A neural network based methodology to predict site-specific spectral acceleration values

A general neural network based methodology that has the potential to replace the computationally-intensive site-specific seismic analysis of structures is proposed in this paper. The basic framework of the methodology consists of a feed forward back propagation neural network algorithm with one hidden layer to represent the seismic potential of a region and soil amplification effects. The methodology is implemented and verified with parameters corresponding to Delhi city in India. For this purpose, strong ground motions are generated at bedrock level for a chosen site in Delhi due to earthquakes considered to originate from the central seismic gap of the Himalayan belt using necessary geological as well as geotechnical data. Surface level ground motions and corresponding site-specific response spectra are obtained by using a one-dimensional equivalent linear wave propagation model. Spectral acceleration values are considered as a target parameter to verify the performance of the methodology. Numerical studies carried out to validate the proposed methodology show that the errors in predicted spectral acceleration values are within acceptable limits for design purposes. The methodology is general in the sense that it can be applied to other seismically vulnerable regions and also can be updated by including more parameters depending on the state-of-the-art in the subject.

Effects of long-term green manure application on the content and structure of dissolved organic matter in red paddy soil

Dissolved organic matter(DOM) plays important roles in soil biogeochemistry activity and nutrients transportation in soils, but studies regarding the long-term effects of green manures on the content and structure of DOM in red paddy soil have not been reported yet. A long-term green manure experiment established in 1982 was utilized to test the DOM contents in different treatments, and the spectral characteristics of DOM were investigated by using ultraviolet-visible(UV-Vis) spectrometry and Fourier transform infrared(FTIR) spectrometry. The experiment included four cropping systems: ricerice-milk vetch(RRV), rice-rice-rape(RRP), rice-rice-ryegrass(RRG) and rice-rice-winter fallow(RRF), among them, milk vetch, rape, and ryegrass are popular winter green manure species in southern China. The results showed that the content of dissolved organic carbon(DOC), which is widely used to estimate the concentration of DOM, was significantly promoted after the incorporation of green manures compared with the other sampling stages. The contents of aromatic groups and the degree of humification of DOM increased in RRV and RRP, suggesting more complex compositions of the soil DOM after long-term application of milk vetch and rape. The contents of phenol, alcohol and carboxylic acid group at the mature stage of early rice were significantly higher than those at the stage of after green manures turned over, especially for the RRV treatment. The absorption ratio of FTIR indicated that winter plantation of rape increased the aromatic-C/aliphatic-C ratio, and ryegrass increased the aromatic-C/carboxyl-C ratio. In conclusion, long-term planting of milk vetch and rape as green manures increased the degree of aromaticity, humification and average molecular weight of DOM, and made the DOM more stable in red paddy soil.

Scattering of elastic wave by a cylindrical shell deeply embeded in saturated soils

The compression of soil grain and pore fluid as well as viscid coupling of pore fluid and soil skeleton is considered, the scattering problem of incident plane P1 wave （fast compressional wave） by an infinite cylindrical shell deeply embedded in isotropic saturated soils is studied by adopting the amended Biot model, amplitude equations about potential functions of scattering and refracting fields are obtained, and the effect of dimensionless frequencies and shell thickness on the back-scattering spectra and dynamic stress concentration factors of two types of cylindrical shells with high and low rigidity are numerically computed and analyzed.

Research on the attribution evaluating methods of dynamic effects of various parameter uncertainties on the in-structure floor response spectra of nuclear power plant

： Consideration of the dynamic effects of the site and structural parameter uncertainty is required by the standards for nuclear power plants （NPPs） in most countries. The anti-seismic standards provide two basic methods to analyze parameter uncertainty. Directly manually dealing with the calculated floor response spectra （FRS） values of deterministic approaches is the first method. The second method is to perform probability statistical analysis of the FRS results on the basis of the Monte Carlo method. The two methods can only reflect the overall effects of the uncertain parameters, and the results cannot be screened for a certain parameter＇s influence and contribution. In this study, based on the dynamic analyses of the floor response spectra of NPPs, a comprehensive index of the assessed impact for various uncertain parameters is presented and recommended, including the correlation coefficient, the regression slope coefficient and Tornado swing. To compensate for the lack of guidance in the NPP seismic standards, the proposed method can effectively be used to evaluate the contributions of various parameters from the aspects ＆sensitivity, acuity and statistical swing correlations. Finally, examples are provided to verify the set of indicators from systematic and intuitive perspectives, such as the uncertainty of the impact of the structure parameters and the contribution to the FRS of NPPs. The index is sensitive to different types of parameters, which provides a new technique for evaluating the anti-seismic parameters required for NPPs.

A novel simple procedure to consider seismic soil structure interaction effects in 2D models

A method is proposed to estimate the seismic soil-structure-interaction （SSI） effects for use in engineering practice. It is applicable to 2D structures subjected to vertically incident shear waves supported by homogenous half-spaces. The method is attractive since it keeps the simplicity of the spectral approach, overcomes some of the difficulties and inaccuracies of existing classical techniques and yet it considers a physically consistent excitation. This level of simplicity is achieved through a response spectra modification factor that can be applied to the free-field 5%-damped response spectra to yield design spectral ordinates that take into account the scattered motions introduced by the interaction effects. The modification factor is representative of the Transfer Function （TF） between the structural relative displacements and the free- field motion, which is described in terms of its maximum amplitude and associated frequency. Expressions to compute the modification factor by practicing engineers are proposed based upon a parametric study using 576 cases representative of actual structures. The method is tested in 10 cases spanning a wide range of common fundamental vibration periods.

Estimating the Texture of Purple Soils Using Vis-NIR Spectroscopy and Optimized Conversion Models

Soil texture is an indicator of soil physical structure which delivers many ecological functions of soils such as thermal regime, plant growth, and soil quality. However, traditional methods for soil texture measurement are time-consuming and labor-intensive. This study attempts to explore an indirect method for rapid estimating the texture of three subgroups of purple soils (i.e. calcareous, neutral, and acidic). 190 topsoil (0 - 10 cm) samples were collected from sloping croplands in Tongnan and Beibei Districts of Chongqing Municipality in China. Vis-NIR spectrum was measured and processed, and stepwise multiple linear regression (SMLR), partial least squares regression (PLSR), and back propagation neural network (BPNN) models were constructed to inform the soil texture. The clay fractions ranged from 4.40% to 27.12% while sand fractions ranged from 0.34% to 36.57%, hereby soil samples encompass three textural classes (i.e. silt, silt loam, and silty clay loam). For the original spectrum, the texture of calcareous and neutral purple soils was not significantly correlated with spectral reflectance and linear models (SMLR and PLSR) exhibited low prediction accuracy. The correlation coefficients and the goodness-of-fits between soil texture and the transformed spectra of all soil groups increased by continuum-removal (CR), first-order differential (R'), and second-order differential (R") transformations. Among them, the R" had the best performance in terms of improving the correlation coefficients and the goodness-of-fits. For the calcareous purple soil, the SMLR exceeds PLSR and BPNN with a higher coefficient of determination (R2) and the ratio of performance to inter-quartile distance (RPIQ) values and lower root mean square error of validation (RMSEV), but for the neutral and acidic purple soils, the PLSR model has a better prediction accuracy. In summary, the linear methods (SMLR and PLSR) are more reliable in estimating the texture of the three purple soil groups when using Vis-NIR spectroscopy inversion.

A Study of the Effect of Soil Improvement Based on the Numerical Site Response Analysis of Natural Ground in Babol City

A series of numerical calculations have been performed to investigate the effect of soil improvement on seismic site response. Seismic site response analyses were also performed using data collected from a study area in Babol city. The improved site is a composite ground and has more or less different mechanical properties than the natural ground. In this research, the influence of the elastic modulus of the pile, the pile distance ratio, ground motion input, distance to fault rupture, and PGA of the earthquakes on seismic response characteristics are especially investigated. The results reveal that the values of the PGA and amplification factor on the surface of the natural and improved grounds depend strongly on the fundamental period of the site, the predominant period, and the intensity of the ground motion input. The acceleration response spectra also are affected by the characteristics of ground motion input and soil layers. Changing the pile distance ratio doesn’t have a significant effect on the seismic response of the site.

基于EPO-PLS回归模型的盐渍化土壤含水率高光谱反演

表层土壤含水率对于指导农业灌溉有重要的作用。研究表明,土壤光谱受到土壤水分和盐分的共同影响,但对于盐渍化地区的土壤含水率高光谱反演却很少涉及。该文通过对11组不同含盐量土壤室内蒸发过程连续监测,获取相关反射率光谱和水分、盐分的变化数据,利用外部参数正交化方法（external parameter orthogonalisation,EPO）预处理土壤光谱,滤除盐分（质量比0.1%~5.0%）的影响,建立经过EPO预处理后的偏最小二乘（partial least squares regression after EPO pre-processing,EPO-PLS）土壤水分预测模型。与偏最小二乘（partial least square model,PLS）模型相比,验证样本的决定系数R2和对分析误差RPD（residual predictive deviation）分别从0.722、1.976上升到0.898、3.145;均方根误差RMSE从5.087 g/（100 g）减少到3.237 g/（100 g）。通过EPO算法预处理后的模型性能提升显著,利用该方法能够有效的消除土壤盐分的影响,很好地实现盐渍化地区的水分含量估测。

基于EPO算法去除水分影响的土壤有机质高光谱估算

野外进行土壤有机质的光谱快速预测时需考虑土壤含水量的影响。在室内设计人工加湿实验分别获取9个土壤含水量梯度(0~32%,间隔4%)的土壤光谱数据,分析土壤含水量变化对光谱的影响,再利用外部参数正交化法(external parameter orthogonalization,EPO)进行湿土光谱校正,并结合偏最小二乘回归和支持向量机回归分别建立土壤有机质预测模型。结果表明,土壤光谱反射率随着土壤含水量的增加呈非线性降低趋势,偏最小二乘回归模型的预测偏差比为1.16,模型不可用;经EPO算法校正后,各土壤含水量梯度之间的光谱差异性降低,能实现土壤有机质在不同土壤含水量梯度的有效估算,偏最小二乘回归和支持向量机回归模型的预测偏差比分别提高至1.76和2.15。研究结果可为田间快速预测土壤有机质提供必要参考。

多光谱数据定量反演土壤营养元素含量可行性分析

运用室内光谱模拟多光谱数据，采用回归分析方法建立土壤营养元素含量预测模型，并进行验证．首先，根据光谱响应函数将实验室光谱数据重采样至多光谱传感器（TIM和ASTER）波段；然后，分别利用多元逐步回归（SMLR）和偏最小二乘回归（PLSR）方法，建立建模土壤样本实测光谱以及模拟光谱（TIM和ASTER）与土壤营养元素含量间的经验模型；最后，利用检验土壤样本进行模型精度验证．与实测光谱模型相比，模拟光谱模型对土壤营养元素含量的预测精度受光谱分辨率的影响并不大．模拟光谱模型对N、P、K元素含量预测精度最高分别为0．89、0．79和0．67．土壤N、P、K元素含量SMLR模型的入选波段分别位于2000-2300nm、1650～1800nm和600～800nm波长附近范围内；土壤N、P、K元素含量PLSR模型的系数表明，近红外（NIR）波段对总氮和总磷元素含量比可见光（VIS）波段敏感，而VIS对K元素含量预测的贡献更大．利用多光谱数据进行土壤营养元素含量的估测具有理论上的可行性；由于不同元素对不同光谱波段的响应不同，在选择多光谱遥感数据时要充分考虑传感器的波段特征．

土壤高光谱噪声过滤评价研究

以ASD FieldSpecProFR测试的土壤高光谱数据为研究对象，探讨光谱噪声分布以及不同滤波器去噪效果定量评价。土壤高光谱曲线目视及其包络线去除、一阶微分和高通滤波分析表明，除起始波段350nrn最前端40nm范围内有部分噪声外，其他UV／vNIR（350-1050nm）范围内基本不存在噪声，而整个SWIR（1000-2500nm）范围内存在一定的噪声，SWIR2（1800-2500nm）后半段噪声较大，并且组成光谱仪的3台分光计在相互结合处，噪声比邻近谱段更大。采用六种不同滤波器进行噪声去除，通过构建光谱平滑指数（SI）、横向特征保持指数（HFRI）和纵向特征保持指数（VFRI）进行定量评价各滤波器去噪能力，评价结果对比发现WD和MA既能达到曲线平滑又能较好地保持波段特征。此外在PLSR模型下，以66个土壤样本的光谱为例，将六种滤波器去噪后的一阶微分光谱作为模型输入，比较分析不同滤波器对砂粒含量预测精度影响，精度对比表明，相比滤波器的特征保持能力，平滑能力更能影响砂粒含量精度。本研究为开展光谱预处理和光谱分析技术提供有益探索，并能为光谱学相关应用提供了科学依据。

基于光谱分析的土壤游离铁预测研究

土壤游离铁含量的高低可作为土壤系统分类中判断土壤类型的诊断指标,同时也对土壤风化程度具有指示作用,并在一定程度上反映了土壤的成土环境。本研究调查了安徽宣城的91个土壤剖面,共398个土壤样品,采集了样品在350~2500 nm波段的漫反射光谱数据,并对游离铁含量进行化学分析。光谱数据包括反射率（R）、反射率一阶导数（FDR）和吸收度（Log（1/R））3种形式。本文采用偏最小二乘回归算法（PLSR）和反向神经网络（BPNN）建模预测游离铁含量,并分析不同形式光谱数据的建模预测效果。结果表明：当存在游离铁〈20 g/kg的样本时,传统建模方法不能准确预测游离铁含量（R2〈0.6,RPD〈1.5）,相对R和Log（1/R）两种光谱数据,以FDR作为自变量建模预测游离铁含量的效果最差。

黑土土壤水分光谱响应特征与模型

为揭示土壤水分对土壤反射率的影响机理,并为其他土壤参数遥感监测提供理论支持,以吉林省德惠市黑土野外、室内高光谱反射率为研究对象,运用光谱分析方法与统计分析方法,分析土壤水分光谱特征,建立土壤水分光谱模型,得出以下结论:土壤光谱反射率在400～2500nm范围内主要有5个吸收谷,随着含水量的增加,土壤光谱吸收谷的面积增大;而秸秆光谱反射率在土壤前2个吸收谷附近没有明显的吸收特征;未翻耕土壤由于秸秆残茬的影响,第二吸收谷左右两部分面积比例增大,这一特征可用于判别土壤是否翻种;土壤表层0～5cm含水量与反射光谱数据的相关性强于5～10cm;土壤含水量与土壤光谱反射率之间是指数或线性关系;土壤含水量高光谱模型稳定性较好、精度较高,可以用于土壤含水量的速测.

中国6种地带性土壤红外光谱特征研究

利用土壤原样研究了中国 6种典型地带性土壤的红外光谱特征。结果表明 ,砖红壤、红壤红外光谱属于高岭石型图谱 ,显著的 36 95 ,36 2 0 ,10 35 cm- 1峰以及 35 2 7cm- 1峰可作为用红外光谱判断热带、南亚热带地带性土壤的特征峰。黄棕壤、褐土、黑垆土、黄绵土红外光谱为蒙脱石型图谱 ,较弱的 36 2 1cm- 1峰 ,3435和 10 2 8～10 32 cm- 1的强宽带 ,以及显著的 14 37cm- 1峰可作为温带半湿润 -半干旱地区石灰性土壤红外光谱的特征标志。黄棕壤是石灰性土壤向热带、南亚热带地区过渡的地带性土壤 ,其红外光谱标志为较宽的 3435 ,10 32 cm- 1吸收带和明显的 36 2 0 cm- 1吸收峰 ,以及较弱的 36 95 cm- 1峰 ,但没有 14 37cm- 1峰。中国从南到北 ,土壤红外光谱的36 96 ,36 2 1cm- 1 峰、石英双峰 (798,780 cm- 1 )的 797cm- 1 峰吸收强度逐渐减弱 ,而石英双峰的 780 cm- 1 峰吸收强度逐渐增强。

免耕条件下不同秸秆覆盖量的土壤有机碳红外光谱特征

【目的】研究不同秸秆覆盖量免耕条件下不同剖面深度土壤的红外光谱特征,探讨免耕对土壤有机碳组成的影响及秸秆来源土壤有机碳在不同土层的分布特征。【方法】以吉林省梨树县黑土秸秆覆盖免耕长期定位试验区不同层次土壤为研究对象,对比分析无秸秆覆盖(CK)、33%秸秆覆盖、67%秸秆覆盖、100%秸秆覆盖4种处理土壤红外光谱特征。【结果】连续5年秸秆覆盖免耕种植后的土壤,红外光谱吸光值随着剖面层次的加深而降低;差减土壤矿物光谱后的谱图表明:0—5 cm土层土壤红外基团吸收明显增加,芳香碳、脂肪碳、烷基碳较之下层土壤具有明显的吸收峰。【结论】秸秆覆盖有利于有机碳在表层的累积,显著增加秸秆来源的芳香碳、脂肪碳及烷基碳含量。

HHT时频分析土壤光谱的重金属铜离子污染信息提取模型

土壤中不同浓度Cu^(2+)含量映射到土壤光谱上的信息量十分微弱,并且这些高光谱数据中也存在着难以避免的噪声,因而本研究的关键是如何在土壤光谱复杂的噪声环境中提取微弱Cu^(2+)信息。经验模态分解算法(EMD)能够有效去除高光谱数据中的噪声,且EMD是Hilbert变换对"非线性非稳定"信号时频分析的前提,当引入Huang变换后,可利用Hilbert-Huang变换(HHT)模型时频分析高光谱数据以实现降噪处理与信息提取。通过时频的HHT分析不同浓度Cu^(2+)污染下的土壤光谱,完成从原始光谱经EMD分解出各本征模态函数(IMF)分量的包络线、调制信号和频谱等曲线中挖掘土壤光谱的Cu^(2+)污染信息。研究结果表明,相同浓度Cu^(2+)污染时的土壤光谱HHT时频分析结果相同,不同浓度时则不同,所以也可依据IMF分量反演土壤Cu^(2+)含量。因此,高光谱数据的HHT时频分析能为土壤光谱的信息挖掘、光谱诊断和Cu^(2+)含量反演等提供一种新的方法和思路。

基于中红外漫反射光谱的土壤重金属元素含量预测研究

研究了中红外漫反射光谱快速预测土壤重金属元素含量的可行性。以在南京江宁区和八卦洲采集的共161个土壤样品为例,利用偏最小二乘回归(PLSR)法对土壤中Ni,Cr,Cu,As,Zn,Pb,Hg和Cd等8种重金属元素数据进行了预测。通过对样品的中红外(MIR)漫反射光谱进行各种预处理,探讨了中红外光谱数据预处理对预测精度的影响,并比较了中红外光谱与可见光-近红外(VNIR)光谱对土壤重金属含量预测的精度。结果表明,依次经平滑、基线校正、多元散射校正预处理能显著提高中红外光谱数据的预测精度;经校正的中红外光谱对异地样品预测的均方根误差是可见光-近红外光谱的21%～73%,比VNIR波段更能准确预测异地样品中土壤重金属元素含量。研究表明,中红外漫反射光谱可以作为一种快速、非破坏方法预测土壤重金属元素含量,且比可见光-近红外精度高。

基于高光谱分析的植烟土壤有机质和全氮含量预测研究

为实现植烟土壤有机质和全氮含量的快速、准确测定,以120份豫中烟区潮褐土为研究对象,在室内条件下采集并研究了土壤有机质和全氮与高光谱400~2400 nm波段的定量反演关系。光谱经Savitzky-Golay平滑滤波后采用偏最小二乘回归(Partial least square regression,PLSR)方法,建立了土壤高光谱与有机质和全氮间的定量反演模型,对比了3种光谱变换形式及5种预处理方法对模型的影响。结果表明,光谱经一阶微分(FDR)或倒数对数(lg(1/R))变换后,建模系数有所提高;不同预处理方法的反演模型精度差异明显,正交信号校正(OSC)明显优于其他光谱预处理方法。采用FDR-OSC建立的土壤有机质和lg(1/R)-OSC建立的土壤全氮模型预测效果最好,预测模型决定系数R^2分别为0.948和0.919,预测均方根误差(RMSEP)分别为1.316 g·kg^(-1)和0.091 g·kg^(-1),相对分析误差(RPD)分别为4.26和4.07,是豫中烟区潮褐土有机质和全氮含量的最佳预测模型。采用高光谱技术结合OSC-PLSR方法对植烟土壤有机质和全氮含量进行高精度预测是可行的。

基于高分辨率反射光谱的土壤营养元素估算模型

研究了土壤中营养元素含量(N、P、K)与土壤可见光/近红外光谱之间的关系。在对原始光谱进行预处理分析后,计算出了4种光谱指标:反射率Reflectance、一阶导数FDR、倒数之对数log(1/R)和波段深度Depth。通过偏最小二乘回归分析建立了营养元素与4种光谱指标的经验模型,并且利用验证样本集对回归模型进行了验证。结果表明,可见光/近红外反射光谱具有快速估算土壤中营养元素含量的潜力。

基于棉花冠层光谱的土壤氮素监测研究

通过连续2年小区氮肥试验,在棉花不同生育期采集冠层高光谱数据并同步测定土壤氮含量,分析棉花冠层高光谱参数与土壤氮含量间的关系,建立基于植株冠层光谱的土壤氮含量估算模型。结果表明:土壤全氮含量随着施氮水平的增加而增加,且差异显著;基于棉花不同时期冠层光谱构建的14种光谱参量与土壤氮含量间的相关性有显著差异。其中,利用冠层光谱参数P_Area1100、Depth980、Area672、PPR(550,540)建立的土壤氮含量监测模型分别在蕾期、花期、铃期、吐絮期4个关键生育期对土壤氮含量的预测均达到了较高的精度,能够很好地反映棉花土壤氮素营养状况。利用植株冠层光谱参数可以很好地监测土壤氮素营养,说明利用植株冠层光谱方法监测土壤氮含量是可行的。