求土壤水力特征的一种迭代法

土壤水力性质包括土壤水分特征曲线(它表明了土壤的基质势h与土壤水分含量θ之间的关系)θ(H)、非饱和水力传导率K和扩散度D。由于土壤的这三种水力性质可以通过关系式K=Ddθ／dh联系起来,因此,它们当中只有两个是独立的。不论是以水分含量为因变量还是以土水势为因变量的Richards方程,其中的参数如非饱和水力传导率、容水度或扩散度都是基质势或水分含量的函数。在研究实际问题时,通常需要知道土壤基质势与土壤水分含量之间的转换关系,以便在不同情况下,采用不同形式的Richards方程,所以,确定土壤的水分含量与基质势之间的定量关系有着非常重要的意义。     

土壤盐分对土壤水分扩散率的影响

土壤水分扩散率是土壤水盐运动的重要参数之一。利用水平土柱吸渗法对不同含盐量土壤的水分扩散率与含水量之间的关系进行了测定,并建立了土壤水分扩散率、土壤含水量与Boltzmann参数间的定量关系。结果表明,不同含盐量土壤之间的水分扩散率存在明显差异,表现在相同土壤含水量情况下,土壤水分扩散率随土壤含盐量的增大而增大;土壤水分扩散率随土壤含水量增大而单调增大,且当含水量接近饱和时,土壤水分扩散率接近无穷,通过建立含水量与土壤水分扩散率的经验函数关系能较好地反映了土壤含水率与土壤水分扩散率间的关系。     

SWCC测定过程产生的容重变化对SWCC参数的影响

土壤水分特征曲线（SWCC）测定过程本身会引起很大的容重变化。随着容重的增大,土壤的持水特性也发生改变。为了明确这种容重变化对土壤水分特征曲线的影响,本研究以4种原状土壤为例,用离心机法测定了不同定容重下的土壤水分特征曲线,分析了容重变化对Brooks-Corey和van Genuchten模型参数的影响。结果表明：土壤的容重越大,相同吸力下所对应的有效饱和度就越大,水分特征曲线就越平缓。4种土壤的参数A和α均随容重的增大而减小,可用幂函数表示。黑垆土和土娄土的λ、n随容重的增加而增大,可用线性函数描述,有别于黄绵土和红壤。从总体上看,无论扰动与否,土壤水分特征曲线测定过程引起的容重变化对粉质粘壤土和粉质粘土的影响情况一致。与van Genuchten模型相比,Brooks-Corey模型参数与容重具有更强的关系性。本研究有益于土壤水分特征曲线测定过程中容重影响因素的修正。     

沙漠非饱和风沙土壤水分特征曲线预测的分形模型

应用VanGenuchten提出的土壤水分特征曲线公式,推导出了沙漠风沙非饱和土壤水分特征曲线的分形模型。通过对古尔班通古特沙漠地9种不同土壤样本利用中子水分仪和负压计实测的水分特征曲线资料反求得到相应的分形维数,分析了分形维数与土壤质地之间的关系,结果表明随着土壤质地从流动风沙土、半固定风沙土到固定风沙土的变化,其分形维数呈逐渐增大。此外,基于土壤颗粒的重量与粒径分布求出了古尔班通古特沙漠地风沙土壤粒径的分形维数。通过对土壤水分特征曲线的分形维数与土壤粒径的分形维数的对比,得知它们之间存在着良好的线性关系。根据此关系,利用易测得的土壤粒径分形维数结合所推导的分形模型,对土壤水分特征曲线进行了预测,模型的预测结果很好地吻合了实测的土壤水分特征曲线。这一结果对于实际工作中根据风沙土壤颗粒大小分布的分形维数来预测沙漠风沙土壤水分特征曲线具有一定的指导意义。     

Green-Ampt模型参数简化及与土壤物理参数的关系

简化模型表达形式从而减少参数个数,对于Green-Ampt入渗模型的实际应用具有重要的现实意义。该文通过推导湿润锋处平均基质吸力与Philip模型中土壤吸湿率关系基础上提出了简化的Green-Ampt入渗模型,基于新疆222兵团两块壤质土壤田块上土壤水分入渗试验资料,分析了Green-Ampt简化入渗模型参数与土壤物理参数之间的关系,建立了模型参数与土壤物理参数之间的定量经验转换函数。结果表明,入渗参数A(组合参数)与土壤初始含水率呈对数负相关,相关系数为0.77,A与土壤紧实度和黏粒含量均呈指数负相关,相关系数分别为0.70和0.74。饱和导水率Ks与土壤紧实度和黏粒呈指数负相关,相关系数分别为0.74和0.73。A和Ks与土壤初始含水率、土壤紧实度和黏粒含量呈高度和中度多元线性相关,相关系数分别为0.9和0.79。研究表明Green-Ampt简化入渗模型能够在一定精度下分析土壤入渗过程。     

经典BRUTSAERT模型应用于土壤水分检测的适用性探讨

BRUTSAERT理论模型是非饱和土壤介质中弹性波传播的理论模型，主要用于描述声波纵波声速与农业土壤饱和度之间的关系。为探究BRUTSAERT理论模型应用在土壤水分检测中的可行性，本文系统介绍了BRUTSAERT理论模型，分析了其在土壤含水量检测中的适用条件，并在经典BRUTSAERT模型中引入了适当的简化和假设，推导出在不同物理条件下各农业土壤的纵波声速值与土壤水分关系曲线，以及进行土壤声速测量的合适声频。结果表明，当声波发射器和拾音器之间的距离为0.2 m时，且声波频率范围固定在380～708 Hz之间，在任何土壤类型中BRUTSAERT理论模型都是有效的。然而，经典BRUTSAERT模型在推导过程中还存在的一些问题：1）对BRUTSAERT理论模型中关于复合流体体积模量模型的书写错误分析；2）经典BRUTSAERT模型中使用土壤固定泊松比对预测结果带来的影响探讨。该研究可为BRUTSAERT理论模型在土壤水分检测领域的应用研究提供参考。     

冬小麦田间墒情预报的经验模型

基于土壤水分变化率与贮水量成正比这一假定,得出了土壤水分的指数消退关系。在此基础上,建立了冬小麦生育期土壤墒情预报的经验递推模型,并对模型进行了检验,表明模型预报效果较好。该模型的特点是模型简单且参数较少;其主要局限性是模型中土壤水分消退系数地域、时域性较强     

钙结石含量对土壤水分蒸发影响的模拟试验

通过在土柱中人工模拟黄土高原北部含钙结石土壤,在土壤总水分一致的情况下研究了钙结石含量对土壤水分蒸发过程的影响,以期为黄土高原特定土壤类型中土壤水平衡的计算和模拟提供试验依据。研究结果表明：不同钙结石质量分数（钙结石质量/（钙结石质量+土壤质量））的土壤水分累积蒸发在最初的7 d内差别不大,随后表现出一定的差异;试验期间不同处理的蒸发率差异很小。土壤水分蒸发量随钙结石质量分数的增加而降低,当钙结石质量分数为0.5时,土壤水分蒸发降低8 mm,占到土壤总水分的10%。土壤水分蒸发与钙结石含量之间的负相关关系与钙结石含量增加所导致的土壤含水率降低有关。钙结石对土壤水分蒸发的作用效果与钙结石吸水性、钙结石含量以及水分在钙结石和土壤之间的分配有关。     

基于遗传算法的土壤水分运动参数识别

土壤水分运动参数的识别是研究土壤水分运动的基础。该文以反映土壤含水率实测值和计算值吻合程度的均方差最小为优化目标,以土壤导水率和扩散率经验参数上下限为约束条件,建立了土壤水分运动参数识别的优化计算模型。采用遗传算法和田间均质土壤一维非饱和运动数值计算相结合的方法,获得土壤导水率和扩散率经验参数最优值。经验证计算,土壤含水率实测值和计算值吻合程度较高,表明这一方法是可行的。     

土壤非饱和导水率温度效应研究

为计算黄土高原3种土壤温度下的非饱和土壤导水率，采用土壤水分动力学方法和数值模拟，利用室内试验分别对3种土壤不同温度下的土壤水分特征曲线，湿润峰下渗速率以及湿润峰湿度与湿润剖面平均湿度的关系进行了定量研究，得到出了以下结果，（1）建立了黄土高原3种土壤非饱和土壤导水率温度效应的定量模型，K(θ）＝△aθ^b＋KTte，栖模型为了解田间土壤水分的动态变化及评价土壤水分有效性提供了理论依据，（2）在已有模型的基础上，计算了不同含水量，温度升高1℃所引起非饱和土壤导水率和土壤有效水的净增加量，并推导出更为直观的温度对土壤水分传导有效性影响的定量模型，可以直接计算不同温度下非饱和土壤导水率的温度效应。     

基于单一和集合土壤转换函数模型对土壤含水量的模拟性能分析

土壤转换函数(Soil pedotransfer function,PTF)是一种高效获取土壤水力参数的方法。由于土壤具有很强的空间异质性,确定最优PTF模型成为模拟土壤含水量的关键。为此,以海河流域3个实验场地(密云站、大兴站、馆陶站)为研究区,采用7种常用的单一PTF模型预测土壤水力参数作为HYDRUS-1D的模型参数,求解Richards方程获得土壤含水量,并与实测土壤含水量进行比较,评价了常用单一PTF模型预测的土壤水力参数对土壤含水量的模拟性能。此外,采用3种方法构建集合PTF模型,评价了集合PTF模型对土壤含水量的模拟性能。结果表明：基于van Genuchten方程构建的单一PTF作为模型参数模拟土壤含水量的均方根误差最小;而其中Rosetta3模型表现更优。在集合PTF模型中,基于遗传算法加权法构建的模型表现最好。集合PTF模型预测土壤水力参数可以较好的捕捉多个单一PTF预测土壤水力参数的整体趋势,弥补单一PTF在某些情况下模拟误差较大的不足。     

Review of modelling crop growth, movement of water and chemicals in relation to topsoil and subsoil compaction

Soil compaction influences crop growth, movement of water and chemicals in numerous ways. Mathematical modelling contributes to better understanding of the complex and variable effects. This paper reviews models for simulating topsoil and subsoil compaction effects. The need for including both topsoil and subsoil compaction results from still increasing compactive effect of vehicular pressure which penetrates more and more into the subsoil and which is very persistent. The models vary widely in their conceptual approach, degree of complexity, input parameters and output presentation. Mechanistic and deterministic models were most frequently used. To characterise soil compactness, the models use bulk density and/or penetration resistance and water content data. In most models root growth is predicted as a function of mechanical impedance and water status of soil and crop yield—from interactions of soil water and plant transpiration and assimilation. Models for predicting movement of water and chemicals are based on the Darcy/Richards one-dimensional flow equation. The effect of soil compaction is considered by changing hydraulic conductivity, water retention and root growth. The models available allow assessment of the effects of topsoil and subsoil compaction on crop yield, vertical root distribution, chemical movement and soil erosion. The performance of some models was improved by considering macro-porosity and strength discontinuity (spatial and temporal variability of material parameters). Scarcity of experimental data on the heterogeneity is a constraint in modelling the effects of soil compaction. Suitability of most models was determined under given site conditions. Few of the models (i.e. SIBIL and SIMWASER) were found to be satisfactory in modelling the effect of soil compaction on soil water dynamics and crop growth under different climate and soil conditions.     

A physically based model of soil erosion during snow melting

A physically based model of snowmelt-induced soil erosion for a random point within a catchment is suggested. This model is based on the (1) the rill erosion model during snow melting, (2) the Mirtskhulava equation for the thawed soil describing the soil particle detachment by water flows, (3) the Kuznetsov equation for the critical flow velocity, (4) the modified Bagnold equation for the transport capacity of a water flow, and (5) general equations of hydraulics and channel flows. The equations obtained include three values, which depend on the hydrothermic conditions of soil thawing and runoff processes. These values are regarded as calibration parameters. The model was verified with the use of long-term data records obtained at runoff plots in Kursk (chernozem) and Orel (gray forest soil) oblasts. Proceeding from the calculations, the accuracy of soil loss assessment for a long period is about 14%. The proposed model agrees with the empirical dependences of soil loss on the slope length and steepness (for slopes of 6°). The model led us to conclude that conditions exist under which soil loss does not depend on the rill network.     

土壤导热率测定及其计算模型的对比分析

土壤导热率是重要的热参数之一,为了获得预测导热率的准确方法,该文对比分析了确定土壤导热率的热脉冲直接测定法和模型间接推求法。根据热脉冲原理在相同体积质量下,测定了不同质地和含水率土壤的导热率值。结果表明在相同含水率条件下,砂粒含量越高,土壤的导热率越大,土壤导热能力越强。利用Horton经验公式对实测值进行了拟合,结果显示Horton经验模型基本可以反映土壤导热率变化特征,并得到了Horton公式经验系数。利用实测值与Campbell模型计算值进行了比较,结果显示Campbell模型计算结果偏差较大,并对其进行了修正。并且用实测值与Johansen模型及其2种改进模型(Coté-Konrad模型和Lu-Ren模型)的计算值进行了对比分析,结果表明Johansen模型计算结果与实测值偏差较大,2种改进型模型的计算结果与实测值更接近。该研究表明土壤导热率可以利用土壤质地、含水率、孔隙度和体积质量进行计算,3种理论模型的计算值与实测值的相关系数均值分别为:0.643、0.937、0.943,推荐使用Coté-Konrad模型和Lu-Ren模型计算土壤导热率,Lu-Ren模型比Coté-Konrad模型的适用范围更广。     

Sensitivity assessment,adjustment, and comparison of mathematical models describing the migration of pesticides in soil using lysimetric data

The water block of physically founded models of different levels (chromatographic PEARL models and dual-porosity MACRO models) was parameterized using laboratory experimental data and tested using the results of studying the water regime of loamy soddy-podzolic soil in large lysimeters of the Experimental Soil Station of Moscow State University. The models were adapted using a stepwise approach, which involved the sequential assessment and adjustment of each submodel. The models unadjusted for the water block underestimated the lysimeter flow and overestimated the soil water content. The theoretical necessity of the model adjustment was explained by the different scales of the experimental objects (soil samples) and simulated phenomenon (soil profile). The adjustment of the models by selecting the most sensitive hydrophysical parameters of the soils (the approximation parameters of the soil water retention curve (SWRC)) gave good agreement between the predicted moisture profiles and their actual values. In distinction from the PEARL model, the MARCO model reliably described the migration of a pesticide through the soil profile, which confirmed the necessity of physically founded models accounting for the separation of preferential flows in the pore space for the prediction, analysis, optimization, and management of modern agricultural technologies.     

Simulating nitrogen dynamics in soils using a deterministic model

Abstract. LEACHN is a deterministic model for simulating nitrogen dynamics in soil. Transport processes are based upon numerical solutions to the Richards equation for water flow and the convection-dispersion equation for solute transport. Transformations of urea, ammonium, nitrate and three organic pools are included, and the influence of water content and temperature can be reflected. Lack of measured input data sometimes limits the more general use of models such as these. Approaches to estimating data values using soil survey information and a limited number of measured data are discussed. Simple model sensitivity studies and a limited number of field measurements can guide the choice of input data values and lead to simulations that reflect the main features of the field soil nitrogen regime. Such an approach provides initial values for a modelling exercise, and improves intuition regarding the relative importance of processes and interactions in the field nitrogen cycle.     

以水面蒸发量为参考推求土壤实际蒸发量的数学模型

为了准确估算土壤在实际条件下的蒸发量,该文以水面蒸发量为参考,结合能量平衡方程及微气象学方法,推导计算土壤实际蒸发量的数学模型.结果表明所建模型所需参数为水面及蒸发土壤表面的日最高温度和日平均温度、水面日蒸发量、风速等.模型的验证结果表明计算的土壤蒸发量与实测蒸发量比较吻合(R2=0.90).模型所引入的参考蒸发面使其避开了土壤蒸发复杂的物理本质,从而使得对土壤蒸发的计算变得简单易行.     

A semi‐analytical model for solute transport in layered dual‐porosity media

The vulnerability of groundwater from chemical leaching through soil is a concern at some locations. Because measurements are laborious, time‐consuming, and expensive, simulation models are frequently used to assess leaching risks. But the significance of simulated solute movement through a layered soil is questionable if vertical homogeneity of physical soil properties has been assumed. In the present study, a semi‐analytical model for solute leaching in soils is presented. The model is relatively simple, but it does account for soil layers having different physical properties. The model includes the mobile‐immobile model (MIM) to describe one‐dimensional (1‐D) nonequilibrium, transient solute transport under steady‐state flow conditions. The MIM is rewritten as a second‐order differential equation and solved by a numerical scheme. Differing from fully analytical or fully numerical solutions, the new approach solves the differential equation numerically with respect to time and analytically with respect to distance. Numerical experiments for a single layered soil profile show that the semi‐analytical solution (SA‐MIM) is numerically stable for a wide range of parameter values. The accuracy of SA‐MIM predictions is comparable to that of analytical solutions. Numerical experiments for a multilayered profile indicate that the model correctly predicts effluent curves from finite layered soil profiles under steady‐state flow conditions. The SA‐MIM simulations with typical parameter values suggest that neglecting vertical heterogeneity of flow paths in a layered soil can lead to inaccurate prediction of soil‐solute leaching. The quality of predictions is generally improved if parameter estimates for the different soil layers are considered. However, the mobile‐immobile‐parameter estimates obtained in a number of previous studies may not be transferable to a field situation that is characterized by a slow and steady flow of water. Further field experiments to determine mobile‐immobile parameters under such conditions are desirable.     

基于PEST的土壤-作物系统模型参数优化及灵敏度分析

农业生产管理系统模型输入参数多,参数率定过程十分耗时费力,大大限制了其推广应用。该研究以华北平原2 a的冬小麦-夏玉米田间试验观测数据为基础,使用PEST(parameter estimation)参数自动优化工具对土壤-作物-大气系统水热碳氮过程藕合模型(soil water heat carbon and nitrogen simulator,WHCNS)的土壤水力学参数、氮素转化参数和作物遗传参数进行自动寻优,同时计算分析模型参数的相对综合敏感度,并将优化结果与土壤实测水力学参数和试错法的模拟结果进行比较。参数敏感度分析结果表明,18个模型参数的相对综合敏感度较高,其中土壤水力学参数普遍具有较高的敏感度,以饱和含水率敏感度最高;作物参数中,作物生长发育总积温和最大比叶面积具有较高的综合敏感度;而氮素转化参数的敏感度远低于土壤水力学参数和作物参数。评价模型模拟效果的统计性指标(均方根误差、模型效率系数和一致性指数)表明,PEST法比实测水力学参数的模拟精度有所提高,其中土壤含水率、土壤硝态氮含量、作物产量和叶面积指数的均方根误差分别降低了61.8%、23.5%、73.6%和23.3%。同时PEST法比试错法对土壤水分和作物产量的模拟精度也有较大提高,但对土壤氮素和叶面积指数的模拟精度提高不明显。由于该方法大大节约了模型校准时间,在较短的时间内获得了明显高于试错法的模拟精度,因此PEST软件在WHCNS模型参数自动优化中是一个值得推广的工具。     

Comparison of the soil hydraulic conductivity predicted from its water retention expressed by the equation of Van Genuchten and different capillary models

Knowledge of water retention and conductivity is essential to study water transport in soil. Determination of the conductivity curve is difficult, and it is often predicted by application of a capillary model to the water retention relationship. Three expressions are predicted from the water retention described by the equation of Van Genuchten. Two expressions are obtained in combination with the capillary models of Mualem and Burdine . The third expression is obtained by a combination with the capillary model of Fatt & Dykstra . The three sets of soil properties were applied to clay in order to compute infiltration and infiltration rates according to the series solution of Philip . Comparison with this solution showed that the results of the first two combinations were severely under-estimated, while those of the third were satisfactory. Similar results were obtained for sand by comparison with experimental data. Because conductivity estimated from the capillary model of Fatt & Dykstra is complicated, it was expressed by a power equation, the exponent of which is obtained by applying the Fatt & Dykstra capillary model to the water retention curve expressed according to Brooks & Corey and having the same asymptotic behaviour as the Van Genuchten equation. Application of this procedure to fifty soils selected from a published database gave satisfactory results. It is concluded that the hydraulic conductivity of a soil can be predicted from its water retention as expressed by the equation of Van Genuchten subject to the condition of the capillary model of Fatt & Dykstra and as expressed by the equation of Brooks & Corey, for which the exponent is obtained according to the same capillary model.