Chinese Antarctic Center of Surveying and Mapping, Wuhan University, Wuhan 430079, China
2.
Key Laboratory of Polar Surveying and Mapping Science, Ministry of Natural Resources, Wuhan 430079, China
Funds:
The National Natural Science Foundation of China under contract No. 42076235; the Fundamental Research Funds for the Central Universities under contract No. 2042022kf0018.
The snow depth on sea ice is an extremely critical part of the cryosphere. Monitoring and understanding changes of snow depth on Antarctic sea ice is beneficial for research on sea ice and global climate change. The Microwave Radiation Imager (MWRI) sensor aboard the Chinese FengYun-3D (FY-3D) satellite has great potential for obtaining information of the spatial and temporal distribution of snow depth on the sea ice. By comparing in-situ snow depth measurements during the 35th Chinese Antarctic Research Expedition (CHINARE-35), we took advantage of the combination of multiple gradient ratio (GR (36V, 10V) and GR (36V, 18V)) derived from the measured brightness temperature of FY-3D MWRI to estimate the snow depth. This method could simultaneously introduce the advantages of high and low GR in the snow depth retrieval model and perform well in both deep and shallow snow layers. Based on this, we constructed a novel model to retrieve the FY-3D MWRI snow depth on Antarctic sea ice. The new model validated by the ship-based observational snow depth data from CHINARE-35 and the snow depth measured by snow buoys from the Alfred Wegener Institute (AWI) suggest that the model proposed in this study performs better than traditional models, with root mean square deviations (RMSDs) of 8.59 cm and 7.71 cm, respectively. A comparison with the snow depth measured from Operation IceBridge (OIB) project indicates that FY-3D MWRI snow depth was more accurate than the released snow depth product from the U.S. National Snow and Ice Data Center and the National Tibetan Plateau Data Center. The spatial distribution of the snow depth from FY-3D MWRI agrees basically with that from ICESat-2; this demonstrates its reliability for estimating Antarctic snow depth, and thus has great potential for understanding snow depth variations on Antarctic sea ice in the context of global climate change.
Figure 1. The distributions of ship routes during CHINARE-35 (blue line), the ship-based observational snow depth (red dots), Alfred Wegener Institute (AWI) snow buoy (orange line) and Operation IceBridge (OIB) airborne measurements (green line) (a) and snow depth estimation involving photogrammetric images (b). The long red line in b denotes the reference ball diameter, the short red lines denote snow depth on sea ice.
Figure 2. Daily TB of FY-3D MWRI at different freque Timency bands in open water from 2018 to 2020.
Figure 3. The scatter diagram between ship-based observational snow depth and combined gradient radio derived from FY-3D MWRI data.
Figure 4. Comparisons of ship-based observational snow depth to the snow depth estimates based on the proposed model (a), Markus98 model (b) and Comiso03 model (c), and comparisons of AWI snow buoy data to the snow depth estimates based on the proposed model (d), Markus98 model (e) and Comiso03 model (f). The color bar represents the density of the points.
Figure 5. The spatial distributions of averaged snow depth on Antarctic sea ice in different seasons (a–d) with its uncertainty (e–h) from 2018 to 2020. Spring: September–November, Summer: December–February, Autumn: March–May and Winter: June–August.
Figure 6. Time series of snow depth with its uncertainty in Antarctic Ocean (a), Weddell Sea: 60°W–20°E (b), Indian Ocean: 20°–90°E (c), Pacific Ocean: 90°–160°E (d), Ross Sea: 160°E–130°W (e) and Bellingshausen-Amundsen Sea: 130°–60°W (f) from 2018 to 2020.
Figure 7. Comparison of FY-3D MWRI snow depth, NSIDC-AMSR2 SD and NTPDC-AMSR2 SD in the Antarctic and five seas from 2018 to 2019 for the statistic index of mean difference (a), mean absolute difference (b), standard deviation (c), and correlation coefficient (d).
Figure 8. The spatial distributions of monthly snow depth (January to June 2020) derived from FY-3D MWRI (a) and ICESat-2 (b), with the probability density functions (PDFs) (c) from FY-3D MWRI (red line) and ICESat-2 (blue line) in the common area.
Figure 9. The spatial distributions of monthly snow depth (July to December 2020) derived from FY-3D MWRI (a) and ICESat-2 (b), with the probability density functions (PDFs) (c) from FY-3D MWRI (red line) and ICESat-2 (blue line) in the common area.
Figure 10. Time series of FY-3D MWRI snow depth and ICESat-2 snow depth from January to December 2020 (a), and the difference between FY-3D MWRI snow depth and ICESat-2 snow depth from January to December 2020 (b).
DownLoad:
