摘要
为定量评估天津城市绿地土壤水分特征曲线模型的适应性,运用张力计法获得绿地土壤水分特征曲线,选取Brooks-Corey模型、Gardner模型、Campbell模型、Van Genuchten模型、Gardner-Russo模型拟合土壤水分特征曲线,利用改进的粒子群算法确定模型参数,使用剩余平方和对模型进行评价。结果表明:Van Genuchten模型较常用的Gardner模型更好地描述天津城市绿地土壤水分特征曲线,Gardner-Russo模型模拟效果最差。通过仿真实验证明改进的粒子群算法计算土壤水分特征曲线模型参数的精度优于非单纯形法、阻尼最小二乘法、混合遗传算法以及随机粒子群算法。
In order to quantificationally evaluate the adaptability of soil water characteristic curve of urban green space of Tianjin, the tensiometer was used to obtain the soil water characteristic curve of urban green space, and the Brooks -Corey model, Gardner model, Campbell model, Van Genuchten model and Gardner- Russo model were selected to fit the soil water characteristic curve. An improved particle swarm optimization was utilized to determine the parameters of five models, and they were evaluated by the residual sum of squares. The results in- dicated that the Van Genuchten model described the soil water characteristic curve of urban green space of Tian- jin better than that the Gardner model did, which was widely used in practice, and the fitting effect of Gardner - Russo model was the worst in the five models. The simulation experiment proved that the accuracy of improved particle swarm optimization in calculation parameters of soil water characteristic curve model was higher than the nonlinear simplex method, damped least square method, hybrid genetic algorithm and stochastic particle swarm optimization.
出处
《干旱区资源与环境》
CSSCI
CSCD
北大核心
2013年第8期115-119,共5页
Journal of Arid Land Resources and Environment
基金
国家水体污染控制与治理科技重大专项(2009ZX07314-002-01)资助