Inversion of the three-dimensional temperature structure of mesoscale eddies in the Northwest Pacific based on deep learning

Mesoscale eddies, which are mainly caused by baroclinic effects in the ocean, are common oceanic phenomena in the Northwest Pacific Ocean and play very important roles in ocean circulation, ocean dynamics and material energy transport. The temperature structure of mesoscale eddies will lead to variations in oceanic baroclinity, which can be reflected in the sea level anomaly (SLA). Deep learning can automatically extract different features of data at multiple levels without human intervention, and find the hidden relations of data. Therefore, combining satellite SLA data with deep learning is a good way to invert the temperature structure inside eddies. This paper proposes a deep learning algorithm, eddy convolution neural network (ECN), which can train the relationship between mesoscale eddy temperature anomalies and sea level anomalies (SLAs), relying on the powerful feature extraction and learning abilities of convolutional neural networks. After obtaining the temperature structure model through ECN, according to climatic temperature data, the temperature structure of mesoscale eddies in the Northwest Pacific is retrieved with a spatial resolution of 0.25° at depths of 0–1 000 m. The overall accuracy of the ECN temperature structure is verified using Argo profiles at the locations of cyclonic and anticyclonic eddies during 2015–2016. Taking 10% error as the acceptable threshold of accuracy, 89.64% and 87.25% of the cyclonic and anticyclonic eddy temperature structures obtained by ECN met the threshold, respectively.     

Long-range forecasting of suspended sediment

This study examines the employment of two methods, multiple linear regression (MLR) and an artificial neural network (ANN), for multistep ahead forecasting of suspended sediment. The autoregressive integrated moving average (ARIMA) model is considered for one-step ahead forecasting of sediment series in order to provide a comparison with the MLR and ANN methods. For one- and two-step ahead forecasting, the ANN model performance is superior to that of the MLR model. For longer ranges, MLR models provide better accuracy, but there is an important assumption violation. The Durbin-Watson statistics of the MLR models show a noticeable decrease from 1.3 to 0.5, indicating that the residuals are not dependent over time. The scatterplots of the three methods (MLR, ARIMA and ANN) for one-step ahead forecasting for the validation period illustrate close fits with the regression line, with the ANN configuration having a slightly higher R² value.     

基于互联网数据的城市社区租金评估及 空间格局制图——以深圳市为例

准确刻画精细化尺度下的城市房租空间格局,对于研究城市居住行为、提高城市规划合理性十分重要。文章提出了一种基于互联网房租数据作为可靠数据源的城市房租空间格局制图方法。以深圳市作为研究区,通过广泛采集开放平台中用户发布的租房信息,绘制社区行政区尺度的房租空间分布图。房租空间制图涉及到对于没有样本数据区域平均房租的评估,因此,选取一系列与房租相关的房产属性、房屋区位及配套设施的评价指标,采用前馈神经网络技术构建评估模型。以2015年深圳市的住宅租赁市场作为研究对象,通过对结果的分析,以及与权威部门发布的统计数据进行比较,表明此方法能够有效地绘制社区尺度下城市房租的空间分布,模型预测结果的误差（%RMSE）为13.87%。所使用的互联网房租数据、POIs数据及前馈神经网络的建模工具均是开源的,而且所提出的方法论具有普适性,能够应用于其他研究区的房租空间格局制图,具有实践意义。     

融合GNSS气象参数的BP神经网络雾霾预测研究

结合IGS中心获取的BJFS站气象参数（气温（T）、气压（P）、大气可降水量（PWV））及同期PM_2.5数据,建立一种融合时序网络和回归网络的雾霾预测模型,对PM_2.5浓度进行预测。研究表明,引入GNSS气象参数的融合网络模型较单一网络模型适应性强、准确度高,在一定精度范围内可准确预测PM_2.5的变化,时效性达3 h。本文结论验证了卫星导航技术应用于雾霾天气监测及预报的可行性。     

基于神经网络模型误差补偿技术的对流层延迟模型研究

针对传统对流层延迟模型精度较低的缺点,基于神经网络模型误差补偿技术,在Hopfield模型基础上建立一个适用于北半球的高精度融合模型。以Wyoming大学提供的2010年全球120多个观测台站的气象探空数据精密解算的天顶对流层延迟(ZTD)作为近似“真值”,分析比较Hopfield模型、传统BP模型和融合模型的计算精度。结果表明,Hopfield模型的均方根误差（RMSE）为35.31 mm,传统BP模型为30.34 mm,融合模型为23.31 mm。     

分部加权的行道树检测网络构建

基于深度学习目标检测框架,提出了一种端到端的训练网络,用于行道树的自动检测。由于行道树之间的遮挡问题,现有的通用物体检测框架无法直接应用于此任务,为此本文提出了一种树形分部加权模块,以减少严重遮挡造成的错误检测。然后对提出的神经网络进行训练和评估。结果显示,本文所建立的分部加权树木检测网络能够在遮挡条件下,有效地检测出街景图像中的行道树,该方法在各种条件下均具有较高的精度和良好的稳健性。     

河口型水源地供水安全风险分类评估方法比选研究

为科学评估河口型水源地供水安全风险,基于水源地特点,在建立相应供水安全风险评估指标体系的基础上,采用层次分析法确定指标权重,结合选取灰色关联度法和BP神经网络法对某河口型水源地分别开展供水安全风险分类评估。以某次咸潮入侵事件数据为例,经采用两种方法评估后对比表明:灰色关联度法能较好地解决评价指标难以准确量化和统计的问题,具有定性与定量相结合评价精度较高的优点,更适合于河口型水源地供水安全风险分类评估。该水源地风险分类评价等级由高到低依次排序:水量风险(较高)→应急风险(较低)→工程风险(较低)→水质风险(低),其结果与实际相符。     

Artificial neural network-based storm surge forecast model: Practical application to Sakai Minato,Japan

The present study describes a novel way of a systematic and objective selection procedure for the development of an Artificial Neural Network-based storm Surge Forecast Model (ANN-SFM) with the 5, 12 and 24 h-lead times and its application to Sakai Minato area on the Tottori coast, Japan. The selection procedure guides how to determine the superiority of the best performing model in terms of the appropriate combination of unit number in the hidden layer and parameter in the input layer. In the application of ANN-SFM to Sakai Minato, it is found that the best 5 and 12 h-forecast ANN-SFMs are established with the most suitable set of 70 units (the number of hidden neurons) and the input components of surge level, sea level pressure, the depression rate of sea level pressure, longitude, latitude, central atmospheric pressure and highest wind speed. The best 24 h-forecast ANN-SFM is determined with 160 units and the input parameters of surge level, sea level pressure, the depression rate of sea level pressure, longitude and latitude. The proposed method of the selection procedure is able to be adaptable to other coastal locations for the development of the artificial neural network-based storm surge forecast model as establishing the superiority of the most relevant set combining unit numbers and input parameters.     

基于多种神经网络的风暴潮增水预测方法的比较分析

简要介绍了利用BP神经网络、小波神经网络、递归神经网络进行风暴潮增水值预测的原理。选取广东省珠江口以南的阳江站2017年风暴潮增水数据进行测试。结果表明,三种神经网络方法针对阳江地区风暴潮增水的预测均具有可靠性和实用性。以当前增水值为输入量的单因子模型更能反映真实风暴潮增水趋势,而从增水极值预测的准确性来看,以台风风力、气压、风向等相关参数为输入量的多因子模型优于单因子模型。BP神经网络更适用于多因子长时间预测,小波神经网络在单因子短时间预测上准确性更高,递归神经网络预测值与实测值相关性更强。在工程运用中,需根据地域时空特点、数据资料的丰富度与预测值评估指标选择合适的方法。     

基于强化学习的锚泊辅助动力定位系统智能决策研究

针对半潜平台锚泊辅助动力定位系统的最优定位点问题,设计了基于强化学习中深度神经网络的Q学习(DQN)控制策略的锚泊辅助动力定位的智能决策系统。该决策系统中DQN方法与比例—积分—微分(PID)控制方法相结合使用,实现系统优化。在基于机器人操作系统(ROS)平台的动力定位时域模拟程序中进行数值仿真,仿真结果验证了该系统在定位点决策问题上的可靠性和有效性,从而使半潜平台在面对未知海况时,均能寻找到最优定位点,在保证锚泊辅助动力定位系统可靠性的同时降低功率消耗,提高经济性。