Spatiotemporal analysis of sea ice in the Weddell Sea of Antarctic based on GTWR

The Geographical and Temporal Weighted Regression (GTWR) method is employed to assess the impact of various environmental factors on sea ice concentration (SIC) in the Weddell Sea. Initially, MODIS-derived SIC was used to evaluate the accuracy of six SIC products derived from different satellite sensors and algorithms. The MWRI/NT2 product demonstrated the highest correlation with the MODIS data, validating its reliability for further analysis. Using the MWRI/NT2 product, along with ERA5 and NCEP/NCAR reanalysis datasets, we investigated the interannual and seasonal trends in SIC and sea ice extent (SIE) from 2011 to 2023. The results indicate a declining trend in SIE at a rate of -6.2 ± 1.9×10³ km²/yr, with the most significant ice loss occurring in autumn. The GTWR analysis highlights significant spatial and temporal variability in the factors influencing SIC in the Weddell Sea. The Latent heat flux (LH) emerged as the most influential factor, with a median standardized regression coefficient of 1.44. The LH primarily promotes sea ice growth by cooling the surface through the condensation of atmospheric water vapor. Zonal winds also played a critical role, particularly by promoting sea ice formation through Ekman transport of cold surface water. However, wind speed had a minimal impact on SIC, likely due to the lack of directional data in the ERA5 dataset. In contrast, the impact of net radiation (NR) varied significantly across the region, complicating its overall influence on sea ice dynamics. Sensible heat flux (SH) generally supported ice growth, except in the central Weddell Sea, where local conditions caused SH to inhibit ice formation. These findings underscore the complex interplay of environmental factors in shaping SIC.