Wavelet regression model as an alternative to neural networks for monthly streamflow forecasting   总被引：3，自引：0，他引：3  

Özgür Kişi 《水文研究》2009,23(25):3583-3597

The accuracy of the wavelet regression (WR) model in monthly streamflow forecasting is investigated in the study. The WR model is improved combining the two methods—the discrete wavelet transform (DWT) model and the linear regression (LR) model—for 1‐month‐ahead streamflow forecasting. In the first part of the study, the results of the WR model are compared with those of the single LR model. Monthly flow data from two stations, Gerdelli Station on Canakdere River and Isakoy Station on Goksudere River, in Eastern Black Sea region of Turkey are used in the study. The comparison results reveal that the WR model could increase the forecast accuracy of the LR model. In the second part of the study, the accuracy of the WR model is compared with those of the artificial neural networks (ANN) and auto‐regressive (AR) models. On the basis of the results, the WR is found to be better than the ANN and AR models in monthly streamflow forecasting. Copyright © 2009 John Wiley & Sons, Ltd.  相似文献   

Spatial analysis of instream nitrogen loads and factors controlling nitrogen delivery to streams in the southeastern United States using spatially referenced regression on watershed attributes (SPARROW) and regional classification frameworks   总被引：1，自引：0，他引：1  

Anne B. Hoos  Gerard McMahon 《水文研究》2009,23(16):2275-2294

Understanding how nitrogen transport across the landscape varies with landscape characteristics is important for developing sound nitrogen management policies. We used a spatially referenced regression analysis (SPARROW) to examine landscape characteristics influencing delivery of nitrogen from sources in a watershed to stream channels. Modelled landscape delivery ratio varies widely (by a factor of 4) among watersheds in the southeastern United States—higher in the western part (Tennessee, Alabama, and Mississippi) than in the eastern part, and the average value for the region is lower compared to other parts of the nation. When we model landscape delivery ratio as a continuous function of local‐scale landscape characteristics, we estimate a spatial pattern that varies as a function of soil and climate characteristics but exhibits spatial structure in residuals (observed load minus predicted load). The spatial pattern of modelled landscape delivery ratio and the spatial pattern of residuals coincide spatially with Level III ecoregions and also with hydrologic landscape regions. Subsequent incorporation into the model of these frameworks as regional scale variables improves estimation of landscape delivery ratio, evidenced by reduced spatial bias in residuals, and suggests that cross‐scale processes affect nitrogen attenuation on the landscape. The model‐fitted coefficient values are logically consistent with the hypothesis that broad‐scale classifications of hydrologic response help to explain differential rates of nitrogen attenuation, controlling for local‐scale landscape characteristics. Negative model coefficients for hydrologic landscape regions where the primary flow path is shallow ground water suggest that a lower fraction of nitrogen mass will be delivered to streams; this relation is reversed for regions where the primary flow path is overland flow. Published in 2009 by John Wiley & Sons, Ltd.  相似文献   

曲率模态小波法用于网壳结构损伤的识别和定位   总被引：1，自引：0，他引：1  

殷志祥  徐文杰  毛丽强 《地震工程与工程振动》2009,29(4)

工程结构损伤的识别与定位研究以往主要针对梁、框架等结构形式,根据大跨度空间结构杆件和节点繁多等特点,提出用曲率模态和小波混合方法对空间结构的损伤进行识别和定位.以跨度100 m的Schwedler网壳结构损伤前、后的曲率模态作为标识量,分别通过离散和连续小波变换,判断网壳结构有无损伤和损伤位置,统计了小波系数差与结构损伤的图形关系,计算了各种损伤工况下该方法判断损伤的准确程度.结果发现基于曲率模态和小波方法的大跨度网壳结构损伤定位精度很高,充分证明该方法对此类结构损伤定位具有有效性和实用性.  相似文献   

结构动力方程的显式级数积分格式     

郭泽英  李青宁 《地震工程与工程振动》2009,29(4)

将Newmark-β法中常平均加速度法的基本假定与精细指数算法结合,根据指数矩阵的Taylor级数展开式,提出了动力方程的显式级数解,并设计了相应的时程积分算法.该算法的精度可根据Taylor级数展开式的项数进行灵活控制.算例的结果表明:在满足稳定性条件的前提下,随着时间步长的增加,其精度优于传统的时程积分法.通过稳定性的分析,指出其稳定性条件是显然满足的.  相似文献   

Ground motion attenuation relations of small and moderate earthquakes in Sichuan region   总被引：1，自引：0，他引：1  

Lanchi Kang    Xing Jin  Fuzhou University  Fuzhou  China Earthquake Administration of Fujian Province  China 《地震学报(英文版)》2009,(3):277-282

On the basis of 10 935 broadband velocity records of 135 earthquakes (ML3.0-6.4 and epicentral distance of 26-623 km) occurred from May 12 to June 10 in 2008, which are collected from 27 bedrock stations included in Sichuan Earthquake Monitoring Network, the corresponding acceleration records are obtained by a real-time simulation method. Then by regression analysis on the data, the relation between the peak ground acceleration and velocity attenuation of small and moderate bedrock earthquakes occurred in S...  相似文献   

Dynamlc rupture process of the 1999 Chi-Chi, Taiwan, earthquake     

Haiming Zhang  Xiaofei Chen 《地震学报(英文版)》2009,(1):3-12

  相似文献   

Estimating overland flow erosion capacity using unit stream power   总被引：2，自引：0，他引：2  

Hui-Ming SHIH  Chih Ted YANG 《国际泥沙研究》2009,24(1):46-62

Soil erosion caused by water flow is a complex problem. Both empirical and physically based approaches were used for the estimation of surface erosion rates. Their applications are mainly limited to experimental areas or laboratory studies. The maximum sediment concentration overland flow can carry is not considered in most of the existing surface erosion models. The lack of erosion capacity limitation may cause over estimations of sediment concentration. A correlation analysis is used in this study to determine significant factors that impact surface erosion capacity. The result shows that the unit stream power is the most dominant factor for overland flow erosion which is consistent with experimental data. A bounded regression formula is used to reflect the limits that sediment concentration cannot be less than zero nor greater than a maximum value. The coefficients used in the model are calibrated using published laboratory data. The computed results agree with laboratory data very well. A one dimensional overland flow diffusive wave model is used in conjunction with the developed soil erosion equation to simulate field experimental results. This study concludes that the non-linear regression method using unit stream power as the dominant factor performs well for estimating overland flow erosion capacity.  相似文献   

Effect of nonuniform boundary conditions on steady flow in saturated homogeneous cylindrical soil columns     

D.A. Barry 《Advances in water resources》2009

  相似文献   

Comparison of cell- and vertex-centered discretization methods for flow in a two-dimensional discrete-fracture–matrix system     

H. Hægland  A. Assteerawatt  H.K. Dahle  G.T. Eigestad  R. Helmig 《Advances in water resources》2009

Simulations of flow for a discrete-fracture model in fractured porous rocks have gradually become more practical, as a consequence of increased computer power and improved simulation and characterization techniques. Discrete-fracture models can be formulated in a lower-dimensional framework, where the fractures are modeled in a lower dimension than the matrix, or in an equi-dimensional form, where the fractures and the matrix have the same dimension.  相似文献   

Efficient uncertainty quantification techniques in inverse problems for Richards’ equation using coarse-scale simulation models     

P. Dostert  Y. Efendiev  B. Mohanty 《Advances in water resources》2009

This paper concerns efficient uncertainty quantification techniques in inverse problems for Richards’ equation which use coarse-scale simulation models. We consider the problem of determining saturated hydraulic conductivity fields conditioned to some integrated response. We use a stochastic parameterization of the saturated hydraulic conductivity and sample using Markov chain Monte Carlo methods (MCMC). The main advantage of the method presented in this paper is the use of multiscale methods within an MCMC method based on Langevin diffusion. Additionally, we discuss techniques to combine multiscale methods with stochastic solution techniques, specifically sparse grid collocation methods. We show that the proposed algorithms dramatically reduce the computational cost associated with traditional Langevin MCMC methods while providing similar sampling performance.  相似文献