基于BP人工神经网络的长江河口地区土壤盐分动态模拟及预测

为开展长江河口地区土壤盐分动态的中长期模拟与预测,采用人工神经网络中应用较为成熟和广泛的BP网络建立长江河口地区土壤盐分与降雨量、蒸发量、长江水电导率、内河水电导率、地下水位、地下水电导率6因子间的非线性神经网络响应模型。网络模型结构为6-11-1,隐含层单元数用"试错法"确定。选择合适的参数训练和学习网络模型后,对河口地区2003年各月平均根层土壤电导率进行预测,并与线性回归模型预测结果进行比较。结果表明:BP网络模型较线性回归模型具有更高的预测精度,平均相对预测误差为7.3%,预测值与实测值相关性良好,可以满足实际应用需求。     

轻质土壤水分特征曲线估计的简便方法

以黄淮海平原封丘地区的潮土和风沙土为研究对象,根据大量的土壤基本物理性质和土壤持水数据,利用多元逐步回归分析方法,建立了轻质土壤在不同基质势下土壤含水量(θ)的传递函数模型,并进行了模型验证。结果表明,利用轻质土壤的基本物理性质估计其水分特征曲线是一种简便可行的方法,并且,在回归方程中,增加-30kPa含水量项可提高-30kPa以上土壤含水量的估计精度;增加-1500kPa含水量项可以明显提高-100kPa至-1500kPa间土壤含水量的估计精度。     

基于遥感与电磁感应仪数据的土壤盐分空间变异性

针对目前黄淮海平原存在的土壤盐渍化问题,以河南省封丘县为研究区,利用遥感影像与磁感应电导率仪 （EM38）在田间定位、定点测量,对研究区土壤盐分的空间变异性进行研究。利用Kriging、指数回归和回归-Kriging 3种不同的方法对土壤盐分空间变异性进行分析。结果表明：土壤电导率在一定的区域范围内具有空间结构特征,符合指数模型分布,具有极强的空间自相关性。通过3种不同的方法对土壤盐分的空间变异性进行分析,结果表明Kriging法、指数回归法及回归-Kriging法在描述土壤盐分空间分布的总体趋势上相似,但指数回归法与回归-Kriging法对于盐分的分布表达更具体,更细致。综合运用多种数据源的方法,即遥感影像数据与EM38测量数据相结合,经典统计学与地统计学方法相结合的方法,能够提高土壤盐分空间预测的精度,结果表明研究区土壤盐分由北向南、自西向东有逐渐升高的趋势,具有明显的趋势效应。     

关中平原农田土壤水力参数空间分异与模拟

土壤水力参数是土壤水分和污染物迁移等陆面过程数值模拟的重要基础参数。为探明关中平原农田土壤水力参数空间分异特征,建立空间分布预测模型,在关中平原网格布设124个样点,采集根层0-20 cm原状和扰动土壤样品,利用van Genuchten模型拟合获取土壤水分特征曲线,获得残余含水量(θr)、饱和含水量(θs)以及系数α和n等土壤水力参数。采用经典统计学、地统计学和结构方程方法分析了θr、θs、α和n的空间变异特征及影响因子,建立了水力参数传递函数预测模型。结果表明:θr和α为强变异,θs为中等变异,n为弱变异。θr、θs、α和n半方差函数最佳拟合模型分别为球状模型、指数模型、指数模型和球状模型。θs和n具有强烈的空间依赖性,变程分别为32.7,54.3 km;θr和α具有中等程度空间依赖性,变程均为52.8 km。土壤质地、容重、pH、有机质和海拔是影响土壤水力参数空间分布的主要因子。基于土壤理化性质和海拔建立的水力参数传递函数模型具有较好的模拟效果,可用于关中地区大尺度农田生态系统土壤水力参数的模拟预测。     

应用土壤质地预测干旱区葡萄园土壤饱和导水率空间分布

田间表层土壤饱和导水率的空间变异性是影响灌溉水分入渗和土壤水分再分布的主要因素之一,研究土壤饱和导水率的空间变化规律,有助于定量估计土壤水分的空间分布和设计农田的精准灌溉管理制度。为了探究应用其他土壤性质如质地、容重、有机质预测土壤饱和导水率空间分布的可行性,试验在7.6 hm2的葡萄园内,采用均匀网格25 m×25 m与随机取样相结合的方式,测定了表层(0～10 cm)土壤饱和导水率、粘粒、粉粒、砂粒、容重和有机质含量,借助经典统计学和地统计学,分析了表层土壤饱和导水率的空间分布规律、与土壤属性的空间相关性,并对普通克里格法、回归法和回归克里格法预测土壤饱和导水率空间分布的结果进行了对比。结果表明:1)土壤饱和导水率具有较强的变异性,平均值为1.64 cm/d,变异系数为1.17;2)表层土壤饱和导水率60%的空间变化是由随机性或小于取样尺度的空间变异造成;3)土壤饱和导水率与粘粒、粉粒、砂粒和有机质含量具有一定空间相关性,而与土壤容重几乎没有空间相关性;4)在中值区以土壤属性辅助的回归克里格法对土壤饱和导水率的预测精度较好,在低值和高值区其与普通克里格法表现类似。研究结果将为更好地描述土壤饱和导水率空间变异结构及更准确地预测其空间分布提供参考。     

基于局部加权回归的土壤全氮含量可见-近红外光谱反演

全氮是土壤肥力的重要指标,对作物产量具有决定性作用,采用土壤可见-近红外(Vis-NIR)光谱预测技术及时获取土壤全氮含量信息具有重要意义。采用来自5省的450个土壤样本来验证局部加权回归方法(LWR)结合Vis-NIR光谱技术预测大面积土壤全氮含量的适用性。结果表明,LWR模型的预测效果优于偏最小二乘回归(PLSR)、人工神经网络(ANN)和支持向量机(SVM),选取主成分数为5,相似样本为40时,模型验证的决定系数(RP2)为0.83,均方根误差(RMSEP)为0.25 g kg-1,测定值标准偏差与标准预测误差的比值(RPD)达到2.41。LWR从建模集中选取与验证样本相似的土样作为局部建模样本,降低了差别大的样本对模型的干扰,从而提高了模型的预测能力。因此,LWR建模方法通过大范围、大样本土壤光谱数据进行大尺度区域的全氮等土壤属性预测时能够发挥更好的作用。     

闽南山地巨尾桉生长的土壤主导因子调控模型研究

根据巨尾桉林分生长及林下土壤调查资料 ,采用人工神经网络方法 ,探讨影响闽南山地巨尾桉生长的土壤主导因子 ,建立了巨尾桉生长与土壤主导因子关系的BP调控模型。结果表明 :限制闽南山地巨尾桉速生丰产的土壤主导因子为土壤毛管持水量、腐殖质层厚度、土层厚度及有效磷 ,其与巨尾桉生长形成是一种非线性映射关系 ,所建立的BP模型对模拟预测巨尾桉生长具有较高的精度 ,平均模拟精度达 96.39% ,效果比传统线性回归模型来的更理想。这不仅为提高巨尾桉人工林生产力 ,培育速生丰产林提供科学依据 ,而且也为人工神经网络在选择适宜造林立地条件研究中的应用开辟新的思路     

土壤持水曲线van Genuchten模型求参的Matlab实现

土壤持水曲线是研究土壤水力学性质必不可少的 ,在已经建立的众多数学模型中 ,vanGenuchten模型是目前运用最广泛的模型 ,而运用该模型的关键是其 4个参数的求解。为此 ,本文对同一组东北褐土的土壤水吸力和对应的土壤含水量数据较详细地介绍了Matlab软件的非线性拟合和非线性回归函数的运用 ,得出了该土壤vanGenuchten模型的 4个参数值 ,分别建立了该土壤的vanGenuchten模型 ,并利用Mat lab强大的绘图功能对它们进行了直观比较。最后运用方差分析和残差分析对该模型的计算值与实测数据进行了分析 ,结果表明 :非线性拟合和非线性回归函数求参结果的显著水平均达到p<0 0 0 0 1,残差平方和均小于 0 0 0 0 5 ,其中非线性回归函数的求参结果较非线性拟合好。因此 ,运用Matlab软件的非线性拟合和非线性回归函数对土壤持水曲线的vanGenuchten模型进行求参是切实可行的 ,从而为土壤学工作者寻求出了一条运用数值计算方法的新途径。     

基于回归克里格和遥感的紫色土区土壤有机质含量空间预测

精确预测紫色土区土壤有机质含量的空间分布,对于指导紫色土区农业生产和培肥土壤具有重要意义。以杜家沟小流域为研究区,以遥感影像作辅助变量,采用回归克里格法,预测土壤有机质含量的空间分布,并与参照方法的预测精度进行比较。结果表明:(1)Landsat ETM+影像的波段2和波段5是土壤有机质含量多元线性回归预测的最佳辅助变量,回归残差的最优半方差函数模型为球状模型,模型的拟合精度较高;(2)土壤有机质含量呈由沟谷逐渐向坡顶递减的趋势,空间变异的细节信息表达较好;(3)回归克里格法在验证点的预测值与实测值的拟合能力更好,预测结果更倾向于无偏的,MAE、RMSE和R2均优于参照方法。因此,回归克里格法是紫色土区土壤有机质含量高精度空间预测的有效方法。     

应用基于PLSR的土壤-环境模型预测土壤属性

土壤-环境模型对于正确理解土壤属性与环境因子间的关系,以及进行土壤属性预测与制图均具有重要的意义。研究区位于陕西省长武县内多年退耕还林还草沟壑区域,采集72个土壤表层样本,选择3/4的样本作为建模集,其余1/4的样本作为验证集;环境因子选择容易获取的地形因子和由遥感影像提取的植被因子和湿度因子,建立基于偏最小二乘回归(PLSR)的土壤-环境模型。结果表明:全氮、速效钾、全钾、有机质与环境因子间均有显著相关性;建立的PLSR模型可解释土壤属性的空间变异从23%(全氮)到27%(全钾);与逐步回归方法构建的模型相比,利用PLSR构建的土壤-环境模型可以更好地表征土壤属性与环境变量间的关系,拟合精度和预测精度也相对较高,说明PLSR建立的模型可以更好地应用于相似区域的土壤属性预测。     

基于土壤传递函数的土壤持水特性间接估计及其空间变异性研究

A total of 107 soil samples were taken from the city of Qingdao,Shandong Province,China.Soil water retention data at 2.5,6,10,33,100,300,and 1 500 kPa matric potentials were measured using a pressure membrane apparatus.Multiple linear regression (MLR) was used to develop pedotransfer functions (PTFs) for single point estimation and van Genuchten parameter estimation based on readily measurable soil properties,i.e.,MLR-based point (MLRP) PTF and MLR-based parametric (MLRV) PTF.The double cross-validation method was used to evaluate the accuracy of PTF estimates and the stability of the PTFs developed in this study.The performance of MLRP and MLRV PTFs in estimating water contents at matric potentials of 10,33,and 1 500 kPa was compared with that of two existing PTFs,the Rawls PTF and the Vereecken PTF.In addition,geostatistical analyses were conducted to assess the capabilities of these PTFs in describing the spatial variability of soil water retention characteristics.Results showed that among all PTFs only the Vereecken PTF failed to accurately estimate water retention characteristics.Although the MLRP PTF can be used to predict retention characteristics through traditional statistical analyses,it failed to describe the spatial variability of soil water retention characteristics.Although the MLRV and Rawls PTFs failed to describe the spatial variability of water contents at a matric potential of 10 kPa,they can be used to quantify the spatial variability of water contents at matric potentials of 33 and 1 500 kPa.     

非饱和水分运动参数空间变异的研究

田块尺度土壤特性的空间变异性对水分与溶质运移具有明显的影响。该研究在野外30m×30m面积、土壤类型为砂壤土的田块的100个空间点上,分别利用张力计和取土样室内测定的方法测定了30cm土层深处土壤水张力、土壤容重ρ、饱和含水率θ_s,与初始含水率θ_i,同时利用圭尔夫仪,测定了该田块同样深度108个空间点上的饱和水力传导度K_s与孔隙大小分布参数α。利用经典统计分析与地质统计分析方法分析上述参数的空间变异特征,研究结果表明:ρ,θ_i,θ_s,K_s和容水度C遵从正态分布,而α具有对数正态分布;K_s,α和C具有较大的空间变异性,而ρ和θ_s的空间变异性则较小;K_s和logα是空间统计相关的;土壤水张力的空间变异具有时不变特征,且土壤水张力方差是其均值的二次函数     

不同盆栽基质水分特征曲线的对比分析与模拟

现有关于盆栽控水模拟土壤干旱条件的试验中多采用含水率作为水分胁迫阈值,然而由于基质配比不同导致含水率相同的基质的水分状况也不尽相同,这导致各研究间结果难以对比和参考。为快速获取盆栽基质水分特征曲线,建立基质水分特征曲线预测模型。该研究以盆栽控水试验常用的泥炭土、蛭石和珍珠岩为基质材料,测定了不同配比基质的水分特征曲线,通过不同方法（多元回归模型、人工神经网络）建立了其预测模型。结果表明,人工神经网络模型对基质水分特征曲线的预测精度高于多元回归;相较于人工神经网络,多元回归模型的稳定性更高。综合考虑模型的精度和稳定性,多元回归模型是预测作物盆栽基质水分特征曲线的最佳模型。该模型为基质水分特征曲线快速获取以及相关作物干旱胁迫研究间的对比提供了方法和依据。     

Impact of Fly Ash Amendment and Incorporation Method on Hydraulic Properties of a Sandy Soil

Coal fly ash has physical and chemical characteristics that makeit useful as a soil amendment, one of the more important beingthe potential to permanently improve the soil water relations ofsandy, drought-prone soils. We axemined changes in theinfiltration rate and water holding capacity of a sandy soilafter application of high rates (up to 950 Mg ha^-1) of aClass F fly ash. Fly ash was applied to large field plots byeither conventional tillage (CT; moldboard plow-disk) orintensive tillage (IT; chisel plow-rotovate-disk), and tomicroplots using a rototiller. Infiltration rate (i) wasmeasured in both studies with a disk permeameter on threeoccasions over a 12-month period. Ash effects on gravimetric water content (θ_g) at the 0–40 cm soil depth were measuredduring a 168 hr period following a 2.5 cm rainfall event andwater release curves (33 to 500 kPa) were constructed in thelaboratory using soils from the large plots. In both studiesi was decreased by ～80% one year after additionof fly ash and θ_gin ash-amended soil was higher than unamended soil throughoutthe 168 hr monitoring period. Soil water distribution variedwith tillage; the IT treatment had the highest θ_g increasesin the 0–20 cm depth while the CT treatment had θ_gincreases throughout the 0–40 cm depth. Soil water content anddistribution in ash-amended microplots were similar to ITtreatments. Fly ash amendment not only increased water holdingcapacity but also increased plant available water by 7–13% inthe 100–300 kPa range. These results suggest fly ash amendmentmay have the potential to improve crop production in excessivelydrained soils by decreasing i and increasing the amountof plant available water in the root zone.     

Evaluation of porous medium hydraulic properties using experimental methods and RETC code

Two experimental procedures were used to determine both hydraulic properties, soil water retention θ(h) curve and unsaturated hydraulic conductivity K(θ), of a sand sample. Knowledge of hydraulic properties is essential, since they generally control soil water dynamics. A steady-state laboratory method was used for the simultaneous determination of θ(h) and K(θ). A one-step outflow method was used for the determination of diffusivity D(θ) and subsequently K(θ) from soil water retention data which were measured independently on the same sample and using the same apparatus. The comparison of K(θ) measured values from the above-mentioned methods showed very good agreement of the results. Also, the comparison between the experimental K(θ) and θ(h) functions and the predictions obtained using retention curve (RETC) code by simultaneous fit of experimental soil water retention and hydraulic conductivity data from outflow data, assuming the Mualem-van Genuchten model, showed very good agreement. It is noted that the main disadvantage of the one-step outflow method is the weakness to predict K(θ) values near saturation. This disadvantage could be overcome using RETC code with the above procedures, since the K(θ) values between the predictive approach and the steady-state method were similar.     

Characterizing soil surface roughness using a combined structural and spectral approach

The ability to quantitatively and spatially assess soil surface roughness is important in geomorphology and land degradation studies. This paper describes the results of an experiment designed to investigate whether hyperspectral directional reflectance factors can describe fine‐scale variations in soil surface roughness. A Canadian silt loam soil was sieved to an aggregate size range of 1–4.75 mm and exposed to five different artificial rainfall durations to produce soils displaying progressively decreasing levels of surface roughness. Each soil state was measured using a point laser profiling instrument at 2 mm spatial resolution, in order to provide information on the structure and spatial arrangement of soil particles. Hyperspectral directional reflectance factors were measured using an Analytical Spectral Devices FieldSpec Pro Spectroradiometer (range 350–2500 nm), at a range of measurement angles (θ_r=?60° to +60°) and illumination angle conditions (θ_i= 28°–74°). Directional reflectance factors varied with illumination and view angles, and with soil structure. Geostatistically‐derived indicators of soil surface roughness (sill variance) were regressed with directional reflectance factors. The results showed a strong relationship between directional reflectance and surface roughness (R²= 0.94 where θ_r=?60°, θ_i= 67°–74°). This fine‐scale quasi‐natural experiment allowed the control of slope, initial aggregate size and rainfall exposure, permitting an investigation into factors affecting a soil’s bidirectional reflectance response. This has highlighted the relationship between fine‐scale variations in surface roughness, illumination angle and reflectance response. The results show how the technique could provide a quantitative measure of surface roughness at fine spatial scales.     

非常规水水质对红壤水分运移参数的影响及其过程模拟

为了探究非常规水水质对亚热带红壤水分运移参数的影响,采用蒸馏水(CK)、再生水(RW)、稀释2,4,6倍再生水(RW-D2、RW-D4、RW-D6)和生活污水(_(WW))等6种不同水质的非常规水模拟红壤水分入渗过程,利用水平土柱吸渗法测定红壤湿润锋运移速率和水分扩散度,应用压力膜仪法测定了土壤水分特征曲线并且借助van Genuchten(VG)模型提取了相应的土壤残余含水量θ_r、土壤饱和含水量θs、α和n等特征参数值。结果表明:不同水质入渗下红壤湿润锋运移速率、水分扩散度均存在着一定差异,湿润锋运移速率由快到慢分别为V_(CK)≈V_(RW-D2)V_(RW-D6)≈V_(RW-D4)V_(WW)V_(RW);在同一含水率下D_(RW-D6)(θ)D_(RW-D4)(θ)D_(WW)(θ)D_(RW-D2)(θ)D_(CK)(θ)D_(RW)(θ)。不同水质处理对VG模型参数变化率的影响程度不同,对α的影响最大,对θ_r的影响最小。因此,非常规水灌溉会降低红壤入渗速率、持水性能以及抗旱能力。