Abstract Classical treatments of ice‐shelf bending suggest that shelf fronts should bend downwards, due to the distribution of hydrostatic water pressure at the front. However, LiDAR data show several instances of upward‐bending ice‐shelf fronts. While this phenomenon has often been attributed to a buoyant force from a submerged ice bench, recent work suggests that vertical viscosity variations within the ice shelf, caused by a temperature gradient, can induce an internal bending moment that causes the front to bend upwards even without a bench. To investigate this novel bending mechanism, we present the first two‐dimensional, viscoelastic models of ice‐shelf‐front bending that assume a standard dependence of ice rheology on temperature and stress. Our results confirm the thin‐plate prediction that an ice‐shelf front can bend upwards with a sufficiently cold surface temperature (demonstrated for −30C) and high ratio of activation energy to flow‐law exponent ( kJ ). The results demonstrate that the temporal evolution of the flexural wavelength and the relationship between deflection amplitude and wavelength are consistent with thin‐plate analytic predictions, though modeled uplift gradually outpaces analytic predictions over time. Our models incorporating vertical viscosity variations and no ice bench were able to reproduce the short wavelength of upward bending observed along the Ross Ice Shelf front. Future work involving further comparison with LiDAR shelf‐front profiles, combined with models using more realistic ice rheology, may produce better constraints on the mechanisms of upward bending and the parameters controlling ice flow.
Glazer et al. (Sun,) studied this question.