北极海冰密集度动态系点值ASI反演算法研究

海冰密集度是极区海冰监测的重要因素,使用AMSR-E (Advanced Microwave Scanning Radiometer for Earth Observing System) 89GHz数据ASI反演算法得到的海冰密集度是目前能够获得的分辨率最高的微波数据.在以前的算法中往往使用固定的系点值,本研究实现了动态系点值ASI (the Arctic Radiation And Turbulence Interaction Study (ARTIST) Sea Ice)算法,更重要的是在统计开阔水系点值的时候消除了云对系点值的影响,使得纯水系点值更接近真实状况.得到2010年平均的开阔水和海冰的系点值分别为50.8K和7.8K,通过每天的系点值得到的反演方程在低密集度区增大了海冰密集度,在高密集冰区减小了海冰密集度,从而在一定程度上改善了微波数据的反演准确度.通过和北极区域选取40幅不受云影响的MODIS 500m分辨率宽频大气层顶反照率(broadband TOA albedo)计算的海冰密集度进行了比较验证.结果显示,40个个例中,95%本文的平均差异比使用固定系点值算法产品的小,而且75%的均方根差异比使用固定系点值算法产品的小.     

基于19GHz修正91GHz频段改进的ASI海冰密集度算法

基于数据融合算法思想，利用低频修正高频微波数据提出改进的ASI海冰密集度反演算法，对北极海冰进行反演研究。目前用于整体海冰密集度反演的算法中，使用低频数据的算法受天气影响较弱，但空间分辨率相对较低；而使用高频数据的算法，空间分辨率相对较高，但受天气影响较大，虽然使用天气滤波器处理，能消除那些被误判成海冰的水点，但并没有改变冰点的密集度。改进的ASI算法，利用低频数据（19GHz）修正高频数据（85.5GHz），进而得到修正后的85.5GHz的极化差P''，将P带入ASI算法，最终得到以2008-2016年每年的1月3日SSMIS数据为例的北冰洋整体海冰密集度反演结果。结果表明，改进后的ASI算法得到的总体海冰面积介于ASI与NASA Team两个结果之间；在边缘海冰区，改进后的ASI算法结果与传统的ASI算法结果在海冰面积与平均海冰密集度上都有较大差异，且前者更接近NASA Team算法。因此改进后的ASI算法，在空间分辨率上优于NASA Team算法，在受天气影响程度上更弱于ASI算法，并且有效变了边缘海冰区像元的海冰密集度。     

Dual-polarized ratio algorithm for retrieving Arctic sea ice concentration from passive microwave brightness temperature

We present a new algorithm for retrieving sea ice concentration from the AMSR-E data, the dual-polarized ratio (DPR) algorithm. The DPR algorithm is developed using vertically and horizontally polarized brightness temperatures at the same channel of 36.5 GHz. It depends on the ratio of dual-polarized emissivity, α, which is determined empirically at about 0.92 by remotely sensed brightness temperature in winter and used for the other seasons as well. The ice concentration retrieved by the DPR is compared with those by the NT2 and ABA algorithms. Since the main difference among these algorithms takes place in marginal ice zones, 17 marginal ice zones are chosen. The retrieved ice concentrations in these zones are examined by the ice concentration obtained by the MODIS data. The mean error, root-mean-square error and mean absolute error of the DPR algorithm are relatively better than those from the other two algorithms. The results of this study illustrate that the DPR algorithm is a more accurate algorithm for retrieving sea ice concentration from the AMSR-E brightness temperature, and can be used for operational purposes.     

基于AMSR-E数据的多年冰密集度反演算法研究

In recent years, the rapid decline of Arctic sea ice area(SIA) and sea ice extent(SIE), especially for the multiyear(MY) ice, has led to significant effect on climate change. The accurate retrieval of MY ice concentration retrieval is very important and challenging to understand the ongoing changes. Three MY ice concentration retrieval algorithms were systematically evaluated. A similar total ice concentration was yielded by these algorithms, while the retrieved MY sea ice concentrations differs from each other. The MY SIA derived from NASA TEAM algorithm is relatively stable. Other two algorithms created seasonal fluctuations of MY SIA, particularly in autumn and winter. In this paper, we proposed an ice concentration retrieval algorithm, which developed the NASA TEAM algorithm by adding to use AMSR-E 6.9 GHz brightness temperature data and sea ice concentration using 89.0GHz data. Comparison with the reference MY SIA from reference MY ice, indicates that the mean difference and root mean square(rms) difference of MY SIA derived from the algorithm of this study are 0.65×106 km2 and0.69×106 km2 during January to March, –0.06×106 km2 and 0.14×106 km2 during September to December respectively. Comparison with MY SIE obtained from weekly ice age data provided by University of Colorado show that, the mean difference and rms difference are 0.69×106 km2 and 0.84×106 km2, respectively. The developed algorithm proposed in this study has smaller difference compared with the reference MY ice and MY SIE from ice age data than the Wang's, Lomax' and NASA TEAM algorithms.     

夏季北极密集冰区范围确定及其时空变化研究

研究夏季北极密集冰区的范围变化是了解北极海冰融化过程的重要手段。密集冰区与海冰边缘区之间没有明确的分界线, 海冰密集度在两者之间平滑过渡, 确定密集冰区范围就需确定一个密集度阈值。文中依据分辨率为6.25 km的AMSR-E遥感数据, 发现不同密集度阈值所围范围在密集冰区边缘处的减小存在由快变慢的过程, 同时与周围格点的密集度差异变化在该处最为显著, 对这两个特征进行统计分析, 获得的阈值同为89%, 具有明确的物理意义和合理性。以此为基础, 运用腐蚀算法剔除海冰边缘区, 同时结合连通域法排除小范围密集冰的影响, 进而确定密集冰区的范围。结果表明, 2002-2006年密集冰区覆盖范围较大, 年际变化较小, 2007年以后明显减小, 2010年与2011年相继出现最小值, 其中2011年的范围最小值仅为2006年的64%。密集冰区范围的变化不同于海冰覆盖范围, 是具有独立特性的海冰变化参数, 反映出高密集度海冰区域的变化特征。海冰的融化与海冰边缘区的变化是导致密集冰区范围发生变化的两个主要因素, 受动力学因素的影响, 海冰边缘区发生伸展或收缩, 发生密集冰区与海冰边缘区互相转化。本文提出了一种研究北极海冰变化的新思路, 密集冰区覆盖范围的减小表明北极中央区域高密集度海冰正持续减少。     

2010-2018年北极夏季中国北极科学考察航行期间被动微波遥感海冰密集度与船基目视观测资料的比较

为了更有效地将卫星数据应用于北极航行导航，被动微波（PM）产品的海冰密集度（SIC）与从中国北极科学考察中收集到的船基目视观测（OBS）资料进行了比较。在2010、2012、2014、2016和2018年的北极夏季总共收集了3667组目测数据。PM SIC取自基于SSMIS传感器的NASA-Team（NT）、Bootstrap（BT）以及Climate Data Record（CDR）算法和基于AMSR-E/AMSR-2传感器的BT、enhanced NT（NT2）以及ARTIST Sea Ice（ASI）算法。使用PM SIC的日算术平均值和OBS SIC的日加权平均值进行比较。比较了PM SIC和OBS SIC之间的相关系数，偏差和均方根偏差，包括总体趋势以及在轻度/普通/严重冰况下的情况。使用OBS数据，浮冰尺寸和冰厚对不同PM产品SIC反演的影响可以通过计算浮冰尺寸编码和冰厚的日加权平均值来评估。我们的结果显示相关系数的范围为0.89（AMSR-E/AMSR-2 NT2）到0.95（SSMIS NT），偏差的范围为-3.96%（SSMIS NT）到12.05%（AMSR-E/AMSR-2），均方根偏差的范围为10.81%（SSMIS NT）到20.15%（AMSR-E/AMSR-2 NT2）。浮冰尺寸对PM产品的SIC反演有显著的影响，大多数PM产品倾向于在小浮冰尺寸情况下低估SIC，而在大浮冰尺寸情况下高估SIC。超过30 cm的冰厚对于PM产品的SIC反演没有明显影响。总体来看，在北极夏季，SSMIS NT SIC与OBS SIC之间有着最好的一致性，而AMSR-E/AMSR-2 NT2 SIC与OBS SIC的一致性最差。     

Comparisons of passive microwave remote sensing sea ice concentrations with ship-based visual observations during the CHINARE Arctic summer cruises of 2010–2018

In order to apply satellite data to guiding navigation in the Arctic more effectively, the sea ice concentrations(SIC)derived from passive microwave(PM) products were compared with ship-based visual observations(OBS)collected during the Chinese National Arctic Research Expeditions(CHINARE). A total of 3 667 observations were collected in the Arctic summers of 2010, 2012, 2014, 2016, and 2018. PM SIC were derived from the NASA-Team(NT), Bootstrap(BT) and Climate Data Record(CDR) algorithms based on the SSMIS sensor, as well as the BT,enhanced NASA-Team(NT2) and ARTIST Sea Ice(ASI) algorithms based on AMSR-E/AMSR-2 sensors. The daily arithmetic average of PM SIC values and the daily weighted average of OBS SIC values were used for the comparisons. The correlation coefficients(CC), biases and root mean square deviations(RMSD) between PM SIC and OBS SIC were compared in terms of the overall trend, and under mild/normal/severe ice conditions. Using the OBS data, the influences of floe size and ice thickness on the SIC retrieval of different PM products were evaluated by calculating the daily weighted average of floe size code and ice thickness. Our results show that CC values range from 0.89(AMSR-E/AMSR-2 NT2) to 0.95(SSMIS NT), biases range from-3.96%(SSMIS NT) to 12.05%(AMSR-E/AMSR-2 NT2), and RMSD values range from 10.81%(SSMIS NT) to 20.15%(AMSR-E/AMSR-2 NT2). Floe size has a significant influence on the SIC retrievals of the PM products, and most of the PM products tend to underestimate SIC under smaller floe size conditions and overestimate SIC under larger floe size conditions. Ice thickness thicker than 30 cm does not have a significant influence on the SIC retrieval of PM products. Overall, the best(worst) agreement occurs between OBS SIC and SSMIS NT(AMSR-E/AMSR-2 NT2) SIC in the Arctic summer.     

Sea-ice flux in the East Greenland Current

The sea-ice export out of the central Arctic through the Fram Strait is a key variable in the Arctic climate system. Satellite data provide the only basis for mapping ice features with a high spatial and temporal resolution in polar regions. An automatic drift algorithm has been employed and optimized to monitor the sea-ice drift velocity in the Greenland Sea with AVHRR data. The combination of the ice drift and the spatial ice distribution provides an insight into the ice transport processes along the coast of Greenland. The combination with sea-ice thickness measurements allows an estimation of the spatial distribution of the sea-ice mass flux. The seasonal and spatial variability of the mass flux allows further predictions of the meridional melting and freezing processes along the East Greenland Current. This investigation covers the years 1993 and 1994. Seasonal and spatial distributions of the sea-ice drift were derived. The derived absolute values in this study are in good agreement with estimates proposed by other authors.     

渤海AVHRR多通道海冰密集度反演算法试验研究

为了得到更精确的渤海海冰密集度反演参数,采用辽东湾不同类型海冰ASD实测数据,在分析光谱特征的基础上,针对NOAA/AVHRR数据确定出合适海冰密集度反演算法阈值。继而,基于线性光谱混合模型的多通道反演算法进行了一系列算法试验。同时实现了基于LandSat5-TM数据的渤海海冰密集度场反演,并利用该结果与AVHRR单通道和多通道算法得到的海冰密集度反演结果进行比对分析。定量误差分析结果表明,当单通道算法或组合算法中包含1通道时,与Landsat5-TM反演结果的平均误差为正值,包含2通道且不包含1通道时,平均误差为负值;同时使用这两个通道较只包含其一的各种组合算法的平均误差明显偏小;在各种组合算法中,1245四个通道组合反演的海冰密集度结果误差最小,可应用于渤海AVHRR数据海冰密集度反演。     

HY-2卫星扫描微波辐射计数据反演北极海冰漂移速度

本文基于最大互相关法,利用海洋二号（HY-2）卫星扫描微波辐射计37 GHz通道多时相垂直极化亮温数据,获取了北极海冰漂移速度。采用2012年和2013年国际北极浮标计划海冰现场观测数据,对利用微波辐射计亮温资料反演的冬季北极海冰漂移速度进行了定量验证,结果表明:流速和流向均方根误差分别为1.12 cm/s和16.37°,从一定程度上说明了HY-2卫星扫描微波辐射计亮温数据反演海冰漂移速度的可行性。此外,使用美国国防气象卫星F-17搭载的专用微波成像仪91 GHz通道垂直极化亮温,采用高斯拉普拉斯滤波方法进行处理,结合最大互相关法反演的海冰漂移速度,优于法国海洋开发研究院海冰漂移速度产品。     

Cross-calibration of brightness temperature obtained by FY-3B/MWRI using Aqua/AMSR-E data for snow depth retrieval in the Arctic

This study cross-calibrated the brightness temperatures observed in the Arctic by using the FY-3B/MWRI L1 and the Aqua/AMSR-E L2A. The monthly parameters of the cross-calibration were determined and evaluated using robust linear regression. The snow depth in case of seasonal ice was calculated by using parameters of the crosscalibration of data from the MWRI T_b. The correlation coefficients of the H/V polarization among all channels T_b of the two sensors were higher than 0.97. The parameters of the monthly cross-calibration were useful for the snow depth retrieval using the MWRI. Data from the MWRI T_b were cross-calibrated to the AMSR-E baseline.Biases in the data of the two sensors were optimized to approximately 0 K through the cross-calibration, the standard deviations decreased significantly in the range of 1.32 K to 2.57 K, and the correlation coefficients were as high as 99%. An analysis of the statistical distributions of the histograms before and after cross-calibration indicated that the FY-3B/MWRI T_b data had been well calibrated. Furthermore, the results of the cross-calibration were evaluated by data on the daily average T_b at 18.7 GHz, 23.8 GHz, and 36.5 GHz(V polarization), and at 89 GHz(H/V polarization), and were applied to the snow depths retrieval in the Arctic. The parameters of monthly cross-calibration were found to be effective in terms of correcting the daily average T_b. The results of the snow depths were compared with those of the calibrated MWRI and AMSR-E products. Biases of 0.18 cm to 0.38 cm were observed in the monthly snow depths, with the standard deviations ranging from 4.19 cm to 4.80 cm.     

中国第四次北极科学考察期间北极点附近海冰和融池的航空遥感观测

An aerial photography has been used to provide validation data on sea ice near the North Pole where most polar orbiting satellites cannot cover. This kind of data can also be used as a supplement for missing data and for reducing the uncertainty of data interpolation. The aerial photos are analyzed near the North Pole collected during the Chinese national arctic research expedition in the summer of 2010(CHINARE2010). The result shows that the average fraction of open water increases from the ice camp at approximately 87°N to the North Pole, resulting in the decrease in the sea ice. The average sea ice concentration is only 62.0% for the two flights(16 and 19 August 2010). The average albedo(0.42) estimated from the area ratios among snow-covered ice,melt pond and water is slightly lower than the 0.49 of HOTRAX 2005. The data on 19 August 2010 shows that the albedo decreases from the ice camp at approximately 87°N to the North Pole, primarily due to the decrease in the fraction of snow-covered ice and the increase in fractions of melt-pond and open-water. The ice concentration from the aerial photos and AMSR-E(The Advanced Microwave Scanning Radiometer-Earth Observing System) images at 87.0°–87.5°N exhibits similar spatial patterns, although the AMSR-E concentration is approximately 18.0%(on average) higher than aerial photos. This can be attributed to the 6.25 km resolution of AMSR-E, which cannot separate melt ponds/submerged ice from ice and cannot detect the small leads between floes. Thus, the aerial photos would play an important role in providing high-resolution independent estimates of the ice concentration and the fraction of melt pond cover to validate and/or supplement space-borne remote sensing products near the North Pole.     

Analysis of Sea-Ice Areas Undetectable by the ASI Algorithm Based on Satellite Microwave Radiometry in the Arctic Ocean

Izvestiya, Atmospheric and Oceanic Physics - In the period of intense ice melting, algorithms retrieving sea-ice concentration from satellite microwave radiometry (SMR) data may fail to detect vast...     

基于高效统计模型预测9月北极航道沿线海冰范围

The rapid decrease in Arctic sea ice cover and thickness not only has a linkage with extreme weather in the midlatitudes but also brings more opportunities for Arctic shipping routes and polar resource exploration, both of which motivate us to further understand causes of sea-ice variations and to obtain more accurate estimates of seaice cover in the future. Here, a novel data-driven method, the causal effect networks algorithm, is applied to identify the direct precursors of September sea-ice extent covering the Northern Sea Route and Transpolar Sea Route at different lead times so that statistical models can be constructed for sea-ice prediction. The whole study area was also divided into two parts: the northern region covered by multiyear ice and the southern region covered by seasonal ice. The forecast models of September sea-ice extent in the whole study area(TSIE) and southern region(SSIE) at lead times of 1–4 months can explain over 65% and 79% of the variances, respectively,but the forecast skill of sea-ice extent in the northern region(NSIE) is limited at a lead time of 1 month. At lead times of 1–4 months, local sea-ice concentration and sea-ice thickness have a larger influence on September TSIE and SSIE than other teleconnection factors. When the lead time is more than 4 months, the surface meridional wind anomaly from northern Europe in the preceding autumn or early winter is dominant for September TSIE variations but is comparable to thermodynamic factors for NSIE and SSIE. We suggest that this study provides a complementary approach for predicting regional sea ice and is helpful in evaluating and improving climate models.     

Improved sea-ice radiative processes in a global coupled climate model

1Introduction Seaiceplaysanimportantroleinmoderating heatandmoistureexchangesbetweentheatmosphere andtheoceanathighlatitudes.Seaicealsointeracts withthebroaderclimatesystembythepositiveice albedofeedback(Curryetal.,1995),whichamplifies projectedclimatewarmingatthehighlatitudes,andby theoceanicfeedbackinvolvingicegrowthandmelt, whichinfluencesglobalthermohalinecirculation(i.e., theNorthAtlanticDeepWaterandtheAntarcticBot- tomWater)(Walsh,1983;Barryetal.,1993). Recently,theimplementationofas…     

Sea surface temperature retrieval based on simulated microwave polarimetric measurements of a one-dimensional synthetic aperture microwave radiometer

Compared with traditional real aperture microwave radiometers, one-dimensional synthetic aperture microwave radiometers have higher spatial resolution. In this paper, we proposed to retrieve sea surface temperature using a one-dimensional synthetic aperture microwave radiometer that operates at frequencies of 6.9 GHz, 10.65 GHz,18.7 GHz and 23.8 GHz at multiple incidence angles. We used the ERA5 reanalysis data provided by the European Centre for Medium-Range Weather Forecasts and a radiation transmission forward model to calculate the model brightness temperature. The brightness temperature measured by the spaceborne one-dimensional synthetic aperture microwave radiometer was simulated by adding Gaussian noise to the model brightness temperature.Then, a backpropagation(BP) neural network algorithm, a random forest(RF) algorithm and two multiple linear regression algorithms(RE1 and RE2) were developed to retrieve sea surface temperature from the measured brightness temperature within the incidence angle range of 0°–65°. The results show that the retrieval errors of the four algorithms increase with the increasing Gaussian noise. The BP achieves the lowest retrieval errors at all incidence angles. The retrieval error of the RE1 and RE2 decrease first and then increase with the incidence angle and the retrieval error of the RF is contrary to that of RE1 and RE2.     

Polarization effects in seaice signatures

Observations of microwave emissivities of multiyear sea ice showed anomalies at horizontal polarization in the frequency range from 5 to 35 GHz during the Norwegian Remote Sensing Experiment (NORSEX) [1] in September and October 1979. The effect can be explained by layers of solid ice present in the dry snow cover throughout the NORSEX area. A special experiment made on a typical multiyear floe confirms this explanation. Since the results also indicate that at 94 GHz the layers do not affect the radiation, a dual-polarized radiometer in the 90-GHz window is a promising sea-ice sensor.     

High-resolution sea ice in long-term global ocean GCM integrations

The resolution of the sea-ice component of a coarse-resolution global ocean general circulation model (GCM) has been enhanced to about 22 km in the Southern Ocean. The ocean GCM is designed for long-term integrations suitable for investigations of the deep-ocean equilibrium response to changes in southern hemisphere high-latitude processes. The space and time scales of the high-resolution sea-ice component are commensurate with those of the resolution of satellite passive-microwave sea-ice data. This provides the opportunity for a rigorous evaluation of simulated sea-ice characteristics. It is found that the satellite-derived continuous high ice concentration of the interior winter ice pack can only be captured when vertical oceanic mixing is modified in a way that less local, intermittent convection occurs. Furthermore, the width and the variability of the coastal polynyas around the Antarctic continent and its ice shelves are best captured when some form of ice-shelf melting is accounted for. The width of the wintertime ice edge is reasonably reproduced, while its variability remains underestimated, closely following the coarse-grid pattern of the ocean model due to its high dependence on ocean temperature. Additional variability besides daily winds, e.g. in form of idealized tidal currents, improves the temporal and spatial ice-edge variability, while leads in the interior ice pack become more abundant, more in line with the fine-scale satellite-derived texture. The coast- or ice-shelf line is described on the fine grid based on satellite passive-microwave data. This method requires parts of a coarse coastal ocean grid cell to be covered by an inert layer of “fast ice” or “ice shelf”. Reasonable long-term global deep-ocean properties can only be achieved when these areas are not inert, i.e. are exposed to heat flux and ice growth, or when the vertical mixing parameterization allows for excessive open-ocean convection. The model area exposed to cold high-latitude atmospheric conditions thus being most decisive for a realistic representation of the long-term deep-ocean properties, suggests that high-latitude coastlines are definitely in need of being represented at high resolution, including ice sheets and their effects on the heat and freshwater flux for the ocean.     

Integrative algorithm of determining ice conditions in Polar Regions by data of satellite microwave radiometry (VASIA2)

In this paper, a new algorithm for determining the concentration of the ice cover in Polar Regions by data of satellite microwave radiometry is considered. The technique of its construction is described in detail; it cardinally differs from the technique of creating present-day algorithms. The new algorithm demonstrates good results in determining the concentration of the ice cover in Polar Regions. The algorithm permits one not only to obtain maps of ice concentration, but also to determine areas of puddles covering the ice-cover surface in summer months. The algorithm is easy-to-use and requires no additional or fitting parameters. At the end of the work, advantages and disadvantages of the new algorithm are discussed.