生物量过程模拟模型可以对夏玉米从出苗到灌浆期间的多个生育阶段生物量动态变化进行估测。出苗到拔节前的生长阶段,生物量积累主要来源于叶片形成,模拟模型可以对生物量进行有效预测,预测值与实测值之间的均方根差(root mean square error,RMSE)为18.31 kg/hm~2,相对误差为3.35%。拔节到抽雄前的生长阶段,由于茎节伸长与节数增加,生物量积累加快,预测值与实测值之间的差异较大。拔节初期生物量预测值为535.5 kg/hm~2,实测值为480 kg/hm~2,相对误差11.56%。抽雄前生物量预测值为7 036.46 kg/hm~2,实测值为5 794 kg/hm~2,相对误差21.44%。拔节到抽雄前生长阶段预测值与实测值之间的RMSE为825.94 kg/hm~2。经过模型参数调整,抽雄前生物量预测值为6 036 kg/hm~2,与实测值较为接近,RMSE为219.43 kg/hm~2,相对误差4.18%。利用参数调整后的模拟模型继续对抽雄到灌浆前生长期间生物量进行预测,预测值与实测值较为一致,灌浆期生物量预测值为11 156 kg/hm~2,实测值为10 785 kg/hm~2,相对误差3.44%,而参数调整前预测值为12 492 kg/hm~2,相对误差15.83%。在玉米拔节期进行模型参数调整,对拔节到抽雄和抽雄到灌浆2生长阶段的生物量预测效果较好。该研究可为县域夏玉米不同生长阶段生物量及其动态变化预测提供参考,可辅助县域农业管理部门进行适时生产措施调整。

Dynamic estimation of summer maize biomass based on parameter adjustment of crop growth model

Biomass is one of the important growth indicators of summer maize (Zea mays). Timely and accurate acquisition of field summer maize biomass data and its dynamic changes is conducive to county agricultural management departments to rationally adjust field production management measures, which is of great significance for increasing maize yield based on the establishment of a simulation model of summer maize biomass process.In this paper, we analyzed above ground biomass and its variation characteristics of summer maize at different growth stages and discussed the feasibility of adjusting parameters of biomass simulation modelusing measured leaf area index and biomass datain Dafengdistrict ofYanchengcity, Jiangsu province, along the east coast of China. The simulation model of biomass process of summer maize could estimate the dynamic changes of aboveground biomass of summer maize at different growth stages from emergence to grain filling.Firstly, the simulation model can predict above ground biomass effectively because of biomass accumulation mainly coming from leaf formation from seedling emergence to pre-jointing growth stage, the root mean square difference (RMSE) of which was 18.31 kg/hm2. During the growth stage from jointing to tasseling stage, the biomass accumulation accelerate due to the elongation of stem nodes and the increase of number of nodes, and the difference between the predicted and measured values waslarge and RMSE was 825.94 kg/hm2. For example, the predicted value of biomass at early jointing stage was 535.5 kg/hm2, and the measured value was 480 kg/hm2, with a difference of 11.56%. Another example, the predicted value of biomass beforetasseling stage was 7036.46 kg/hm2, and the measured value was 5 794 kg/hm2, the difference was 21.44%. Then, adjusting the parameters of the model, the predicted biomasswas 6 036 kg/hm2, which was close to the measured biomasswith RMSE 219.43 kg/hm2. Finally, by using the simulation model adjusted by the parameters, we predictedbiomass during the growth period from tasseling to grain filling that predicted values were in good agreement with the measured values, and the determinant coefficient between them was 0.978 and RMSE was 182.95 kg/hm2. For example, the predicted biomass at grain filling stage before parameter adjustment was12 492 kg/hm2, the measured biomass was 10 785 kg/hm2, the relative error was15.83%, and the predicted biomass after parameter adjustment was 11 156 kg/hm2with the relative error of 3.44%.This study provided an effective informatics method for forecasting the aboveground biomass and its dynamic changes of summer maize at different growth stages, and could assist the county agricultural management departments to adjust production measurestimely. Combining remote sensing data with crop process simulation model to estimate regional crop biomass or yield was a hot topic in the remote sensing application in agriculture. In this paper, crop process simulation model was used to analyze the trend of biomass change at three growth stages of summer maize, namely, emergence to jointing, jointing to tasseling, tasseling to filling, and to clarify the law of biomass accumulation and nutrient uptake characteristics at corresponding growth stages. In the follow-up study, LAI and biomass inversion using remote sensing data will be further considered, and assimilation and application of remote sensing data and crop process simulation model will be studied to enhance the universality and effectiveness of maize biomass process simulation model in maize planting areas along the East coast in China.