Acquiring flow structure characteristics in specific areas and constructing large-scale flow fields are necessary tasks for analyzing the evolution of river channels and developing water conservancy projects. The Acoustic Doppler Current Profiler (ADCP) is commonly used to observe large-scale flow fields in rivers. However, the conventional method of constructing flow fields has several challenges, including inaccurate estimation of flow velocity, distorted identification of flow structure in specific areas, and inability to explain the sediment transport mechanism. Therefore, it is necessary to explore a new method for large-scale flow field construction. This paper utilizes vessel-mounted ADCP data with dense interval single transects, along with single-beam data, to perform optimal flow velocity estimation based on the actual spatial positions of radial velocities. Experiments demonstrate that the method proposed in this paper is superior to the conventional method in estimating magnitude and direction of flow velocity at different water depths. Furthermore, the flow field analysis using the proposed method successfully identifies flow features such as plunging flow and secondary flow. The maximum difference in grid velocity exceeds 0.2 m/s, while the root-mean-square error is reduced by 50 % compared to the conventional method. Results reveal a significant correlation between eddies and scour holes, providing a novel perspective on understanding the sediment transport mechanism of scour holes and exploring the relationship between vertical structures and localized scour.

Study region: Xuliujing section, Yangtze River Estuary, China. Study focus: Predicting discharge in tidal rivers is challenging due to complex flow dynamics influenced by topography, tides, runoff, and weather. Traditional forecasting methods struggle with fixed parameters, limiting their adaptability and accuracy over time. To address this, we propose an enhanced deep learning model, A KAN-augmented LSTM framework, which integrates a Kolmogorov-Arnold network (KAN) with a long short-term memory (LSTM) network. This model retains LSTM's ability to handle long-term dependencies while replacing the fully connected layer with a KAN layer. A learnable B-spline activation function in the KAN layer improves the model's capacity to capture nonlinear dynamics and long-term dependencies in time series data, enhancing forecasting accuracy. This paper applies the LSTM-KAN model to the Xuliujing section of the Yangtze River Estuary and compares its performance with traditional harmonic analysis (HA) and four neural network models: LSTM, XGBoost, DLinear, and Informer. New hydrological insights for the region: The results demonstrate that the LSTM-KAN model significantly enhances discharge forecasting accuracy, outperforming all comparative methods across short-term (6 h), medium-term (12-24 h), and long-term (36-48 h) forecasts. Specifically, it achieved relative accuracy improvements of 12.1 %-35.2 % over HA and 7 %-52.8 % over the traditional LSTM model. These findings suggest that the complex interplay of tidal forcing, runoff, and weather in the Yangtze Estuary is better represented by the adaptive, function-learning paradigm of KAN than by models with fixed nonlinearities. The model's superior performance offers new insights for studying complex flow dynamics, indicating that deep learning techniques with learnable activation functions provide a more powerful and accurate tool for operational forecasting in highly dynamic tidal river environments.

The complexity of river-tide interaction poses a significant challenge in predicting discharge in tidal rivers. Long short-term memory (LSTM) networks excel in processing and predicting crucial events with extended intervals and time delays in time series data. Additionally, the sequence-to-sequence (Seq2Seq) model, known for handling temporal relationships, adapting to variable-length sequences, effectively capturing historical information, and accommodating various influencing factors, emerges as a robust and flexible tool in discharge forecasting. In this study, we introduce the application of LSTM-based Seq2Seq models for the first time in forecasting the discharge of a tidal reach of the Changjiang River (Yangtze River) Estuary. This study focuses on discharge forecasting using three key input characteristics: flow velocity, water level, and discharge, which means the structure of multiple input and single output is adopted. The experiment used the discharge data of the whole year of 2020, of which the first 80% is used as the training set, and the last 20% is used as the test set. This means that the data covers different tidal cycles, which helps to test the forecasting effect of different models in different tidal cycles and different runoff. The experimental results indicate that the proposed models demonstrate advantages in long-term, mid-term, and short-term discharge forecasting. The Seq2Seq models improved by 6%-60% and 5%-20% of the relative standard deviation compared to the harmonic analysis models and improved back propagation neural network models in discharge prediction, respectively. In addition, the relative accuracy of the Seq2Seq model is 1% to 3% higher than that of the LSTM model. Analytical assessment of the prediction errors shows that the Seq2Seq models are insensitive to the forecast lead time and they can capture characteristic values such as maximum flood tide flow and maximum ebb tide flow in the tidal cycle well. This indicates the significance of the Seq2Seq models.

The artificial visual method is currently commonly used to decipher multi-beam water column images to obtain the position and state of lost shipping containers. However, the recognition efficiency and accuracy of this method need to be improved. The You Only Look Once (YOLO) series model has strong real-time target detection capability. Meanwhile, the Segment Anything Model (SAM) has strong zero-shot transferability. A detection and extraction method for lost shipping containers, which combines the two models mentioned above, is proposed in this study. First, the YOLO series model is employed to detect lost shipping container targets in a single-frame water column image. On this basis, the output of bounding box positions by the optimal target detection model is used as a prompt for the SAM. Finally, the SAM is used to extract lost shipping container targets in images through zero-shot transferability. Experimental results in the Pearl River estuary show that the combined modelling method of YOLOv5-n and EdgeSAM-3x achieves the best overall performance. The precision and recall for the detection of lost shipping containers by this method is better than 95%. In terms of target extraction, YOLOv5-n and EdgeSAM-3x have the best Intersection over Union, recall, and F1 scores.

In recent years, the surge in marine activities has increased the frequency of submarine pipeline failures. Detecting and identifying the buried conditions of submarine pipelines has become critical. Sub-bottom profilers (SBPs) are widely employed for pipeline detection, yet manual data interpretation hampers efficiency. The present study proposes an automated detection method for submarine pipelines using deep learning models. The approach enhances the YOLOv5s model by integrating Squeeze and Excitation Networks (SE-Net) and S2-MLPv2 attention modules into the backbone network structure. The Slicing Aided Hyper Inference (SAHI) module is subsequently introduced to recognize original large-image data. Experimental results conducted in the Yellow Sea region demonstrate that the refined model achieves a precision of 82.5%, recall of 99.2%, and harmonic mean (F1 score) of 90.0% on actual submarine pipeline data detected using an SBP. These results demonstrate the efficiency of the proposed method and applicability in real-world scenarios.

针对感潮河段受径流和潮汐共同影响导致流量预报精度较低的问题,结合变分模态分解(variational mode de-composition,VMD)处理非平稳信号的能力与反向(back propagation,BP)神经网络处理非线性问题的优势,提出了基于潮流调和分析及VMD-BP神经网络组合模型的感潮河段流量预报方法.首先,采用潮流调和分析方法对原始流量进行潮流、余流分离;然后,根据误差逆向传播算法构建BP神经网络,并对潮流数据和经VMD处理后的余流数据进行仿真训练;最后,将仿真训练输出的潮流和余流分量叠加重构进而得到最终的流量预报结果.在长江口徐六泾断面开展流量预报实验,结果表明,单独采用BP神经网络方法相对于传统潮流调和分析方法的流量预报精度提高了约3400 m3/s,相对精度提高了约6％;所提组合模型方法的流量预报精度相对于传统潮流调和分析方法提高了约5 000 m3/s,相对精度提高了约9％.基于调和分析及VMD-BP神经网络的组合模型可以有效提高感潮河段流量预报精度,同时也为流况多变水域的流量预报提供新思路.

Three long-term fixed acoustic Doppler current profilers were first used for investigating the vertical structure of tidal currents in Xuliujing Section of Changjiang River Estuary. Moreover, three different periods (spring, summer and fall) were also considered for investigating seasonal variations. The semi-diurnal tides were the most energetic, with along-channel speed of up to 80 cm/s for M2 constituent, which dominates at all stations with percent energy up to 65％–75％ during seasons. The shape of tidal ellipses of the most energetic semi-diurnal constituent M2 showed obvious polarization of the flow paralleling to the riverbank, with the minor semi-axis being generally less than 20％ of the major one. The maximum velocity of mean current is appeared in top layers at all the three stations, and the velocity decreased with the depth. The seasonal variations of direction are also observed, which is probably caused by complex local topography since the erosion and deposition in riverbed. Observed vertical variation of four parameters of M2 ellipses, agreed well with the optimally fit frictional solutions in top and middle layers. However, there was an obvious difference between frictional model and observed data in the lower water column. Discrepancies are probably on account of stratification, which strengthens in summer and fall due to the freshening influence of the Changjiang River Estuary outflow.

多波束数据的广泛利用目前存在两个问题:一是多波束数据多以二进制等形式存储,难以直接进行读取使用;二是由于不同厂商的数据格式不同,批量处理多个来源数据的难度较大.为此,研究了 S7K、ALL、HSX、XTF、XSE、GSF共6种主流多波束数据格式及数据结构,并比较分析了不同数据格式之间的差异性.在此基础上,设计开发了一款能够自动解译并快速显示和存储所需数据的软件系统.该软件通过数据库技术管理海量数据,可以按照自定义格式存储关键数据并以文本格式输出,数据可视化图形实时检索定位,有助于提高多波束数据解译和分析的效率.

The quantitative study of vadose zone water (i.e., the water contained in the rock formation between the ground surface and the groundwater level) is of vital importance for understanding the groundwater flow mechanism, mass exchange, and circulation. However, the vadose zone is closely related to the saturated zone, and it is difficult to directly observe the vadose zone water, causing the quantitative separation of the vadose zone water storage to be a challenging problem. Considering the insufficient spatial resolution of Gravity Recovery and Climate Experiment (GRACE), we draw on the practice of converting the observed groundwater level changes into the water storage changes in the study of groundwater storage changes using GRACE data, and propose a new method for implementing the quantitative separation of vadose zone water storage changes based on a single superconducting gravimeter and groundwater level observation data. This new method is validated with measurement data collected using a superconducting gravimeter (GWR-C032) at the National Geodetic Obser-vation Station in Wuhan, China (Wuhan SG Station) and groundwater level observation data obtained during the same period. The calculation results show that the changes in the vadose zone water storage obtained using the superconductive gravity technique are in good agreement with that obtained using the local hydrological modeling method, indicating that the quantitative separation of the changes in the local vadose zone water storage can be achieved using superconductive gravity and groundwater level observations. During the study period from May 2008 to April 2010, the peak-to-peak changes of the vadose zone water storage at Wuhan SG Station reached 1580 mm, and the seasonal variations were the inverse of the variations of the total groundwater storage changes and the saturated zone water storage changes, indicating that the vadose zone plays a pro-nounced role in slowing down the total groundwater storage changes.

GPS data are usually used to measure boat velocity during boat-mounted acoustic Doppler current profiler (ADCP) measurements when bottom tracking is biased by moving bed. GPS cannot provide consistently accurate boat velocity reference because of multipath errors, satellite signal reception problems, and heading errors. In addition, the computation of water velocity from an ADCP mounted onto a moving boat is a vector-algebra problem, thus the discharge calculation is subject to the compass error when GPS is used for boat velocity reference. This paper proposes a method for correcting discharge based on the idea that the discharge calculation is independent of the boat path. The processing of two sets of boat-mounted ADCP measurements integrated with differential GPS and non-differential GPS was simulated to verify the method. The results show that the proposed method performs well in both differential and non-differential GPS conditions. The relative errors range from 0.1% to 1.5% for all measurements with the mean relative errors of 0.7%. Analytical assessment of the GPS errors shows the proposed method is insensitive to the positioning accuracy of GPS, but positioning error of non-differential GPS may induce relative discharge error of more than 1% when the river or stream is narrow. On the contrary, a relatively small compass or heading error can cause a significant error in water velocity and discharge when using GPS as the boat velocity reference. Therefore, integrating a differential GPS and maintaining a slow boat speed are best practices for discharge measurement, especially for narrow streams or short boat paths.

Buildings are extremely important artificial feature objects. Extracting buildings can provide technical support for urban planning, population estimation, and landscape analysis. Object-oriented classification is an effective method for extracting ground objects and has been widely used in the extraction of building information. The object-oriented morphological building index method has good practicability, but the effect of extracting sparse buildings still needs to be improved. To solve this problem, the median absolute deviation is applied to the object-oriented building extraction, and the two situations of dense and sparse buildings are analyzed. Precision, recall, and F-1 score are used to evaluate the extraction results. Experimental results show that the object-oriented median absolute deviation method extracts sparse buildings significantly better than the object-oriented classification and object-oriented morphological building index methods.

长江三角洲是我国经济发达、人口密集的地区之一,其丰富的水资源对经济发展作用巨大.获取该水域准确的时空流场及实时流量信息,对于水利工程规划建设、防洪减灾及水资源科学管理和合理分配等意义重大.受河道、径流潮流、变化流速等因素综合影响,长江口水域流速直接获取困难、流场组成不单一、流量实时准确估计困难,严重影响了该水域流场及流量信息的准确获取及其应用.据此,开展了长江口局域时空流场构建及实时流量精确估计方法研究.论文的主要内容如下.

The supervised classification technology of remote sensing images is widely used in the field of information extraction and change detection in which the selection of training samples is very important and the quality of training samples directly determines the accuracy of classification However due to the limitation of conditions and human error some impure or wrong training samples may be selected resulting in a decrease in classification accuracy In order to solve this problem the median absolute deviation method is used to detect and eliminate impure and wrong training samples in the supervised classification of remote sensing images based on the spectral information of the image The optical remote sensing image data obtained from Landsat-8 in some areas of Nanchang city is selected the support vector machine is used to supervise and classify the two situations that contain abnormal training samples and eliminate abnormal training samples and compare the classification results Experimental results show that the classification accuracy of removing abnormal training samples is significantly better than that of abnormal training samples

基于长江口徐六泾断面3个测站2011年春、夏、秋3个季节对应的不同径流量下的ADCP流速剖面数据,采用调和分析方法对徐六泾感潮河段的余流和潮流进行分离.研究表明,余流流速变化规律与长江口径流流量季节性变化一致,各测站流向沿水深变化平稳,基本与河岸线平行;潮流椭圆极化严重,几乎接近于直线,且基本与河岸线平行;潮流振幅和倾角的垂直变化和季节变化规律与余流的流速和流向的变化规律一致,即潮流的强度与径流强度呈正相关,且潮流椭圆的倾角变化幅度与平均流速变化幅度基本一致.

The premise of dealing with uncertainty in this article is to figure out the impact of uncertainties on project performance. This article proposes a framework for quantitative analysis that combines a fuzzy mathematics approach with system dynamics theory. Simulations generated using a model based on the framework show that uncertainties affect project performance by changing project parameters, and the mechanisms involved depend on the sources of the uncertainties. Furthermore, while technical and financial uncertainties mainly influence project costs, uncertainties related to markets, regulations, and organizational structures affect both project duration and cost. Finally, strategies for coping with uncertainties in a resource allocation perspective are presented.

This paper proposes a real-time discharge estimation method based on double vertical combination as an alternative to the index flow velocity method,which is based on the vertical data of a single acoustic Doppler current profiler(ADCP).The index flow velocity method is unsuitable for large tidal reaches with large section width, large lateral flow velocity difference,and strong seasonal changes of runoffs.Experiments performed at the Xuliu-jing section of the Yangtze Estuary show that the flow velocity estimation precision of the traditional single vertical line at the section is low;the maximum value of its root mean square error(RMSE)is greater than 0.1 m/s;the relative standard deviation RSD for flow estimation is 11.3%; the RMSE of flow velocity in the double vertical method are all less than 0.07 m/s;and the relative standard deviation(RSD)for flow estimation is 7.3%.How-ever,the estimation precision of the total section flow increases by 2 300 m3/s.Therefore, for such a large tidal reach as the Yangtze Estuary with an annual average discharge of 2.8 ×104m3/s,the 4%improvement of the rel-ative estimation precision in the double vertical method is very significant.

This paper analyses the characteristics and application of Green and Gauss functions and concludes that the number and the position of nodes is the key of detiding.Threshold method was proposed to determine the number of nodes based on fitting error and gradient algorithm was put forward to determine the position of nodes based on the changes of flow field.The tidal reach in the vicinity of Xuliujing of the Yangtze River estuary was taken as the study area.Traditional Candela method and the combination of threshold method and gradient algorithm proposed were adopted to separate the mean flow and tidal current from vessel-mounted ADCP data of three periods of spring,middle and neap tide in the eight sections of the water.The results show that the accuracies of proposed method is much better than that of traditional method.In addition,the interpolation and extrapolation accuracies for the same period are consistent,which indicates that the interpolation scheme proposed in this paper is valid and effectiveness in short-time extrapolation.

The discharge at Datong station may be not entirely representative of net discharge into the East China Sea due to water abstraction and precipitation along the lower Yangtze River. Therefore, Xuliujing station, which is much nearer to the river mouth, is chosen to accurately estimate the net discharge in this paper. The Multi-Profile Method that considers the cross-sectional variability of flow is applied at Xuliujing station to estimate the tidal discharge. Then, the low-pass filter is used to remove the effects of the tide and the net discharge is obtained. Lastly, the net discharge in Datong and that in Xuliujing are compared. The results show that the net discharge of Xuliujing is obviously larger than that of Datong in the rainy season due to precipitation, but the net discharge of Xuliujing is slightly smaller than that of Datong in the dry season due to persistent abstraction and less precipitation. Therefore, in some degree, the net discharge of Xuliujing is more representative than that of Datong regarding the discharge into the East China Sea.

ADCP data processing software WinRiver has practical limitations of the instrument itself, namely,to obtain the vertical data with specific distance from initial point at every hour is the software in ADCP must interpret several R files at the same time and interpolate the interpreted data vertically by time and space.For improving the data processing efficiency, an ADCP data processing software system is developed by interpreting original ADCP R file and the theoretical study of ADCP data processing.The practical test in the estuary of Changjiang showed that the software ran well in data processing and post processing.