Stochastic simulation of soil water dynamics and irrigation water requirement in farmland

【作者】 黄仲冬；

【导师】 齐学斌；

【作者基本信息】 中国农业科学院 ， 农业水土工程， 2016， 博士

【摘要】 土壤水分和灌溉需水量是农田生态系统水循环过程和农业水资源管理的核心与基础,准确认知土壤水分动态过程和灌溉需水量变化规律并建立估算模型具有极其重要的现实意义和科学价值。土壤水分和灌溉需水量是大气、土壤、作物、地下水、灌溉等因子长期综合作用的结果,其动态过程具有强烈的非线性和不确定性,因此,有必要采用概率统计的思想进行研究。本文以华北南部山前平原夏玉米为研究对象,以土壤水分动态过程为核心,采用水量平衡原理、概率统计和微积分理论对农田生态系统的水分通量、灌溉需水量的变化特征及规律进行了系统的理论分析和定量模拟,主要结论如下:(1)系统分析了夏玉米生长季水文气象因子在年尺度上的动态变化特征和随机分布规律,探讨了土壤水分动态与气候波动的响应关系。研究结果表明,平均降水量(α)和降水频率(λ)年变异系数分别为0.323和0.178,年变化服从Logistic分布;日平均潜在蒸散量(Emax)年变异系数为0.088,年变化服从为Log Normal分布;土壤水分概率密度函数曲线在α、λ、Emax随机波动影响下呈现双峰现象,表明处于随机波动条件下的夏玉米SPAC系统可能会趋向于两个比较适宜的水分状态,一个为土壤相对湿润的状态,另一个为土壤相对干燥的状态。(2)基于Budyko水热耦合平衡假设和土壤水分动态随机模型定量分析了α、λ和Emax长期随机波动对夏玉米蒸散量(E)的影响。研究结果表明,α、λ和Emax长期随机波动导致夏玉米生长季的蒸散量减少,当干旱指数(Ep/P,Ep和P分别为潜在蒸散量和降水量)等于1时,蒸散量的减少量最大,蒸发比(E/P)的减少比率接近10\%;E/P的最大减少比率分别与α、λ和Emax的变异系数呈线性递增关系;E的年变异性随着α、λ和Emax年变异性的增加而增加,且α对E年变异性的影响显著大于λ和Emax。研究区夏玉米生长季E、P和Ep的定量关系可采用基于Budyko水热耦合平衡假设的傅抱璞方程进行描述,方程的参数ω取值为5.3。(3)构建了具有一定普适性的农田土壤水分动态随机模型。通过引入土壤饱和导水率(Ks)和水力参数(β),建立了相对完整的土壤水分损失函数表达式,根据土壤水分概率密度函数与土壤水分损失函数的解析关系,推导出了不同灌水控制条件下的土壤水分概率密度函数表达式。(4)改进了灌溉需水量估算方法。基于土壤水分动态随机模型,建立了具有物理基础的灌溉需水量计算模型,模型综合反映了灌溉需水量与降水、土壤、作物、灌溉等因子的定量关系,克服现有经验模型的不足,为变化环境下适宜灌溉方式的选取提供了理论依据。(5)构建了潜水蒸发作用下的农田土壤水分动态随机模型。在分析了潜水蒸发与潜水埋深和土壤相对含水率定量关系的基础上,建立了潜水蒸发作用下的土壤水分损失函数,并根据土壤水分概率密度函数与土壤水分损失函数的解析关系,推导出了土壤水分概率密度函数的解析表达式,最终形成综合考虑土壤、作物、大气、地下水和灌溉等因素的农田土壤水分动态随机模型。

【Abstract】 Hydrologic cycle is the fundamental process in agroecosystem, and it is necessary and valuable to understand and simulate the water balance in cropland. Soil moisture and irrigation requirement, the foundation and key varaiable of cropland water cycle, play an extremely important role in crop growth and development. Meanwhile, it is a significant parameter for improving water use efficiency, water-saving irrigation management, and rational allocation of water resources. Water balance dynamics of cropland, taking the core as soil moisture, are the long-term result of various factor comprehensive roles, including atmosphere, soil, crop, groundwater, irrigation activity. Therefore, they manifest strong nonlinearity and uncertainty. Accordingly, it is essential to interpret and model the dynamics of soil moisture and hydrological elements, providing scientific basis for the evolution analysis of hydrologic cycle and irrigation decision in agroecosystem.This thesis mainly investigated the hydrologic cycle and transformation of summer maize, taking soil moisture dynamics as core, in south of North China Plain, using water balance principle, probability statistics and calculus. The main results are outlined as follow:(1) The annual fluctuations and random distributions of hydrometeorological factors were analyzed during the growing season of summer maize, and the response relationship between soil moisture and climate fluctuations was investigated. The annual coefficient of variation of mean depth of rainfall events(α) and arrival rate of rainfall events(λ) were 0.323 and 0.178 respectively. Moreover, α and λ were testified following a Logistic distribution. The annual coefficient of variation of daily average potential evapotranspiration(Emax) was 0.088, while following a Log Normal distribution. The probability density function of soil moisture emerged a bimodal curve, which suggested that the SPAC system of summer maize trend to switch between two preferential states, one characterized by high average soil moisture and the other characterized by low average soil moisture conditions.(2) The long-term effect of α, λ and Emax variations on evapotranspiration(E) of summer maize was studied, based on Budyko framework of water and energy balance and stochastic soil moisture dynamics. Annual long-term fluctuations of α, λ and Emax reduces the long-term evapotranspiration. This reduction is the maximum when the dryness index(Ep/P) equals 1, and maximum reduction in the evaporation ratio(E/P) can reach about 10\%. The relations between the maximum reductions and the coefficient of variation of α, λ 和 Emax follow linear law. The interannual variability of E increases as the increasing of interannual variability of α, λ and Emax, and α is more significant than λ and Emax. The relationship of E, P and Ep during the growing season of summer maize in the study can described by the Fu’s equation with ω of 5.3.(3) A certain universal model of stochastic soil moisture dynamics was built, aiming at the shortcoming of previous researches. Soil saturated hydraulic conductivity(Ks) and a hydraulic parameter(β) were introduced to establish a comparatively complete soil moisture loss function. Afterwards, the probability density functions of soil moisture with different irrigation schemes were derived according to the analytic relation between loss function and probability density function of soil moisture.(4) A novel estimating method of crop irrigation requirement was developed. The model for calculating irrigation requirement is physic-based on stochastic soil moisture dynamics model, which reflect the quantitive relationship among rainfall parameters(α, λ), soil parameters(n, s*, sfc, Ks, β), crop parameters(Emax, Zr) and irrigation parameters(si、st). The new physic-based method is better than empirical method. The irrigation requirement model coupled with water production function provides a theoretical basis to the selection of appropriate irrigation scheme in a changing environment.(5) The capillary flux on the soil moisture and groundwater depth was investigated, and a stochastic model of water transform in soil-plant-atmosphere-groundwater continuum for irrigated cropland was built, offering theoretical basis for studying hydrologic cycle and irrigation management with a shallow groundwater depth.

【关键词】 土壤水分； 水量平衡； 随机模型； 灌溉需水量； 潜水蒸发；
【Key words】 soil moisture； water balance； stochastic model； irrigation requirement； capillary flux；