D thick ice. Despite the fact that these observations of 1 day per year
D thick ice. Though these observations of one particular day per year for seven years can not represent the all round continuous spatiotemporal variations of lead fraction, this common spatial pattern agrees with that of prior lead studies [5,18,19,39]. Figure 5b portrays the averaged area of individual leads for the 25 km track segment, and Figure 5c portrays the ratio of the number of lead-included images towards the total number of images for the 25 km segment. The lead fraction (Figure 5a) was determined by the individual lead area (Figure 5b) and also the frequency of leads (Figure 5c). For example, despite the fact that big leads have been observed in 2013 for 000 km (Figure 5b), lead frequency for this portion was low (Figure 5c) as a result of the small quantity of huge leads. As a result, the averaged lead fraction for this segment was not high due to the fact from the low lead frequency. Additionally, the lead frequency in 2018 for 1000500 km was relatively high, however the averaged lead fraction was not so ��-Carotene Autophagy higher due to the massive variety of modest leads.Remote Sens. 2021, 13,for the total number of pictures for the 25 km segment. The lead fraction (Figure 5a) was determined by the individual lead area (Figure 5b) and the frequency of leads (Figure 5c). As an example, though significant leads had been observed in 2013 for 000 km (Figure 5b), lead frequency for this portion was low (Figure 5c) as a consequence of the smaller number of huge leads. Consequently, the averaged lead fraction for this segment was not higher for the reason that of the low lead frequency. In addi11 of 18 tion, the lead frequency in 2018 for 1000500 km was comparatively high, however the averaged lead fraction was not so higher on account of the substantial number of little leads.Figure 5. (a) Averaged lead fraction for every single 25 km; (b) averaged region of person leads for each and every 25 km; (c) frequency Figure five. (a) Averaged lead fraction for each 25 km; (b) averaged area of individual leads for just about every 25 km; (c) frequency of lead-included pictures for every 25 km. Gray parts indicate missing/invalid information. of lead-included pictures for each and every 25 km. Gray components indicate missing/invalid data.4.two.two. Retrieval of Freeboard 4.two.two. Retrieval of Freeboard Depending on the DMS lead detection outcome, we calculated the 400 m imply sea ice freeboard Depending on the DMS lead detection outcome, we calculated the 400 m mean sea ice freeboard fromthe ATM surface height data (Figure six). The MYI area (near centralcentralOcean) at track in the ATM surface height data (Figure six). The MYI area (near Arctic Arctic Ocean) at track 1200 km showed larger a greater (i.e., thicker ice) in comparison with that of to FYI distance distance 1200 kmashowedfreeboard freeboard (i.e., thicker ice) comparedthe that in the FYI location (close to the Beaufort Sea with a track distance beyond 1200 km). As shown in Table 7, the FYI region always showed a lower freeboard than the MYI area. Additionally, the freeboard retrieved from our lead detection shows an excellent correlation with all the ATM freeboard Phenmedipham supplier product provided by NSIDC [32]–correlation coefficient (R) was 0.832, and root mean square distinction (RMSD) was 0.105 m (Table 8). It’s also noted that 2015, 2016, and 2017 showed fairly reduced R and higher root mean square error (RMSE) than the other years (Table 8 and Figure 7), which might be resulting from the decrease classification accuracy of those years (Table six). Some misclassified leads could make substantial variations in estimation of sea surface height, at some point major to the variations among our freeboard estimation plus the NSIDC freeboard solution.