基于CryoSat-2卫星测高数据的北极海冰体积估算方法

近30年来,北极海冰正发生着剧烈的变化。海冰体积是量化海冰变化的重要指标之一。本文以2015年CryoSat-2卫星测高数据和OSI SAF海冰类型产品为基础。提取了浮冰出水高度、积雪深度、海冰密集度、海冰类型等属性信息,通过数据内插、投影变换、栅格转换、空间重采样等工作将海冰属性信息统一为25 km×25 km分辨率的栅格数据集。根据流体静力学平衡原理,逐个估算栅格像元对应的海冰厚度值,将其与对应的海冰面积相乘,估算了北极海冰密集度大于75%海域的海冰体积,并分析了海冰厚度和体积的月变化和季节变化特征。用NASA IceBridge海冰厚度产品对反演的海冰厚度进行验证。结果表明二者相关系数为0.72,有较高的一致性。北极海冰平均厚度春季最大,夏季最小,分别约为2.99 m和1.77 m,最厚的海冰集中在格陵兰沿岸北部和埃尔斯米尔半岛以北海域。多年冰平均厚度大于一年冰。冬季海冰体积最大,约为23.30×103 km3,经过夏季的融化,减少了近70%。一年冰体积季节波动较大,而多年冰体积相对稳定,季节变化不明显。     

Modelling the annual cycle of landfast ice near Zhongshan Station,East Antarctica

A high resolution one-dimensional thermodynamic snow and ice(HIGHTSI) model was used to model the annual cycle of landfast ice mass and heat balance near Zhongshan Station, East Antarctica. The model was forced and initialized by meteorological and sea ice in situ observations from April 2015 to April 2016. HIGHTSI produced a reasonable snow and ice evolution in the validation experiments, with a negligible mean ice thickness bias of(0.003±0.06) m compared to in situ observations. To further examine the impact of different snow conditions on annual evolution of first-year ice(FYI), four sensitivity experiments with different precipitation schemes(0, half, normal, and double) were performed. The results showed that compared to the snow-free case,the insulation effect of snow cover decreased bottom freezing in the winter, leading to 15%–26% reduction of maximum ice thickness. Thick snow cover caused negative freeboard and flooding, and then snow ice formation,which contributed 12%–49% to the maximum ice thickness. In early summer, snow cover delayed the onset of ice melting for about one month, while the melting of snow cover led to the formation of superimposed ice,accounting for 5%–10% of the ice thickness. Internal ice melting was a significant contributor in summer whether snow cover existed or not, accounting for 35%–56% of the total summer ice loss. The multi-year ice(MYI)simulations suggested that when snow-covered ice persisted from FYI to the 10 th MYI, winter congelation ice percentage decreased from 80% to 44%(snow ice and superimposed ice increased), while the contribution of internal ice melting in the summer decreased from 45% to 5%(bottom ice melting dominated).     

基于双极化SAR图像的U-Net海冰多分类模型

北极多年冰在近几十年有明显的减少趋势,与北极海冰的厚度、体积和夏季最小海冰范围的减少密切相关。合成孔径雷达(synthetic aperture radar, SAR)具有全天时、全天候成像能力,基于SAR卫星影像的海冰分类对监测北极多年冰具有重要意义。基于深度学习U-Net模型,以SAR图像的双极化信息为模型输入,构建了像素级的海水、一年冰和多年冰多分类模型。与已有SAR图像海冰分类方法(支持向量机、随机森林和卷积神经网络)进行对比,基于双极化SAR图像的U-Net海冰分类模型的准确率、平均重叠度和Kappa系数,分别达到了90.73%、0.831和0.849,优于其他对比模型,分别提升了4.08%～19.04%, 0.063～0.321和0.111～0.335。此外,针对SAR图像水平-垂直极化(horizontal-vertical polarization, HV)有明显的条状热噪声和水平-水平极化(horizontal-horizontal polarization, HH)受入射角效应而亮度不均匀的特点,设计敏感性实验,研究HV噪声、入射角和灰度共生矩阵(gray leve...     

利用2012年中国北极考察现场观测数据和卫星遥感资料评估海洋2号卫星扫描辐射计数据反演北极海冰密集度的研究

A retrieval algorithm of arctic sea ice concentration(SIC) based on the brightness temperature data of "HY-2" scanning microwave radiometer has been constructed. The tie points of the brightness temperature were selected based on the statistical analysis of a polarization gradient ratio and a spectral gradient ratio over open water(OW), first-year ice(FYI), and multiyear ice(MYI) in arctic. The thresholds from two weather filters were used to reduce atmospheric effects over the open ocean. SIC retrievals from the "HY-2" radiometer data for idealized OW, FYI, and MYI agreed well with theoretical values. The 2012 annual SIC was calculated and compared with two reference operational products from the National Snow and Ice Data Center(NSIDC) and the University of Bremen. The total ice-covered area yielded by the "HY-2" SIC was consistent with the results from the reference products. The assessment of SIC with the aerial photography from the fifth Chinese national arctic research expedition(CHINARE) and six synthetic aperture radar(SAR) images from the National Ice Service was carried out. The "HY-2" SIC product was 16% higher than the values derived from the aerial photography in the central arctic. The root-mean-square(RMS) values of SIC between "HY-2" and SAR were comparable with those between the reference products and SAR, varying from 8.57% to 12.34%. The "HY-2" SIC is a promising product that can be used for operational services.     

一种融合纹理特征与NDVI的随机森林海冰精细分类方法

海冰的精准分类对于掌握海冰生长发育状况,保障航海安全等具有重要意义。由于受数据源和分类方法等影响,使得海冰分类精度提高受限。本文面向高空间分辨率的光学遥感影像,提出了一种融合纹理特征和归一化差分植被指数（NDVI）的海冰精准分类方法,运用随机森林分类器构建海冰分类方法。以青岛胶州湾为实验区,高分二号（GF-2）为实验数据,进行了海冰类型提取,并与其他分类方法进行对比。结果显示:针对GF-2高分辨率光学遥感数据,融合纹理特征和NDVI的随机森林方法,相比于传统的随机森林、支持向量机、自动决策树和融合纹理特征的最大似然分类方法,总体分类精度分别提高13.70%、11.60%、19.22%、29.37%。Kappa系数分别提高0.16、0.13、0.22、0.44。相比于融合纹理特征和归一化水指数（NDWI）的随机森林方法,总体分类精度提高了9.67%,Kappa系数提高了0.09。这表明本文构建的海冰分类方法可有效提高海冰分类精度,为海冰的精确分类提供了一种有效的技术手段。     

结合纹理特征的SVM海冰分类方法研究

海冰分类是遥感监测领域中的重要应用之一，海冰分类的准确性对于评估海冰冰情、保证航海安全和开辟北极航道具有重要的意义。针对海冰分类问题，本文选用Sentinel-1遥感数据，结合纹理特征分析，提出了一种改进的SAR海冰分类方法。该方法选用灰度共生矩阵提取特征值，通过实验得到适宜用于海冰分类的多特征组合，在此基础上利用支持向量机开展SAR海冰类型的分类研究。实验结果表明，该方法可以实现对海冰SAR图像中一年冰、多年冰和海水3种类型识别，与传统的海冰分类方法神经网络和最大似然法相比较，使用SVM分类方法，结合纹理特征开展海冰类型监测是可行的，同时也表明多特征组合有利于提高SAR图像的分类精度，从而验证了本方法的有效性，为海冰分类提供了一种新思路。     

Estimation of Sea Ice Freeboard from SARAL/AltiKa Data

An attempt has been made to derive sea ice freeboard from Ka-band Altimeter (SARAL/AltiKa) over Arctic region for 15 March–15 April 2013 (spring) and 15 September–15 October 2013 (autumn). A waveform template matching technique is employed for classification of leads and floe pixels. The estimated sea ice freeboards were found in close agreement with “Operation IceBridge quick look” freeboards (RMSD = 0.30 m). The differences between the two freeboards were largely due to snow layer over sea ice (R = 0.8). The estimated freeboards were of the order of 0.08–0.15 m during the two seasons.     

基于最优特征集的HY-1C卫星海岸带成像仪影像海冰分类方法研究

基于海洋一号C（HY-1C）卫星海岸带成像仪（CZI）遥感影像,提出了一种基于最优特征集的支持向量机海冰分类方法。分别提取CZI影像的光谱特征和纹理特征,采用基于距离可分性的判据进行特征选择,得到最优特征集,以最优特征集作为支持向量机分类器输入,分别对3期辽东湾海域CZI影像开展海冰分类实验和结果分析。结果表明:本文方法得到的海冰分类结果精度优于仅利用光谱特征或纹理特征的海冰分类精度;基于本文方法的3期影像的海冰分类精度均较高,2020年12月19日、2021年1月10日与2021年1月16日的海冰分类总体精度分别为93.67%、91.75%、84.89%,均在80%以上;利用海冰分类结果图估算海冰面积,发现3期辽东湾海冰面积依次增大,最大约为11 998.98 km2。     

北极海冰输出区域海冰运动变化特征及其与北极航道适航性的关系

随着北极地区气候变暖的加剧，北极海冰正在急剧消融，海冰的减少增加了北极地区航道的适航性。本文利用遥感数据反演得到的海冰运动产品对北极海冰输出区域以及东北航道以北区域的海冰运动特征进行了量化。结果显示，从北极中央海域向弗拉姆海峡以及格陵兰海流出海冰的南向位移量呈现出显著增长趋势，海冰的平均南向位移量在2007-2014年间达到1511 km，是2007年之前（617 km）的两倍以上，反映了北极穿极流（TDS）强度在不断增强。通过长时间序列分析发现，春季东北航道以北区域的海冰北向漂移速度在喀拉海呈现+0.04 厘米/秒/年的显著增长趋势（P<0.05）。海冰北向漂移对于东北航道的开通具有显著的影响，在拉普捷夫海与喀拉海，海冰北向运动速度与航道适航期的决定系数分别达到0.33（P<0.001）和0.15（P<0.05）。东西伯利亚海、拉普捷夫海以及喀拉海存在冰间湖区域的春季海冰面积变化与航道的适航期密切相关，海冰的北向漂移对拉普捷夫海和喀拉海的海冰面积减少也有显著影响，这说明北向漂移促进了海冰的离岸输送，造成海冰面积减少的同时形成冰间水道或冰间湖促使航道开通。为探究大气环流指数对海冰运动以及东北航道适航期的影响，本文利用大气再分析数据计算了中央北极指数（CAI）和北极大气偶极子异常（DA）指数。相关性分析表明，CAI比DA更能解释东北航道的适航期，而且CAI能够解释北极海冰输出区域海冰南向位移量变化的45%。最近10年，夏季正相位的CAI进一步加强，通过加强海冰离岸输运和冰间湖活动加剧了东北航道区域海冰变薄及其强度变弱，从而促进了东北航道的开通。     

黄河口邻近海域海冰变化特征及机制分析

黄河口邻近海域海冰是渤海海冰的一部分,为了解其独特的变化特征及机制,本研究基于北海预报中心提供的黄河口周边海洋台站观测数据以及CMEMS (Copernicus Marine Environment Monitoring Service)全球海冰密集度再分析数据,使用统计分析和两种滑动相关分析,结合小波相干方法及大气过程的影响,得到长期变化分析的结果。黄河口冰情在1979—2020年间整体呈减轻趋势(–0.25%/a),显然其直接因素为局地温度整体升高;海冰密集度与黄河径流量呈明显正相关,相关系数为0.46,其原因为径流增大导致盐度降低,海冰增加;与北极涛动指数(AOI,Arctic Oscillation Index)呈明显负相关,相关系数为–0.44,因为当北极涛动为正位相时,东亚大槽强度减弱,北极冷空气南侵受阻隔,冬季黄河口的整体气温升高,导致海冰减少;1997年和2016年左右与北极涛动的相关性都出现了显著正异常,其原因为两次强厄尔尼诺事件的影响,同时海冰密集度在1985年左右的跃变可能与AOI和黄河径流量的突变有关。短期变化分析的结果显示:从2010年和2020年冬季逐日的典型寒潮过程与海冰密集度的变化分析可知,海冰与前6 d负积温的相关性最大,平均相关系数为–0.77,寒潮的出现时间、强度及间隔,控制海冰的生成,而整体气温的低频变化控制海冰的维持和发展。     

误差订正对2018年夏季次季节尺度海冰预测的作用

北极海冰次季节尺度预测在针对破冰船和商船的实际服务中十分重要，但常常受制于气候模拟的模拟能力。本研究提出了一种误差订正方法并分别应用到两个气候模式：海洋一所地球系统模式（FIOESM）和美国国家环境预报中心（NCEP）的气候预报系统（CFS），来改善北极海冰60天尺度的预测。本研究的预测工作是中国第9次北极科学考察和2018年夏季中远集团北极商业航行的业务化海冰服务保障的重要部分。模式起报时间分别是2018年7月1日、8月1日和9月1日，预报时效均是60天。结果显示，FIOESM整体上低估了海冰密集度的数值，平均偏差可达30%。误差订正对海冰密集度（SIC）的均方根偏差（RMSE）的改进比例可达27%，对海冰外缘线（SIE）的整体偏差（IIEE）的改进比例为10%。而对于CFS，SIE在边缘区域的过高估计是其主要特点。误差订正导致了SIC的RMSE改进了7%，而对SIE的IIEE改进了17%。在海冰范围预测方面，FIOESM预测的最小范围数值和时间点都和观测接近，而CFS的预测结果偏差较大。另外和其他S2S模式的结果比较发现，本研究提出的误差订正方法对存在较大偏差的预测结果改进更为有效。     

Bayesian classification of the ice cover of the Arctic seas

A classification of sea ice in the Arctic by age (multiyear; first-year; and first-year deformed ice, nilas, etc.) is developed based on the Bayesian approach using satellite radar data and taking into account regional peculiarities of these types of ice for different sectors of the Arctic. Estimations of a priori probabilities for each ice type, which are required for the use of the Bayesian classification, are obtained by the analysis of ice charts in the Arctic seas developed at the AARI in 2008–2013 using satellite data. A posterior probabilities are estimated visually by an expert. Types of sea ice distinguished by the expert on satellite images make it possible to create sample values of the radar-scattering cross section (RSCS). Examples of the proposed Bayesian classification of ice in the Laptev Sea according to Envisat satellite data are given.     

The effects of resolving the Canadian Arctic Archipelago in a finite element sea ice model

Though narrow straits may have a strong influence on the large-scale sea ice mass balance, they are often crudely represented in coarse resolution sea ice models. Unstructured meshes, with their natural ability to fit boundaries and locally increase the mesh resolution, propose an alternative framework to capture the complex oceanic areas formed by coasts and islands. In this paper, we develop a finite element sea ice model to investigate the sensitivity of the Arctic sea ice cover features to the resolution of the narrow straits constituting the Canadian Arctic Archipelago. The model is a two-level dynamic-thermodynamic sea ice model, including a viscous-plastic rheology. It is run over 1979–2005, forced by daily NCEP/NCAR reanalysis data. Confronting qualitatively numerical experiments with observations shows a good agreement with satellite and buoys measurements. Due to its simple representation of the oceanic interactions, the model overestimates the sea ice extent during winter in the southernmost parts of the Arctic, while the Baffin Bay and Kara Sea remain ice-covered during summer. In order to isolate the benefits from resolving the Canadian Arctic Archipelago, a numerical experiment is performed where we artificially close the archipelago. Focusing on the large-scale sea ice thickness pattern, no significant change is found in our model, except in the close surroundings of the archipelago. However, the local and short-term influences of the ice exchanges are nonnegligible. In particular, we show that the ice volume associated to the Canadian Arctic Archipelago represents 10% of the Northern Hemisphere sea ice volume and that the annual mean ice export towards Baffin Bay amounts to 125 km³ yr⁻¹, which may play an important role on the convective overturning in the Labrador Sea.     

CICE5.0与BCC_CSM2.0模式的耦合及对北极海冰的模拟评估

本文将美国Los Alamos国家实验室发展的最新海冰模式CICE5.0引入国家气候中心气候系统模式BCC_CSM2.0,替代原有的海冰模式SIS,形成一个新的耦合模式。在此基础上,评估新耦合模式对1985-2009年北极海冰的模拟性能,检验引入CICE5.0后对耦合模式中北极海冰、海洋和大气模拟结果的改进。结果表明,引入CICE5.0后,模式能较好地模拟出北极海冰的空间分布、季节以及年际变化特征。相比于旧版本耦合模式,新耦合模式模拟的北极多年冰增多、一年冰减少,同时,海冰增厚、海冰流速减慢,模拟效果得到显著改进,对波弗特涡流模拟的改善尤为明显。进一步分析发现,相比于SIS,CICE5.0对北极海冰特别是海冰厚度模拟性能的提升,在耦合进入BCC_CSM2.0后,会触发冰-温的正反馈机制,改进了模式对海平面气压场、表层气温和海表温度的模拟,由此进一步提高了模式对北极海冰的模拟能力。     

地球系统模式FIO-ESM对北极海冰的模拟和预估

评估了地球系统模式FIO-ESM(First Institute of Oceanography-Earth System Model)基于CMIP5(Coupled Model Intercomparison Project Phase 5)的历史实验对北极海冰的模拟能力,分析了该模式基于CMIP5未来情景实验在不同典型浓度路径(RCPs,Representative Concentration Pathways)下对北极海冰的预估情况。通过与卫星观测的海冰覆盖范围资料相比,该模式能够很好地模拟出多年平均海冰覆盖范围的季节变化特征,模拟的气候态月平均海冰覆盖范围均在卫星观测值±15%范围以内。FIO-ESM能够较好地模拟1979-2005年期间北极海冰的衰减趋势,模拟衰减速度为每年减少2.24×104 km2,但仍小于观测衰减速度(每年减少4.72×104 km2)。特别值得注意的是:不同于其他模式所预估的海冰一直衰减,FIO-ESM对21世纪北极海冰预估在不同情景下呈现不同的变化趋势,在RCP2.6和RCP4.5情景下,北极海冰总体呈增加趋势,在RCP6情景下,北极海冰基本维持不变,而在RCP8.5情景下,北极海冰呈现继续衰减趋势。     

协同主动学习和半监督方法的海冰图像分类

海冰遥感光谱影像分类中标签样本难以获取,导致海冰分类精度难以提高,但是大量包含丰富信息的未标签样本却没有得到充分利用,针对这种情况,提出一种协同主动学习和半监督学习方法用于海冰遥感图像分类。在主动学习部分,结合最优标号和次优标号、自组织映射神经网络以及增强的聚类多样性算法来选择兼具不确定性和差异性的样本参与训练;在半监督学习部分,利用直推式支持向量机,并且融合主动学习思想从大量未标签样本中选取相对可靠且包含一定信息量的样本进行迭代训练;然后协同主动学习分类结果和半监督分类结果,通过一致性验证保证所加入伪标签样本的正确性。为了验证方法的有效性,分别采用巴芬湾地区30 m分辨率的Hyperion高光谱数据(验证数据为15 m分辨率的Landsat-8数据)和辽东湾地区15 m分辨率的Landsat-8数据(验证数据为4.77 m分辨率的Google Earth数据)进行海冰分类实验。实验结果表明,相对其他传统方法,该协同分类方法可以在只有少量标签样本的情况下,充分利用大量未标签样本中包含的信息,实现快速收敛,并获得较高的分类精度(两个实验的总体精度分别为90.003%和93.288%),适用于海冰遥感图像分类。     

High Arctic sea ice conditions influence marine birds wintering in Low Arctic regions

Ocean climate change is having profound biological effects in polar regions. Such change can also have far-reaching downstream effects in sub-polar regions. This study documents an environmental relationship between High Arctic sea ice changes and mortality events of marine birds in Low Arctic coastal regions. During April 2007 and March 2009, hundreds of beached seabird carcasses and moribund seabirds were found along the east and northeast coasts of Newfoundland, Canada. These seabird “wrecks” (i.e. dead birds on beaches) coincided with a period of strong, persistent onshore winds and heavily-accumulated sea ice that blocked bays and trapped seabirds near beaches. Ninety-two percent of wreck seabirds were Thick-billed Murres (Uria lomvia). Body condition and demographic patterns of wreck murres were compared to Thick-billed Murres shot in the Newfoundland murre hunt. Average body and pectoral masses of wreck carcasses were 34% and 40% lighter (respectively) than shot murres, indicating that wreck birds had starved. The acute nature of each wreck suggested that starvation and associated hypothermia occurred within 2–3 days. In 2007, first-winter murres (77%) dominated the wreck. In 2009, there were more adults (78%), mostly females (66%). These results suggest that spatial and temporal segregation in ages and sexes can play a role in differential survival when stochastic weather conditions affect discrete areas where these groups aggregate. In wreck years, southward movement of Arctic sea ice to Low Arctic latitudes was later and blocked bays longer than in most other years. These inshore conditions corresponded with recent climate-driven changes in High Arctic ice break-up and ice extent; coupled with local weather conditions, these ice conditions appeared to be the key environmental features that precipitated the ice-associated seabird wrecks in the Low Arctic region.     

北极海冰密集度数值预报及验证评估研究

本文系统地评估了国家海洋环境预报中心于我国第七次北极科学考察期间开展的北极海冰密集度数值预报结果。该预报系统基于麻省理工大学通用环流模式，并采用牛顿松弛逼近（Nudging）资料同化方法，计算输出未来1~5 d的北极海冰密集度预报产品。本文将数值预报结果同卫星观测的海冰密集度、再分析资料和"雪龙"号第七次北极考察期间观测的海冰密集度数据进行了对比分析。结果表明，预报的北极海冰密集度小于卫星观测值，24 h、72 h和120 h预报结果的偏差分别为-2.7%、-3.1%和-3.2%；数值产品的预报技巧好于气候态结果和惯性预报，但是在海冰出现快速融化或冻结时，基于Nudging同化的数值预报技巧仍有不足。另外，相比船测数据，数值预报结果在海冰边缘区的偏差相对较大，24 h、72 h和120 h预报结果的偏差分别为8.8%、12.0%和14.5%。     

1979-2012年北极海冰运动学特征初步分析

利用美国冰雪数据中心(NSIDC)发布的海冰速度和范围数据,本文分析了1979—2012年间北极海冰的运动学特征,以及北极海冰运动与分布范围演变之间的关系。结合欧洲中期天气预报中心(ECMWF)发布的2007和2012年高分辨率的气压场、风场数据,探讨了北极风场和气压场与海冰运动、辐散辐合和海冰面积的关系。结果表明,在1979-2012年间北极海冰平均运动速度呈显著增强的趋势,冬季海冰平均运动速度增加趋势明显强于夏季;北极、波弗特-楚科奇海域和弗拉姆海峡的冬、夏季海冰平均运动速度的增加率分别为2.1%/a和1.7%/a、2.0%/a和1.6%/a以及4.9%/a和2.2%/a。1979-2012年北极海冰平均运动速度和范围的相关性为-0.77,二者存在显著的负相关关系。北极冬季和夏季风场的长期变化趋势与海冰平均运动速度的变化趋势一致,冬季和夏季的相关系数分别为0.50和0.48。风场和气压场对海冰的运动、辐散及重新分布发挥着重要作用。2007年夏季,第234~273天波弗特海域一直被高压系统控制,波弗特涡旋加强,使得波弗特海域海冰聚集在北极中央区;顺时针的风场促使海冰向格陵兰岛和加拿大北极群岛以北聚合。2012年,白令海峡和楚科奇海域处于低压和高压系统的交界处,盛行偏北风,海冰从北极东部往西部输运,加拿大海盆的多年海冰因离岸运动而辐散,向楚科奇海域的海冰输运增加,受太平洋入流暖水影响,移入此区域的海冰加速融化,从而加剧海冰的减少。     

基于卫星观测的2003-2013年北极海冰体积变化估算

北极海冰正处于快速减退时期，北极海冰体积变化是全球气候变化的重要指示因子。本文利用两种卫星高度计数据（ICESat和CryoSat-2）反演得到的海冰厚度数据，结合星载辐射计提取的海冰密集度数据以及海冰年龄数据，估算了近期的北极海冰体积以及一年冰和多年冰体积变化。CryoSat-2观测时段（2011-2013年）与ICESat观测时段（2003-2008年）相比，北极海冰体积在秋季（10-11月）和冬季（2-3月）分别减少了1 426 km3和412 km3。其中，秋季和冬季的一年冰的体积增加了702 km3和2 975 km3。相反，多年冰分别减少了2 108 km3和3 206 km3。多年冰的大量流失是造成北极海冰净储量下降的主要原因。