基于动态规划理论的棉花根长生长模拟方法

作物根系生长不仅取决于生理因素,还取决于生态环境因素,而土壤水分环境与作物根系生长之间的关系则是局部灌溉技术设计的理论依据之一。为了进一步探索影响棉花根系生长的主要因素,本文在根?冠水量平衡的基础上,结合作物系数与叶面积的关系模型、根长密度分布函数以及根系吸水效率函数,应用动态规划理论,建立了棉花根系生长模型,并以桶栽棉花试验结果进行验证。结果表明,该模型纳入了土壤水分环境、大气蒸腾力和叶面积等影响根系生长的因子,具有揭示根系生长耗水机理的作用。该模型模拟出的棉花总根长变化趋势与实测结果基本一致,当以多年月平均参考蒸散量(ET0)作为输入条件时,模拟结果总体误差为15.41%,可以用于工程设计。对模型敏感性分析结果表明,所建模型能够反映棉花根?叶生长的同步性,以及进入生殖生长期以后根—叶之间的水量平衡关系。棉花根系生长对土壤水分环境变化的敏感性高于对叶面积变化的敏感性,体现了棉花根系生长的机理,建模方法可行。本文研究成果对完善局部灌溉技术中灌溉制度的设计理论具有重要意义。

Simulation method of cotton root length growth based on dynamic programming theory

Crop root length and soil moisture distribution are important determinants of crop root water uptake potential. Crop root length changes with growth stage, which also requires changes in soil moisture environment. Therefore, establishing a root growth model to predict crop root growth conditions under normal water use has theoretical and application significance regarding the determination of irrigation quota and soil moisture environment indicators, which are required to design appropriate local irrigation technology. However, current root growth simulating models are more suitable for the determination of irrigation quotas of whole irrigation technology. These models, which are mostly statistical, can not completely reflect the relational mechanisms of the growth of plant root system, crop water use and soil moisture environment. In view of the above and, a root-canopy water balance combined with crop coefficient vs. leaf area relationship model, root length density distribution function and root water uptake efficiency function, a cotton root growth model was developed based on dynamic programming theory and verified by experimental results of a barrel-cultivated cotton. The main results of the study showed that the model well accounted for the effects of root growth factors such as soil moisture environment, atmospheric transpiration rate and leaf area, which revealed to a certain extent the mechanism of crop water use due to root growth. The growth characteristics of cotton root length simulated by the model were consistent with measured dates in the barrel experiment. When multi-year average monthly mean reference evapotranspiration (ET0) was used as input condition, the overall error of the simulation result was 15.41%. Therefore the model was applicable in engineering designs. Based on sensitivity analysis, the established model well reflected the synchronization between the growth of cotton root and leaf, as well as the water balance between root and leaf after entering the reproductive period. The sensitivity of cotton root growth to changes in soil moisture environment was higher than to changes in leaf area, reflecting the processes of cotton root growth and the feasibility of the modeling method. The research significantly improved the design theory of irrigation systems for the development of localized irrigation technology.