Ekman模型时间变化风应力系数的伴随参数估计

本文基于时间分布参数设置,利用伴随同化方法,反演了Ekman模型中随时间变化的风应力拖曳系数,并在孪生实验和实际实验中对该方法进行了验证。在孪生实验中,研究了参数反演结果对不同影响因素的响应,包括:风速分布、风应力系数分布、风应力系数初始猜测值、风应力系数独立变量个数、观测数据误差和观测的深度。孪生实验结果验证了伴随同化方法反演Ekman模型中时变风应力系数的有效性,具体包括如下五个方面结论:1)不同风速分布下均能成功反演出不同风应力拖曳系数分布; 2)反演结果对初始猜测值较为敏感,风应力系数初始猜测值越接近给定值,反演结果越好;3)风应力系数独立点个数的选取会显著影响反演结果,合理的选择有利于提高反演效率及减小观测数据误差;4)观测误差能够影响反演结果,观测数据误差在20%以下时能取得合理的反演结果; 5)反演结果对观测数据的表层和次表层流速更为敏感,这是由Ekman流的物理性质决定的。实际实验,利用百慕大锚系试验平台的风速和流速数据,去除周期性潮流和地转流成分后得到Ekman流成分,并作为观测输入到该同化模型,反演出了适用于该区域和该时段的随时间变化的风应力系数。通过比较模拟流速和观测流速,证明利用伴随同化方法能从实测数据中反演出合理的时变风应力系数,对于海洋模型风应力系数的确定是一项有益的尝试。     

Experimental study on a sampling technique based on a freeze-sediments valve for deep-sea microorganism

Deep-sea microorganism research has become a hot topic in life science, owing to its special value in high pressure and low temperature environment. If not kept the in-situ low temperature and high pressure, it will cause the microorganism to lose its activity and lead to a waste of resources. A freeze-sediments valve presents a promising solution to maintain the low temperature and high pressure in situ, improve the success rate of sampling and shorten the sampling time. A series of experiments were conducted to characterize the freeze-sediments valve, including the sampling time without leakage, adaptation to different sediments and the effect of the length, etc. Most important of all, temperature change of sediments is observed in all experiments. Experimental results indicated that the freeze-sediments valve was formed in 5 min without leakage and the freeze-sediments valve could be applied to different sediments without limitations on sediments type and mineral content. Considering the pressure retaining capacity and formation time of freeze-sediments valve, it was found that the freeze-sediments valve with a length of 60–80 mm has good pressure retaining capacity whose value was 41.3–48.4 Mpa and short formation time which was 5 min. Precooling the sampler could be used to shorten the sampling time in the actual environment application. This sampling technique can play a vital role in maintaining in-situ low temperature at −2 to −4 ° Celsius in all the experiments. According to the tests, the sampling technique based on the freeze-sediments valve had the potential to solve the problem of keeping in-situ temperature, low sampling success rate and long sampling time in the future.     

Stochastic bivariate time series models of waves in the North Sea and their application in simulation-based design

In this paper, we present and evaluate three long-term wave models for application in simulation-based design of ships and marine structures. Designers and researchers often rely on historical weather data as a source for ocean area characteristics based on hindcast datasets or in-situ measurements. The limited access and size of historical datasets reduces repeatability of simulations and analyses, making it difficult to assess the sampling variability of performance and loads on marine vessels and structures. Markov, VAR and VARMA wave models, producing independent long-term time series of significant wave height (H_s) and spectral peak period (T_p), is presented as possible solutions to this problem. The models are tested and compared by addressing how the models affect interpretation of design concepts and the ability to replicate statistical and physical characteristics of the wave process. Our results show that the VAR and VARMA models perform sufficiently in describing design performance, but does not capture the physical process fully. The Markov model is found to perform worst of the tested models in the applied tests, especially for measures covering several consecutive sea states.     

Large-time response of ice cover to a load moving along a frozen channel

Unsteady response of an ice cover in a channel with vertical walls is studied for large times. The ice deflection is caused by a load moving along the frozen channel at a constant speed. The ice cover is modelled as a thin elastic plate clamped to the walls of the channel. The time-dependent problem is solved by using the Fourier transform along the channel and the method of separating variables. In the system moving along the channel together with the load, the large-time deflection of the ice cover consists of steady deflection and standing waves in front and behind the load. The number of waves, their frequencies and wavenumbers depend on the speed of the load and the values of the critical speeds for the channel. The number of the waves and their amplitudes are calculated for a given load and its speed. The maximum stress in the ice as a function of the load speed is estimated.     

时间尺度重构EEMD-GRNN改进模型预测PM2.5的研究

合理构建PM_2.5浓度预测模型是科学、准确地预测PM_2.5浓度变化的关键。传统PM_2.5预测EEMD-GRNN模型具有较好的预测精度,但是存在过于关注研究数据本身而忽略其物理意义的不足。本研究基于南京市2014-2017年PM_2.5浓度时间序列数据,分析PM_2.5浓度多尺度变化特征及其对气象因子和大气污染因子的尺度响应,基于时间尺度重构进行EEMD-GRNN模型的改进与实证研究。南京市样本数据PM_2.5浓度变化表现为明显的天际尺度和月际尺度,从重构尺度（天际、月际）构建GRNN模型更具有现实意义;同时,PM_2.5对PM₁₀、NO₂、O₃、RH、MinT等因子存在多尺度响应效应,以其作为GRNN模型中的输入变量更具有时间序列上的解释意义。改进后的EEMD-GRNN模型具有更高的PM_2.5浓度预测精度,MAE、MAPE、RMSE和R²分别为6.17、18.41%、8.32和0.95,而传统EEMD-GRNN模型的模型有效性检验结果分别为8.37、27.56%、11.56、0.91。对于高浓度天（PM_2.5浓度大于100 μg/m³）的预测,改进模型更是全面优于传统EEMD-GRNN模型,MAPE为12.02%,相较于传统模型提高了9.03%。     

GNSS变形监测网短基线时间序列噪声特性分析

利用时间跨度为5 a的GNSS短基线时间序列对噪声特性进行分析,发现长周期噪声分量(随机游走噪声)。选取最优噪声模型,评估不同噪声模型对测站周期振幅和线性速度估值的影响。结果表明,短基线时间序列中有色噪声应顾及闪烁噪声和随机游走噪声,对于表现出随机游走噪声的分量,可能与测站的真实运动有关;假设只有白噪声时求得的速度估值与最优噪声模型下求得的速度估值存在0.4～0.6 mm/a的偏差,对周期振幅的影响可以忽略。     

基于位置大数据的国庆假期青藏高原人群分布时空变化模式挖掘

人类活动是引起青藏高原生态环境发生改变的重要因素。很多学者对青藏高原史前人类活动和近几十年的人口分布格局与人口流动开展了大量研究,但有关人群时空分布的精细尺度研究相对缺乏。海量的位置大数据为认识高原人群短期的动态变化提供了新途径。本文利用手机定位数据、人口迁徙数据等高时空分辨率的位置大数据,通过时间序列分解方法和基于统计检验的异常判别方法,分析了2017年国庆期间青海与西藏的人群分布时空变化特征,并探讨了假期旅游行为及人口迁徙与变化特征之间的关系。研究结果显示：① 在省级和城市整体尺度上,定位请求量的假期变化在时间上呈现先降后升的“潮汐”变化模式;② 在精细网格尺度上,西宁和拉萨城市及周边地区的人群分布变化在空间上呈现中心跌、周边涨的“离心化”变化模式。国庆假期人们向城市周边热门景点移动聚集的旅游行为和城市之间的人口迁徙都是导致西宁和拉萨周边地区定位请求量上涨的重要潜在原因,而两座高原城市中心定位请求的下跌不仅与人口迁徙有关,还与假期人类日常行为及定位请求频次的变化等因素有关。通过位置大数据挖掘节庆假期人群分布的时空变化,不仅加深了对高原人口分布格局与流动变化的认识,也为高原城镇化与生态保护的精细化管理与决策提供支撑。     

南海海洋地磁固定观测站建设初步研究

为填补我国海洋地磁固定观测的空白,支撑科学研究、防灾减灾、海洋资源开发,维护国家的海洋权益和履行我国在南海海域负有的国际责任,急需在南海建设岛礁和海底地磁固定观测站。通过初步分析国外海洋地磁观测站案例、南海岛礁基础设施现状以及海底观测网计划实施情况,认为南海海洋地磁固定观测站建设的条件已经成熟。文中给出了合理空间布局、分步实施、因地制宜及智能化观测的建议。     

实时卫星钟差基准精化方法

在基于精密单点定位(PPP)的授时方法中,卫星钟差产品的高精度时间基准至关重要.针对实时卫星钟差产品时间基准不够稳定的问题,本文采用一组具有原子钟外部输入的国际全球卫星导航系统(GNSS)服务(IGS)跟踪站建立了顾及原子钟变化特性的基准精化方法.该方法首先采用阿伦方差对不同的IGS跟踪站外接原子钟进行稳定度分析,挑选出一组稳定度高的原子钟用以精化时间基准.在此基础上,利用阿伦方差分析各台原子钟的噪声参数特征,并确定不同原子钟之间的权比关系.最终,建立时间基准改正量的随机模型,并计算出精化后的时间基准.通过实例验证表明:与IGS事后精密钟差产品定义的时间基准比较,改正后的实时钟差基准单天内的标准差(STD)优于0.1 ns,相比于改正前最高提升了93%.同时,基准改正后的天内万秒稳达到10^-15量级,实现了一个量级的提高.此外,通过相对钟差精度的分析,表明钟差基准修正不影响PPP的定位精度.     

基于支持向量机的GNSS时间序列预测

在深度学习的理论框架下,针对预测全球卫星导航系统(GNSS)时间序列,传统的经验风险最小化预测模型误差大精度低,泛化性能差且对历史数据的经验依赖大的问题.提出一种采用结构风险最小化原则的基于支持向量机(SVM)的时间序列预测模型.通过和多层的BP神经网络预测模型预测效果比较,结果证明SVM预测模型拥有更好的时间序列预测效果.