摘要
南极海冰表面的积雪显著影响着海冰的生长与消融,在全球气候变化中起着至关重要的作用。应用我国自主海洋卫星HY-2B微波辐射计(SMR)数据获取高精度南极海冰表面积雪深度研究较少,本研究利用我国第35次南极科学考察“雪龙”号走航观测的积雪深度数据,构建HY-2B SMR的被动微波积雪深度遥感反演模型,并与传统的反演模型进行比较和验证。结果表明,新构模型反演的南极海冰表面积雪深度与现场观测数据的平均偏差仅为-1.70 cm,反演精度优于传统常用的Markus98模型和Comiso03模型。与美国国家冰雪数据中心发布的GCOM-W1 AMSR-2南极海冰表面积雪深度产品对比发现,应用HY-2B SMR数据和新构模型反演的时序积雪深度大于AMSR-2积雪深度产品,整体变化趋势一致,但HY-2B SMR反演雪深更接近于实际观测值。两个产品的差异在积雪积累期和稳定期(4月至10月)主要集中在东南极海域和罗斯海,在消融期(11月至次年3月)差异主要分布在西南极海域以及威德尔海南部。开展国产HY-2B SMR数据的南极海冰表面积雪深度反演研究,可为我国业务化南极冰雪动态监测及南极海冰与气候变化研究提供大范围、高精度的数据支撑和技术参考。
Snow cover on Antarctic sea ice significantly influences the processes of sea ice growth and melting,regulates energy exchange between the atmosphere and ocean,and plays a critical role in global climate change.This study explores the application of China’s HY-2B satellite microwave radiometer(SMR)data for retrieving high-precision snow depth information over Antarctic sea ice.Using HY-2B passive microwave brightness temperature(BT)data at 6.925 GHz,10.7 GHz,18.7 GHz,23.8 GHz,and 37.0 GHz,we analyzed the brightness temperature gradient ratios(GRs)derived from pairwise combinations of the BTs.Among these,only the 23.8 GHz channel provides data in vertical polarization(V),while the other channels offer both vertical(V)and horizontal(H)polarization modes.A total of 10 GRs were calculated and correlated with ship-based snow depth observations to identify the GR with the strongest relationship to snow depth.The optimal GR,derived from vertically polarized BTs at 18.7 GHz and 37.0 GHz,was used alongside snow depth data collected during China’s 35th Antarctic scientific expedition aboard the research vessel Xuelong and Antarctic Sea Ice Processes and Climate(ASPeCt)ship-based observations to develop a snow depth retrieval model for HY-2B SMR data.The model construction excluded the influence of open water and liquid water on the BTs of sea ice surfaces.To evaluate the performance of the newly developed model,four statistical metrics—Bias,mean absolute error(MAE),root mean square error(RMSE),and correlation coefficient(r)—were used to compare the retrieved snow depths with those from traditional models(Markus98 and Comiso03)and the GCOM-W1 AMSR-2 Antarctic sea ice snow depth product released by the National Snow and Ice Data Center(NSIDC).The results show that the snow depths retrieved using the new model exhibit spatial distributions similar to those derived from the Markus98 and Comiso03 models,all displaying clear seasonal variations.Compared to the traditional Markus98 and Comiso03 models,the new model demonstrated higher accuracy,with a bias of only-1.70 cm,and outperformed the existing GCOM-W1 AMSR-2 Antarctic sea ice snow depth product.The snow depths retrieved from HY-2B SMR data for 2019 and those from the GCOM-W1 AMSR-2 product both ranged between 10 and 30 cm,with similar daily and monthly mean snow depth trends.The overall bias between the two products ranged from-5 to 8.4 cm,with larger differences observed in December,January,and February.During the snow accumulation and stabilization periods( April to October), the average bias between the two datasets was only 1. 71 cm,with a high correlation coefficient of 0. 8, indicating strong consistency between the two snow depth datasets during this period. This consistency further validates the reliability of the newly developed model. The main differences during this period were concentrated in the southeastern Antarctic Ocean and the Ross Sea. During themelting period (November to March), the average bias increased to 3. 12 cm, with the largest differences observed in the southwestern Antarctic Ocean and the southern Weddell Sea. Although the biases between HY-2BSMR-derived snow depths and the GCOM-W1 AMSR-2 product were larger during the melting period, the overall annual mean bias between the two datasets was 1. 42 cm. Furthermore, the snow depths retrieved from HY-2B SMR data were generally higher than those from the GCOM-W1 AMSR-2 product and closer to actual fieldobservations. This study demonstrates the potential of using HY-2B SMR data for retrieving snow depths overAntarctic sea ice and provides high-accuracy, large-scale datasets to support operational monitoring of Antarcticice and snow dynamics. It also offers valuable technical references for advancing research on Antarctic sea iceprocesses and their interactions with global climate change.
作者
刘娜娜
季青
于梦琴
王伟
LIU Nana;JI Qing;YU Mengqin;WANG Wei(School of Geography and Tourism,Anhui Normal University,Wuhu 241002,Anhui,China)
出处
《冰川冻土》
2025年第1期42-56,共15页
Journal of Glaciology and Geocryology
基金
国家自然科学基金项目(42076235)
安徽省中青年教师培养行动项目(YQZD2023009)
安徽省高等学校省级质量工程项目(2023jyxm0164)
安徽师范大学大学生创新创业训练计划项目(202410370002)资助。
关键词
南极海冰
积雪深度
HY-2B卫星
被动微波
卫星遥感
Antarctic sea ice
snow depth
HY-2B satellite
passive microwave
satellite remote sensing
