| Sea ice is not only an important component of the cryosphere but also an important regulator of the polar climate system and a sensitive factor of global climate change.Many studies have shown that sea ice has important effects on the atmosphere,ecology,environment,ocean,and human activities.Therefore,monitoring polar sea ice parameters and obtaining information on their temporal and spatial distribution have important scientific value and practical significance.Sea ice concentration and sea ice thickness are two important sea ice parameters that characterize the changes of sea ice in the"horizontal direction"and"vertical direction",respectively.Compared with traditional sea ice concentration and sea ice thickness measurement methods,remote sensing covers a wide range and can achieve a large-scale continuous observation.Therefore,research on sea ice concentration and sea ice thickness remote sensing inversion methods is of great significance for obtaining and understanding the temporal and spatial distribution and changing trends of sea ice in polar regions.This dissertation takes polar regions as the research area.Aiming at the shortcomings of the current passive microwave remote sensing sea ice concentration inversion method and the newly launched ICESat-2 laser altimeter,the research work on the sea ice concentration and sea ice thickness remote sensing inversion method was carried out respectively.Moreover,the spatio-temporal changes of sea ice parameters in the polar region with long time series are analyzed using the sea ice concentration data produced by this dissertation.The main research content and conclusions of the dissertation include the following four aspects:(1)A comprehensive accuracy evaluation and consistent evaluation of the international popular passive microwave sea ice concentration products was carried out.First,high-resolution remote sensing images(MODIS)were used to evaluate the accuracy of four passive microwave sea ice concentration products(SSMIS/ASI,AMSR2/BT,FY3B/NT2,and FY3B/NT2)produced by different sea ice concentration algorithms(NT2,BT,and ASI algorithm).Secondly,ERA-Interim Reanalysis data,which has a temporal and spatial resolution similar to that of passive microwave sea ice concentration,was used to compare the consistency between ERA-Interim and four different passive microwave sea ice concentration products in polar regions.Finally,the performance of various sea ice concentration products(algorithms)in the polar regions and the temporal and spatial distribution of errors were summarized and corresponding solutions were proposed,laying the foundation for the subsequent newly developed sea ice concentration fusion algorithm.(2)According to the error characteristics of different passive microwave sea ice concentration algorithms,a new passive microwave sea ice concentration fusion method was proposed.Through the assessment of the consistency of the current international popular passive microwave sea ice concentration products in the Arctic and Antarctic regions,it is found that the existing sea ice concentration products have large errors in marginal ice zones and during the summer period.Among four passive microwave sea ice concentration products,ASI and BT algorithms have the best performance in the Arctic and Antarctic regions,respectively.Therefore,based on the opposite characteristics of underestimation and overestimation of the two algorithms in marginal ice zones and the summer period,the ASI and BT algorithms were chosen as the benchmark algorithm to effectively integrate a new hybrid method to eliminate the influence.Furthermore,ship-based observations data of the Chinese National Arctic Research Expedition from 2012 to 2014 in Arctic and Antarctic were used to evaluate the accuracy of the new hybrid method proposed by the thesis.The results show that the accuracy of the sea ice concentration retrieved by the new algorithm is better than the ASI and BT algorithms,and the root mean square error(RMSE)is 9.55%in Arctic and 11.37%in Antarctic.Besides,the sea ice concentration retrieved by the new algorithm can well capture the dynamic change trend of sea ice.The coefficient of determination R~2 in Arctic and Antarctic is above 0.8,indicating the proposed hybrid method can achieve high precision and long-term production of sea ice concentration products in polar regions.(3)A systematic analysis of the temporal and spatial changes of sea ice concentration,sea ice area,and extent in the Arctic and Antarctic regions in the past 20years was carried out.Based on the new sea ice concentration fusion algorithm proposed in this thesis,combined with AMSR-E and AMSR2 multi-channel radiometer brightness temperature data,a long-time series daily-scale sea ice concentration product from 2002 to 2020 was generated.The temporal and spatial distribution and change trend of sea ice concentration,sea ice area,and extent in polar regions and their sub-regions were analyzed from different time scales(annual,seasonal,monthly).The results show that the annual average sea ice concentration in the Arctic and Antarctic regions as a whole shows a decreasing trend in last 20 years.Among them,the decreasing trend of sea ice in Arctic is more obvious than that in Antarctic,especially in the Barents Sea and the Kara Sea.The results further reveal the characteristics of sea ice changes in the polar regions since the 21st century and help to further understand the changing characteristics of sea ice and the similarities and differences between the Antarctic and Arctic sea ice changes in the context of global warming.(4)The influence of different parameter combination schemes on the inversion results of sea ice thickness in the Arctic region based on the ICESat-2 laser altimeter was discussed.The existing altimeter-based methods for obtaining sea ice thickness are mainly based on the assumption of hydrostatic equilibrium.Different snow depth,snow density,and sea ice density input data have relatively greater uncertainties for the sea ice thickness inversion results.The dissertation first analyzed the spatial distribution of sea ice freeboard in the Arctic region and the temporal changes of the monthly average sea ice freeboard obtained by the ICESat-2 laser altimeter in the two winters of 2018-2019 and 2019-2020.Then,a total of 18 parameter combinations with different snow depth,snow density,and sea ice density schemes were carried out to compare the sea ice thickness inversion results based on the ICESat-2 laser altimeter.The difference between the inversion results of the parameterized schemes for one-year and multi-year ice regions was also analyzed.Finally,the inversion results of the sea ice thickness using different schemes based on the ICESat-2 were compared with the ICEBird airborne data set to determine the optimal scheme,which was adopted to map the spatial distribution of the sea ice thickness of the ICESat-2 in the Arctic region in the winter of 2020.Three Cryo Sat-2 sea ice thickness products(AWI,CMOP,and GSFC)were used to analyze the difference with ICESat-2.The results showed that the average deviations of the sea ice thickness retrieved by ICESat-2 and the sea ice thickness of AWI,CMOP,and GSFC were-0.15 m,-0.26 m,and-0.13 m,respectively. |
PDF Full Text Request