基于小波分解和最小二乘支持向量机的ENSO集成预测

用小波分解和最小二乘支持向量机相结合的方法,建立了ENSO的集成预报模型。该方法将复杂海温系统分解为相对简单的带通分量信号,然后建立分量信号的独立预报模型,最后对预报结果进行集成。试验结果表明,模型在保留预报对象主要特征的前提下,有效地降低了预报难度,集成预报准确率和预报时效均较传统方法有明显的改进和提高。     

热带太平洋次表层三维海温距平场的EOF分析

应用太平洋次表层海温距平资料构造了一个立体的西低东高的四级阶梯模型,使用EOF方法对此模型进行时空分解,重点讨论了分解结果与ENSO循环的关系,并对Nino 3距平指数进行试预报,结果表明,用 EOF分解的第一主分量代表暖池-厄尔尼诺(拉尼娜)模态,第二主分量代表次表层温度距平的东西运移模态,厄尔尼诺事件正是东西运移模态突变的结果.时滞相关分析估算,一次ENSO循环的平均周期约为41个月,厄尔尼诺、拉尼娜模态与温度距平的东西运移模态的位相差平均约为9.7个月.应用逐步回归方法得到超前Nino 3距平指数3、6和12个月的3个预报方程.预报结果表明,第一、第二时间系数对Nino 3距平指数均具有一定的预报价值,预报时效可达1 a左右.     

基于EEMD和ARIMA的海温预测模型研究

类型丰富、时空分辨率高的海洋探测数据,为信号分解和机器学习算法的应用提供了可能。本文针对如何建立有效的海温预测模型这一问题,使用高时空分辨率的海表温度(SST)融合产品,引入信号处理领域的集合经验模态分解(EEMD)和机器学习领域的自回归积分滑动平均模型(ARIMA)。首先利用最适于分解自然信号的EEMD方法,将海温数据分解成多个确定频率的序列;再利用ARIMA分别对各个频率的序列进行预测,最后将各个序列的预测结果进行组合。该方法在丰富数据的支撑下,比以往直接使用海温数据所建立的预测模型精度更高,为更好地进行海温预测提供了新方法。     

基于相空间重构与模糊神经网络耦合的海温预测模型

在海温预报中引入混沌理论,将相空间重构理论与模糊神经网络相结合,建立了海温预测模型.通过相空间重构,把一维海温时间序列拓展为多维序列,而多维序列包含着各态历经的信息,从而挖掘出了丰富的海温变化空间的信息,有利于模糊神经网络的训练.利用建立好的模糊神经网络模型,对海温预报问题进行了建模、训练和预测.实际的预测结果表明,该模型预报精度较高,预测结果可以为业务工作提供一定的参考与借鉴.     

基于机器学习的青岛市区近岸海雾集成预报方法

利用2014-2017年青岛小麦岛海洋站观测资料,采用机器学习方法建立了青岛市区近岸海雾集成预报模型,通过主成分分析方法对预报因子进行了优选.结果表明:采用能见度、风向、风速、气压、露点、气温、海温、气温露点差、气海温差、相对湿度、云量、气温24h变温12个预报因子建立的海雾集成预报模型,对2018年海雾预报的TS评分...     

基于BP神经网络方法的近岸数值海温预报释用技术

为了提高近岸精细化海温预报精度,利用神经网络方法,分析了海温数值预报及观测数据在释用中的作用,研究了定点近岸海温影响因子的最优配置方案,建立了定点海温精细化数值预报释用模型,评估了释用模型性能。误差分析结果显示,数值海温产品及其观测在建模中起到了降低和稳定模型误差的作用;释用模型将定点数值预报的误差从2.2°C减少至0.7°C;预报误差较调训误差略高,但考虑到预报误差的稳定性,数值释用与人工经验预报水平持平,因此,该方法具有十分广阔的拓展空间和应用前景。     

太平洋次表层海温距平的立体EOF分析及其与ENSO的关系

应用太平洋次表层海温距平资料构造了一个立体的西低东高的四级阶梯模型,使用EOF方法对此模型进行时空分解,并讨论与ENSO循环的关系。结果表明,EOF分解的前两个主分量方差贡献达到42.58%,其中第一主分量代表了暖池-厄尔尼诺(拉尼娜)模态;第二主分量代表了次表层温度距平的东西运移模态。后者存在着突变和渐变两个过程,在赤道及其附近,由冷位相变暖位相为渐变过程,暖位相变冷位相是突变过程,在赤道外南北纬10°附近则反之,厄尔尼诺正是东西运移模态突变的结果。时滞相关分析估算,一次ENSO循环的平均周期约为41个月,即3.42a,厄尔尼诺、拉尼娜模态与温度距平的东西运移模态的位相差平均约为9.7个月。运用逐步回归法求得一个对nino3指数预报效果比较理想的预报方程,其中EOF分解的第一时间系数对nino3距平指数有重要的预报价值,第二时间系数对nino3距平指数预报意义虽然不如第一时间系数明显,但仍具有一定的参考价值。     

变分伴随数据同化方法在断面海温数值计算中的应用研究

以二维断面海温分布模型为例,利用海温实际观测数据,将变分伴随方法应用于断面海温初始场的优化。讨论了变分伴随方法的基本思想,分别从模型方程的连续和离散形式出发推导伴随模型系统,并对这两种途径建立的伴随系统之间的相互关系进行了分析。数值试验的结果表明了变分伴随数据同化方法在海温数值计算和数值预报业务中的良好的应用前景。     

基于Stacking机器学习模型的南海北部海温预报

基于Stacking(ET-ET)的机器学习算法,利用美国国家环境预报中心再分析数据和MGDSST海温融合数据,建立了一套高效的海温长期预报方法,并在南海北部海域开展了1 a的表层海温长期预报实验。结果表明：基于Stacking(ET-ET)机器学习模型的表层海温长期预报的均方根误差降至0.52℃,平均绝对百分比误差降至1.58%,明显优于基于支持向量机、人工神经网络和长短期记忆模型的预报结果。     

广西四季温度气候变化与同期海温场异常的关系及预测

利用奇异值分解(SVD)分析广西春(3～5月)、夏(6～8月)、秋(9～11月)和冬(12～2月)四季气温变化与同期北太平洋海温场异常的关系。研究表明:广西各季气温气候变化与同期北太平洋海温场具有较好的相关性,相关性显著区域主要是Nino区、黑潮区和太平洋中东部的夏威夷海区。广西春季、冬季气温与Nino区、黑潮区海表温度异常表现显著的正相关,与夏威夷海表温度异常呈反相关;夏季气温与北太平洋海表温度呈正相关;秋季与黑潮区、亲潮区呈正相关,与Nino区呈反相关。在均生函数预报模型的基础上,提出了均生函数残差模型。实验结果表明,残差模型在拟合效果和预测实验比均生函数预报模型理想,为气候预报提供了一种较为理想而有用的方法。     

海上大直径钢管桩打桩过程中桩周土强度弱化研究

海上大直径钢管桩打桩过程中,桩周土体受到强烈扰动而发生强度弱化,掌握桩周土体强度弱化规律对于准确预测打桩过程、保证工程安全具有重要意义。为研究土体强度弱化规律,开展了环剪试验模拟打桩对桩周土体的扰动,测试土体强度随剪切速率的变化规律,建立了描述土体强度弱化规律的拟合公式,引入到打桩分析软件中。研究结果表明：土体的强度折减程度不仅与土体本身的性质有关还受到土体的埋深和剪切速率的影响,埋深越深土体强度折减程度越低,剪切速率越高土体强度折减越高,在打桩分析中可采用这里推荐的线性折减方法来模拟不同深度处土体强度的折减规律。     

Phylogenetic diversity and phenotypic characterization of cultivable bacterioplankton isolated from polar oceans

A set of 27 marine planktonic bacteria isolated from the polar regions was characterized by 16S rDNA sequencing and physiological and biochemical testing. More than half of these bacteria were positive for caseinase, gelatinase and 13-glucosidase, and could utilize glucose, maltose or malic acid as carbon source for cell growth. Twelve isolates expressed nitrate reduction activities. Except for one antarctic isolate BSwlO175 belonging to Actinobacteria phylum, these isolates were classified as γ-Proteobacteria, suggesting that γ-Proteobacteria dominated in cultivable marine bacterioplankton at both poles. Genus Pseudoalteromonas was the predominant group in the Chukchi Sea and the Bering Sea, and genus ShewaneUa dominated in cultivable bacterioplankton in the Prydz Bay. With sequence similarities above 97%, genus Psychrobacter was found at both poles. These 27 isolates were psychrotolerant, and significant 16S rDNA sequence similarities were found not only between arctic and antarctic marine bacteria （ 〉 99% ）, but also between polar marine bacteria and bacteria from other aquatic environments （ ≥ 98.8% ）, including temperate ocean, deep sea, pond and lake, suggesting that in the polar oceans less temperature-sensitive bacteria may be cosmopolitan and have a bipolar, even global, distribution at the species level.     

Estimation of Sediment Properties during Benthic Impact Experiments

Sediment properties, such as water content and density, have been used to estimate the dry and wet weights, as well as the volume of sediment recovered and discharged, during benthic impact experiments conducted in the Pacific and Indian Oceans. The estimates show that the weights of dry (355-1332 t) and wet sediment (1651-4888 t) recovered during the experiments, are not only a function of the total duration and distance covered during the experiment, but also depend on water content and density of the sediment. Estimation of dry sediment and its volumetric ratio in the discharge, are key parameters for calculating the discharged volume. The estimated volume of wet sediment recovered (1427-4049 m 3 ) as well as discharged (2693-6951 m 3 ) during the experiments, provide important inputs to evaluate the sediment resettlement and migration. Using these estimates, the average depth of excavation on the seafloor, can also been calculated. However, the stages of sediment recovery and discharge are expressed differently in some of the experiments. In order to standardize these, different stages have been identified, and definitions of certain terms have been suggested, for use in the future. The methods of calculating different properties, as well as weight and volume of discharged sediment are described in the paper for use in other applications concerning deep-sea discharges.     

A Tidal Flow Model for the Gulf of Kachchh, India

A vertically integrated model has been used to study the tidal circulation and currents in the Gulf of Kachchh along the west coast of India. The model is fully nonlinear and uses a semiexplicit finite difference scheme to solve the basic hydrodynamic equations on a staggered grid. The model is forced by prescribing the tides along the open boundary of the model domain. The flow is simulated both with and without the presence of the proposed tidal barrage across the Hansthal Creek in the Gulf of Kachchh. The results show a considerable change in the behavior of the tidal flow in the presence of the barrage.     

利用中国GF-3卫星SAR和HY-2A卫星微波散射计的首次台风联合定量观测

<正>The typhoon,as a mature tropical cyclone that develops in the western part of the North Pacific Ocean with high wind speed and heavy rainfall,is one of the most lethal and costly of natural disasters for the densely populated countries of East Asia.It can be easily detected by space-borne sensors operated at microwave,visible or infrared bands(Liu et al.,2014).Synthetic Aperture Radar(SAR)is     

离子交换—原子吸收光谱法测定海水中痕量铜、铅、锌、镉、铁、锰

本文介绍了用国产D401型螯合树脂分离富集海水中铜、铅、锌、镉、铁、锰等痕量元素,并用原子吸收光谱仪测定其含量的方法。讨论了各元素的分离条件选择及干扰元素的影响,并与溶剂萃取法的结果作了比较。各元素检测的定量下限为:铜0.5μg/L、铅0.1μg/L、锌1.0μg/L、镉0.01μg/L、铁2.0μg/L、锰2.0μg/L。方法精密度在4—8％之间,回收率为90—102％。     

Biogeochemistry of Mangrove Soil Organic Matter: a Comparison BetweenRhizophoraandAvicenniaSoils in South-eastern Brazil

Soil core samples fromRhizophora mangleL. andAvicennia schauerianaStapf & Leech forests from south-eastern Brazil were analysed for their total organic matter content and their sugars, amino acid and amino sugars composition. Organic carbon and nitrogen contents were higher inAvicenniathan inRhizophorasoils. The contribution of sugars and amino acids to the total organic carbon pool was constant with depth inRhizophorasoils whereas inAvicenniasoils it increased. Spectral distribution of sugars and amino acids showed a dominance of Ca-affine monomers, particularly acidic amino acids, and the sugar arabinose. Biogeochemical indicators derived from ratios of individual sugar and amino acid monomers confirm previous studies which showed that organic matter in both soils is mainly of mangrove origin. The results further suggest accumulation of organic matter inRhizophorasoils and a continuing degradation of organic matter inAvicenniasoils. The latter may thus release more nutrients to adjacent ecosystems thanRhizophorasoils.     

A review of orange roughy Hoplostethus atlanticus fisheries,estimation methods,biology and stock structure

Orange roughy Hoplostethus atlanticus are unusual fish. They form dense aggregations that have fuelled lucrative fisheries at great depths (600–1 400 m), especially off Namibia, New Zealand and Australia. They are thought to be very long-lived (>100 years, maturity at 22 – 40 years), and to have exceptionally low natural mortality (M = 0.045–0.064 year^?1) and slow growth rates (K = 0.055–0.070 year^?1). In addition, they spawn large eggs and have low fecundity. These factors combine to make orange roughy highly susceptible to overfishing; most stocks are below 30% of pristine levels. Assessments are obtained from indices of catch rate and trawl, acoustic and egg surveys. Acoustic estimates are the most direct, but are confounded by the species' low target strength (?50 to ?53 dB)–attributable to the wax-filled swim bladder. Extracellular wax esters are stored in abundance and comprise mostly mono-unsaturated fatty acids, with low concentrations of the ω-3 fatty acid family. This unusual composition (resultant from the species' diet) ensures neutral buoyancy. Stock separation has been inferred mainly from biological studies, but genetic studies have also found differences among stocks within New Zealand and Australia. Deep-water habitat may be damaged by trawling operations and may take many years to recover, so in some quarters there is a call for a portion of suitable habitat to be set aside for preservation. Although Namibian orange roughy are shallower, smaller and younger than those in other stocks, the Namibian fishery sustained high catches for only a few years before quotas were reduced, from 12 000 to 1 875 tons. Three management lessons are suggested for developing orange roughy fisheries based on the Namibian experience: (1) imposition of catch limits during exploratory fishing; (2) starting the acoustic surveys earlier in the fishery, if possible; (3) greater reliance on trends in catch rate until a survey series has been established.