Retrieval And Application Of Snow Depth On Arctic Sea Ice Based On Active And Passive Microwave Data

With the rapid development of human industrial activities,climate and environment change is accelerating.The increase and decrease of snow depth on sea ice are closely related to climate change.Global warming has led to a significant reduction in polar sea ice,and to obtain accurate information on Arctic shipping routes is essential.Obtaining accurate and high-resolution snow depth on sea ice and retrieving small-scale sea ice thickness are key issues in Arctic sea ice changes and route planning research.However,the existing snow depth data is inadequate to meet the requirements of Arctic sea ice change research and ship route planning.Therefore,this paper conducts research in the following areas:(1)Research on the difference and accuracy of passive microwave remote sensing snow depth productsFirstly,this paper analyzes the issues of the snow depth observed by Ice Mass Balance buoys(IMB),and achieves automated and high-precision acquisition of IMB snow depth.Secondly,the differences in temporal and spatial coverage as well as numerical values of snow depth retrieved by various sensors from different organizations are compared and analyzed,and then the accuracy of different passive microwave remote sensing products is evaluated.Finally,the spatial distribution and seasonal variation characteristics of snow depth across the entire Arctic sea ice region are quantitatively analyzed.(2)Research on snow depth retrieval from active microwave remote sensing dataThe impact of snow humidity on snow depth retrieval and the coverage range of dry snow on the sea ice is quantitatively obtained.The spatial and temporal matching of the sea ice drift between the Synthetic Aperture Radar(SAR)images and Operation Ice Bridge(OIB)observations is conducted,and then developing the empirical model and machine learning model to retrieve snow depth by taking the OIB snow depth data as the observed values.The accuracy differences between different models’ snow depths are compared,and the spatial distribution characteristics of high-resolution snow depth are calculated.The long-term spatial distribution of snow depth is retrieved,revealing the changes in snow cover extent and depth over time.(3)Application research on snow depth retrieval from active and passive microwave remote sensing dataThe spatial and numerical differences of snow depth retrieved by active and passive microwave remote sensing are quantitatively analyzed,and their applicable conditions are clarified.Based on the satellite altimetry data,the correlation between active and passive microwave remote sensing snow depth and sea ice freeboard is investigated,and an sea ice freeboard retrieval model is developed to obtain high-resolution spatial distribution of sea ice thickness.Combining the high-resolution sea ice concentration data with sea ice thickness data,the research on the application of ship route planning in a small scale Arctic sea ice area is carried out.The main research conclusions of this paper are as follows:(1)The accuracy of different passive microwave remote sensing snow depth products varied greatly.The NSIDC AMSR-E snow depth product has the highest accuracy,with an average difference of less than 1 cm compared to OIB snow depth.The next accurate products are NASA SSM/I and UB SSMIS,with average differences of less than 10 cm compared to OIB snow depth.The UB AMSR-E snow depth product has the lowest accuracy,with a difference of up to 12 cm.However,the different passive microwave remote sensing snow depth products show consistent snow depth variation with time,which can effectively reflect the temporal variation of snow depth on the surface of Arctic sea ice.(2)The wet snow index is constructed using Sentinel-1 data,and the threshold method is used to quantitatively calculate the coverage of dry snow,eliminating the influence of wet snow on snow depth retrieval.The empirical model of SAR snow depth retrieval based on HV/HH polarization ratio is developed,with a correlation coefficient of 0.58 compared to OIB snow depth.The correlation coefficient between the snow depth retrieved by machine learning model and the OIB snow depth is 0.73,indicating its better performance.The study finds that the variation of snow depth in the Smith Strait is consistent with the growth and melting of sea ice and exhibits clear seasonal characteristics,with the maximum snow depth occurring in February and the minimum in August and September.A comparison analysis is conducted between SAR and different snow depth products in the Smith Strait,which finds that the retrieval results of SAR have a higher upper limit of snow depth.(3)The snow depth values retrieved by active microwave remote sensing are generally higher than those estimated by passive microwave.For areas with fine scales or thick snow,the snow depth retrieved by SAR is more suitable.Both active and passive microwave snow depth data are highly correlated with satellite altimetry data,and then the high-resolution sea ice freeboard and sea ice thickness products are generated based on the empirical model.A route planning model is established based on sea ice concentration,sea ice thickness,and sailing distance and applied to the navigation analysis of the coastal area of Svalbard.This paper conducts an in-depth research about the retrieval and application of snow depth on Arctic sea ice based on active and passive microwave remote sensing data,and retrieves high-resolution and accurate snow depth data on Arctic sea ice,and applies it to the retrieval of sea ice thickness and ship route planning,which provides a data basis and technical support for the small scale research of Arctic sea ice and snow parameters,as well as for the application research on ship navigation.