设施农业对农田排涝模数影响

EFFECTS OF FACILITY AGRICULTURE ON FARMLAND DRAINAGE MODULUS

关键词:  超渗—蓄满垂向耦合模型 Sobol'方法 敏感性分析 SCE-UA算法 排涝模数

Key words:  Excess infiltration-excess storage vertical coupled runoff model  Sobol' method  sensitivity analysis  SCE-UA algorithm  drainage modulus

摘要:

［目的］研究设施农业发展对农田设计排涝模数的影响。［方法］以天津市西青区一处封闭农田为对象，进行了农田排涝现场观测试验。在现场观测数据的基础上，构建了超渗—蓄满垂向耦合产流模型，用Sobol'法进行参数敏感性分析，甄别出模型的敏感性参数，并将SCE-UA算法与人工率定相结合对模型参数进行率定。利用水量平衡条件、地下水位相对变化量以及土壤含水量变化量对模型的模拟精度进行评价，并与单一超渗产流模型、单一蓄满产流模型进行对比，进一步验证所建模型的适用性。最后针对不同重现期下的设计暴雨，利用耦合产流模型推求产水量，进而给出相应设计排涝模数。［结果］模型各项评价指标的相对误差在20%之内，Nash系数在0.6以上，模拟精度高； 与单一超渗模型和蓄满模型相比，有效性分别提高了0.869和0.912，因此所建立的超渗—蓄满垂向耦合模型是合理有效的。在3年、5年、10年和20年一遇的重现期下，设计排涝模数与原排涝模数相比均有所增加，增大幅度分别为16.44%、8.63%、7.50%和3.77%。通过降雨趋势分析，基准期与评价期的降雨没有显著线性变化趋势，相较于降雨，设施农业占比增加对农田设计排涝模数有更大的影响。［结论］在设施农业迅速发展背景下，重新修订农田小区排涝模数是必要的，对有效控制农田洪涝灾害有重要现实意义。

Abstract:

In order to study the effect of facility agricultural development on the design drainage modulus in farmland areas, taking a closed farmland area in Xiqing district of Tianjin city as an example, a field observation experiment of farmland drainage was carried out. On the basis of the field drainage observation data, an infiltration excess-saturation excess vertical coupled runoff model was established. Sobol' method was used to analyze the sensitivity of the model parameters and identify the sensitive parameters of the model. The Shuffled Complex Evolution Algorithm (SCE-UA) and manual calibration methods were applied to optimize model parameters. The water balance conditions, the change of the underground water level and soil moisture content were adopted to evaluate the simulation accuracy of the proposedly coupled model. In addition, the model was compared with the single infiltration and saturation excess runoff model. Finally, according to the design storm of different return periods, the runoff yield was calculated by using the vertical coupled runoff model, and then the design drainage modulus under different return periods was given. The results showed that the relative errors of each evaluation index of the model accuracy were within 20%, and the Nash efficiency coefficient was above 0.6, which showed the proposed model had a high simulation accuracy. Compared with the single infiltration excess runoff model and the single saturation excess runoff model, the effectiveness of the vertical coupled runoff model was higher, with the increase of 0.869 and 0.912, respectively. The infiltration excess vertical coupled runoff model was reasonable and effective in this farmland area. For the return periods of 3 years, 5 years, 10 years and 20 years, the design drainage modulus increased by 16.44%, 8.63%, 7.50% and 3.77%, respectively, compared with the original drainage modulus. And through the rainfall trend analysis, there were no significant linear change trend during the baseline period and the evaluation period. Therefore, the increase of the proportion of facility agriculture had greater impact on the design drainage modulus. With the rapid development of facility agriculture, it was necessary to revise the drainage handbook in farmland areas, which was of great practical significance for effective control of flood and waterlogging in farmland.