Figure 3 . a . The 1963–1986/1988 flow velocities. The background image is the georeferenced DISP image showing the extended ice front in the south in 1963. The dashed lines divide the velocity data into five regions (Areas 1–5). b . Box and Whiskers plots of the time-series flow velocities for the five areas, generated based on the velocity measurements of the locations where the 1963–1986/1988 velocity data are available for each period. The “x” markers indicate the mean value.
4.4 Flow accelerations associated with rifting and calving
The most significant velocity change was strongly related to the rift development within the downstream portion of the C5–C6 units (Figure 4). There was a sharp velocity change along the rift line during 2015–2016 (Figure 4e) immediately before the calving occurred. Along the C6 flow line (Figure 4h), the average velocity of the portion that calved in 2017 increased by more than 50% from 570±16 m/year during 1986–1988 to 890±36 m/year during 2015–2016. The largest increase over this period was between 2013–2015 and 2015–2016, when the velocity increased by 21.6±2.5% from a speed value of 731±25 m/year. In comparison, the central ice shelf (C4; Figure 4g) exhibited much less variability in velocity before calving. The calved portion of C4 had an average velocity of 723±21 m/year during 2015–2016, increasing by only 3.1±0.7% relative to the previous period 2013–2015. However, there was a velocity increase of 9.1±2.3% between 1986–1988 and 2000–2002 from the ice front to an upstream distance of ~80 km. This acceleration was very likely due to the velocity adjustment in response to the front geometry change caused by the 1986 retreat. Despite the sparse velocity data before 2000, this observed acceleration pattern may indicate a lagged response of flow velocity to front retreat. The lagged response suggests that although the 2017 calving event has not yet induced any significant flow accelerations, there is potential for a future flow velocity response to the 2017 event.