Greenland's ice sheet mass loss significantly contributes to sea-level rise, driven by an imbalance between snowfall, melt, and calving. The Northeast Greenland Ice Stream (NEGIS) is a major contributor, draining approximately 12% of the ice sheet. Unlike other Greenlandic ice streams, NEGIS's position and width aren't topographically controlled; instead, factors like hydrology, substrate, thermal feedbacks, or crystal orientation fabric likely play crucial roles. Recent large dynamical changes have occurred at all three major NEGIS outlets (Zachariæ, 79°N, and Storstrømmen glaciers), including speedups and frontal retreat. Ice-flow modeling suggests that these outlet changes, driven by ocean warming, will continue for at least a century. However, the rate at which these changes propagate inland is uncertain and crucial for accurate sea-level rise estimations. This study investigates how far these changes have propagated inland and examines velocity changes within NEGIS's interior, decoupled from coastal ice-flow dynamics, to help disentangle the causes of observed ice dynamic changes.

Previous research has highlighted the significant mass loss of the Greenland Ice Sheet (Fox-Kemper et al., 2021) and the substantial dynamical changes at NEGIS outlets, such as Zachariæ and Nioghalvfjerdsfjorden glaciers (Mouginot et al., 2015; Lu An et al., 2021). Studies have explored the role of thermal weakening (Holschuh et al., 2019), crystal orientation fabric (Minchew et al., 2018), and geothermal heat flux in controlling shear margin positions. The surge behavior of Storstrømmen glacier has also been investigated (Reeh et al., 1994; Mouginot et al., 2018). Modeling efforts have attempted to predict the response of NEGIS outlets to ocean forcing (Choi et al., 2017) and the impact of crystal orientation fabrics on basal friction and mass flux (Rathmann & Lilien, 2022). The response of the subglacial drainage system to surface elevation changes has also been studied (Karlsson & Dahl-Jensen, 2015). Finally, long-term changes in accumulation have been addressed (Box et al., 2013; Karlsson et al., 2020), and bed topography and subglacial landforms in the onset region of NEGIS have been mapped (Franke et al., 2020). These previous findings provide a framework for understanding the complexities of ice stream dynamics in Northeast Greenland.

The study utilized two primary datasets: (1) ITS_LIVE annual ice-velocity maps (Gardner et al., 2019; Hvidberg et al., 2020) derived from optical feature tracking of Landsat scenes covering 1985-2018, and (2) GPS observations from a 2015-2019 campaign at EastGRIP. Along-flow acceleration was calculated by fitting linear models to x and y velocities separately and then computing the along-flow component. GPS data from 63 stakes were used, correcting for advective acceleration. To analyze the spatial pattern of acceleration, an average profile across the southeastern shear margin was created. Furthermore, four experiments were conducted using an idealized 2D cross-section model of an ice stream to simulate the surface velocity response to changes in boundary conditions and ice-flow parameters. These simulations investigated the effects of ice stream thickness, sliding at the base, and softening of the ice column beneath the shear margin. The kinematic wave theory was applied to assess the propagation of signals from the outlet glaciers to the interior of the ice stream (Williams et al., 2012).

Analysis of the ice velocity data revealed that the shear margins of the Northeast Greenland Ice Stream (NEGIS) are accelerating, indicating a widening of the ice stream. This acceleration signal is consistent over several hundred kilometers from the onset of the ice stream to approximately 77°N. The acceleration is most prominent just inside the shear margin. GPS observations at EastGRIP corroborated the accelerating ice flow, although uncertainties in both the remotely sensed and GPS data limited a direct comparison. The study observed a distinct region with no clear pattern of acceleration separating inland changes from frontal changes, suggesting the interior acceleration isn't directly linked to recent dynamical changes at the outlets. Modeling showed that the spatial fingerprint of acceleration is most consistent with a softening of the ice column beneath the shear margin, a process potentially driven by warming or evolving ice fabric. In contrast, simulations of changes in ice stream thickness or basal sliding produced different spatial patterns of acceleration.

The findings challenge the assumption that ice stream interior acceleration is solely a response to external forcing at the outlets. The spatial separation of the outlet dynamics and the interior acceleration suggests a different mechanism is at play. The model results strongly support the hypothesis that the ice stream widening is due to an internal dynamical instability, potentially involving shear margin softening through either warming or changes in ice fabric. The observed widening of NEGIS, deep within the ice sheet interior, is indicative of greater dynamical variability than previously recognized. This raises questions about the susceptibility of NEGIS to even slight changes in forcing, and whether the observed widening is a precursor to a large-scale reorganization of ice flow in Northeast Greenland, similar to past events in other ice streams. The study underscores the importance of carefully separating internal dynamics from external forcings when projecting future ice sheet mass loss.

This study demonstrates that the Northeast Greenland Ice Stream (NEGIS) is experiencing significant acceleration deep within its interior, leading to ice stream widening. This acceleration is not simply a response to recent changes at the ice stream outlets, but rather is likely due to internal processes such as shear margin softening. This finding highlights the complex dynamics of ice streams and underscores the need for further research to fully understand the interplay between internal processes and external forcing in driving ice sheet mass loss. Future work should focus on detailed investigations of ice fabric evolution, temperature changes within the ice stream, and the potential for large-scale reorganization of ice flow in Northeast Greenland.

The uncertainties associated with both the remotely sensed and GPS velocity data limited the precision of the acceleration estimations. The idealized 2D model used in the study may not fully capture the complexity of three-dimensional ice flow. The study focused primarily on the southeastern shear margin, and further investigation of the northwestern shear margin would enhance understanding of the ice stream widening phenomenon. Additionally, while the study considered various factors, other un-modeled processes may contribute to the observed acceleration.