Data Assimilation Modeling of the Barotropic Tides in the Korea/Tsushima Strait

During 1999–2000, 13 bottom mounted acoustic Doppler current profilers (ADCPs) and 12 wave/tide gauges were deployed along two lines across the Korea/Tsushima Strait, providing long-term measurements of currents and bottom pressure. Tidally analyzed velocity and pressure data from the moorings are used in conjunction with other moored ADCPs, coastal tide gauge measurements, and altimeter measurements in a linear barotropic data assimilation model. The model fits the vertically averaged data to the linear shallow water equations in a least-squares sense by only adjusting the incoming gravity waves along the boundaries. Model predictions are made for the O₁, P₁, K₁, μ₂, N₂, M₂, S₂, and K₂ tides. An extensive analysis of the accuracy of the M₂ surface-height predictions suggests that for broad regions near the mooring lines and in the Jeju Strait the amplitude prediction errors are less than 0.5 cm. Elsewhere, the analysis suggests that errors range from 1 to 4 cm with the exception of small regions where the tides are not well determined by the dataset. The errors in the model predictions are primarily caused by bias error in the model’s physics, numerics, and/or parameterization as opposed to random errors in the observational data. In the model predictions, the highest ranges in sea level height occur for tidal constituents M₂, S₂, K₁, O₁, and N₂, with the highest magnitudes of tidal velocities occurring for M₂, K₁, S₂, and O₁. The tides exhibit a complex structure in which diurnal constituents have higher currents relative to their sea level height ranges than semi-diurnal constituents.     

Environmental controls on food web regimes: A fluvial perspective

Because food web regimes control the biomass of primary producers (e.g., plants or algae), intermediate consumers (e.g., invertebrates), and large top predators (tuna, killer whales), they are of societal as well as academic interest. Some controls over food web regimes may be internal, but many are mediated by conditions or fluxes over large spatial scales. To understand locally observed changes in food webs, we must learn more about how environmental gradients and boundaries affect the fluxes of energy, materials, or organisms through landscapes or seascapes that influence local species interactions. Marine biologists and oceanographers have overcome formidable challenges of fieldwork on the high seas to make remarkable progress towards this goal. In river drainage networks, we have opportunities to address similar questions at smaller spatial scales, in ecosystems with clear physical structure and organization. Despite these advantages, we still have much to learn about linkages between fluxes from watershed landscapes and local food webs in river networks. Longitudinal (downstream) gradients in productivity, disturbance regimes, and habitat structure exert strong effects on the organisms and energy sources of river food webs, but their effects on species interactions are just beginning to be explored. In fluid ecosystems with less obvious physical structure, like the open ocean, discerning features that control the movement of organisms and affect food web dynamics is even more challenging. In both habitats, new sensing, tracing and mapping technologies have revealed how landscape or seascape features (e.g., watershed divides, ocean fronts or circulation cells) channel, contain or concentrate organisms, energy and materials. Field experiments and direct in situ observations of basic natural history, however, remain as vital as ever in interpreting the responses of biota to these features. We need field data that quantify the many spatial and temporal scales of functional relationships that link environments, fluxes and food web interactions to understand how they will respond to intensifying anthropogenic forcing over the coming decades.     

Estimating hypocentral uncertainties for marine microearthquake surveys: A comparison of the generalized inverse and grid search methods

For microearthquake surveys conducted with small networks in regions where the seismic velocity structure has large vertical gradients, the formal errors accompanying hypocentral solutions obtained by a generalized inverse method may be misleading since they do not incorporate the effects of nonlinearity in travel times. An alternative method for estimating uncertainties involves calculating travel time residuals over a regular grid and using the F statistic to contour confidence volumes. We present a statistical expression for the latter confidence limits that is applicable when an independent estimate of arrival time errors is available from observations accumulated for a number of earthquakes. Synthetic experiments comparing the results of the grid search and generalized inverse methods show that in cases where solutions are obtained either without S wave information or for epicenters which lie well outside the network, the effects of nonlinearity on the shape of the confidence regions may be significant. However, for the well-observed events both methods yield comparable confidence volumes in good agreement with the distribution of hypocenters obtained from repeated locations incorporating random errors. The generalized inverse method has the advantage that it requires fewer calculations, so the examination of systematic errors in hypocentral parameters produced by uncertainties in the seismic velocity structure can be studied in a more computationally efficient manner. Except in the cases of poorly resolved earthquakes, the effects of nonlinearity on uncertainties in hypocentral parameters can be observed by the application of the F statistic to the variation of the generalized inverse travel time residuals with focal depth.     

一种多层神经网络的学习算法

提出一种基于序贯预测误差方法（ＳＰＥ）的多层神经网络（ＭＮＮ）的学习算法。经模拟计算，它比传统的基于最陡下降方法的误差反传（ＳＤＢＥＰ）算法具有更好的收敛性能。并对这两种算法进行了模拟计算的比较．     

GMRES方法在大型离岸结构水弹性分析中的应用

使用GMRES方法求解了大型离岸结构水弹性分析所获得的复系数线性方程组，并按照向后误差分析的方法，给出了它的终止准则。数值试验表明对于所考虑的问题，GMRES方法优于直接方法。     

Effect of fluid column dynamics on longitudinal vibrations of an ultra deepwater riser covered by an orifice plug

The paper studies longitudinal vibrations of an ultra-deepwater drilling riser whose bottom end is disconnected from the sea floor and covered by a plug (a blind one or the one with a hole). An elastic shell is used as a model of a riser pipe. The fluid column motion in the riser is described by the Navier–Stokes equation for a compressible fluid. Losses of fluid pressure during its flowing through a hole in the plug are taken into account. Solution of the riser equations is carried out in the frequency domain. Analysis of the effect which the riser length, wave period and height, and the plug-hole diameter on the amplitude of the riser's tension vibrations is conducted. A riser manufactured of strong aluminium alloys, some of prospective materials for ultra deepwater drilling, was considered as a prototype. It has been shown that risers 2000–4000 m long, plug-hole, and a riser 6000 m long with plug-hole of 0.087 m in diameter can withstand wave height of >15 m throughout the frequency range.     

不等保护能力纠错编码在水声信息传输中应用的研究

通过分析水声信道的特性和水声信息源的特点 ,结合信道编码理论与技术的最新进展 ,提出了一种全新的、能适合于水声信息传输的、高性能的不等保护能力纠错编解码方法。     

误差方差二次型估计模型的转化

研究一般的回归模型中误差方差的二次型估计的容许性，研究方法是模型的整体转化和局部转化，结果有：（１）二次约束下的线性模型等价于相应的无约束的线性模型。（２）线性（齐次或非齐次）等式约束下的线性模型等价于某个无约束的线性模型。（３）单个非齐次不等式约束下的线性模型等价于某个无约束的线性模型。（４）通过例子证明了多个线性不等式约束的线性模型不能等价于某个无约束的线性模型。（５）某类非齐次二次型估计的容许性等价于相应的齐次二次型估计的容许性     

水声通信中基于小波变换的图像编码研究

提出了一种高误比特率传输条件下的图像编码方法，它适用于水声信道的图像传输。针对水下图像的特点，选取合适的小波基和变换参数对图像进行离散小波变换；依据小波系数的能量分布特性，对不同的子带采用不同的量化和定长编码，编码率为0．8比特／像素。水声通信试验表明，在传输误比特率达到10^-2时，仍能得到可接受的图像质量。     

Data assimilation methods for estimating a heterogeneous conductivity field by assimilating transient solute transport data via ensemble Kalman filter

An ensemble Kalman filter (EnKF) is developed to identify a hydraulic conductivity distribution in a heterogeneous medium by assimilating solute concentration measurements of solute transport in the field with a steady‐state flow. A synthetic case with the mixed Neumann/Dirichlet boundary conditions is designed to investigate the capacity of the data assimilation methods to identify a conductivity distribution. The developed method is demonstrated in 2‐D transient solute transport with two different initial instant solute injection areas. The influences of the observation error and model error on the updated results are considered in this study. The study results indicate that the EnKF method will significantly improve the estimation of the hydraulic conductivity field by assimilating solute concentration measurements. The larger area of the initial distribution and the more observed data obtained, the better the calculation results. When the standard deviation of the observation error varies from 1% to 30% of the solute concentration measurements, the simulated results by the data assimilation method do not change much, which indicates that assimilation results are not very sensitive to the standard deviation of the observation error in this study. When the inflation factor is more than 1.0 to enlarge the model error by increasing the forecast error covariance matrix, the updated results of the hydraulic conductivity by the data assimilation method are not good at all. Copyright © 2012 John Wiley & Sons, Ltd.