Arctic Sea Ice Drift Fields Extraction Based on Feature Tracking to MODIS Imagery
Yan Fang,Xue Wang,Gang Li,Zhuoqi Chen,Fengming Hui,Xiao Cheng
DOI: https://doi.org/10.1016/j.jag.2023.103353
IF: 7.5
2023-01-01
International Journal of Applied Earth Observation and Geoinformation
Abstract:Moderate-resolution optical imagery holds great potential in deriving Arctic sea ice drift fields because of its higher resolution than microwave radiometers and scatterometers, as well as its larger swath widths than most other optical and synthetic aperture radar (SAR) images. However, the application of such imagery is hindered by cloud influences and a lack of texture. In this study, we propose a method of deriving Arctic sea ice drift fields based on applying feature tracking to Moderate Resolution Imaging Spectroradiometer (MODIS) imagery. To enhance the quality of the feature tracking step, a bundle of digital image processing techniques is first introduced, including histogram equalization which is based on the Cumulative Distribution Function (CDF) of the sea ice area, and Laplacian filtering which enhances image texture. Various MODIS bands and A-KAZE parameter settings are subsequently compared to balance the quality of sea ice drifting fields and calculation efficiency. Three pairs of MODIS images observed in different zones of the Arctic Ocean are selected to evaluate the performance of the proposed method. International Arctic Buoy Programme (IABP) buoy data are employed for validating the derived drift vectors with MODIS imagery. The results show that our proposed method effectively increases the number of vectors and their coverage rates of the sea ice drift fields extracted with MODIS images. The coverage rates of sea ice drift fields in three regions increase from 4.8%, 2.3%, and 2.5% to 56.5%, 23.5%, and 53.0% compared to using the A-KAZE algorithm directly, respectively. The MAEs of the derived sea ice motion vectors are 707 m/d in speed and 6.4° in direction, superior to the sea ice drift products based on the Advanced Very High Resolution Radiometer (AVHRR) imagery. The proposed method enables MODIS and other medium-resolution optical data to be utilized in deriving Arctic sea ice drift fields, which is of great significance to the long-term and large-scale Arctic environment, climate, and oceanography research in the future.