基于SWAP-IES作物同化模型的宁夏玉米产量模拟

Simulation of Maize Yield in Ningxia Based on SWAP-IES Crop Assimilation Model

作物长势和田间水分预测对于农业精准管理至关重要。为准确模拟宁夏玉米产量,利用2019-2020年田间观测数据,整合SWAP(soil-water-atmosphere-plant)模型和迭代集成平滑算法(iterative ensemble smoother,IES)构建了适用于宁夏干旱地区玉米的SWAP-IES作物同化模型。比较了同化叶面积指数(leaf area index,LAI)、土壤含水率(soil water content,SW)及其组合对宁夏干旱地区玉米种植区土壤含水率模拟和产量估算的影响。研究结果表明,当同时同化LAI和SW数据时,土壤含水率模拟的决定系数(R2)显著提升,从初始时的-0.07增加到0.71。这表明,将LAI和SW数据同时纳入模型显著增强了模型预测土壤含水率的准确性。而同时同化LAI和SW相比仅同化LAI或SW能更好的模拟土壤含水率,这表明2个观测变量之间并不是孤立的,二者的耦合能更好地提升模型的模拟精度。同时同化LAI和SW时估产精度最高,RMSE降低到914.113 kg/hm^(2),显著低于其他情景。说明所构建的SWAP-IES玉米同化模型,在同时同化LAI和SW的情况下,可以准确模拟土壤含水率变化过程和玉米产量,为干旱地区农田灌溉优化和玉米估产提供参考。

Crop growth and field moisture prediction are essential for accurate agricultural management.In order to accurately simulate the maize yield in Ningxia,based on the field observation data from 2019 to 2020,the Soil-Water-Atmosphere-Plant(SWAP)model and iterative ensemble smoother(IES)algorithm were used to construct a SWAP-IES assimilation modeling system for maize in Ningxia.The effects of assimilating leaf area index(LAI),soil water content(SW)and their combination on soil moisture content simulation and yield estimation in maize fields in Ningxia were compared.The results showed that the R2 of soil moisture content simulation increased from-0.07 to 0.71 without assimilation when assimilating LAI and SW at the same time,indicating that the simulation accuracy of soil moisture content of the model could be greatly improved by assimilating LAI and SW at the same time.However,the simultaneous assimilation of LAI and SW can better simulate the soil moisture content than the assimilation of LAI or SW alone,which indicates that the two assimilation variables are not isolated,and the coupling of the two can better improve the prediction accuracy of the model.The yield estimation accuracy was the highest when assimilating LAI and SW at the same time,and the RMSE decreased to 914.113 kg/hm^(2),which was significantly lower than that in other scenarios.The results show that the constructed SWAP-IES maize assimilation model can accurately simulate the process of soil moisture change and crop yield under the condition of simultaneous assimilation of LAI and SW,and provides a reference for optimizing farmland irrigation and improving crop yield estimation.