Bering Glacier is rapidly retreating and thinning since it surged in 1993–1995. From 2002 to 2007 we have mapped the terminus position and measured the surface ablation from the terminus region up-glacier to the snowline in the Bagley Ice Field. Since the last surge the terminus has retreated, primarily by calving, ~0.4–0.5 km/a, and the terminus position is at the 1992 pre-surge position. The glacier surface in the terminus region is presently downwasting by melting at ~8–10 m/a and 3.5–6.0 m/a at the approximate altitude of the equilibrium line, 1200 m. The average daily melt for Bering Glacier is ~4–5 cm/d at mid-glacier, and this melt rate appears to be steady, regardless of insulation and/or precipitation. The melt from the Bering Lobe of the glacier system generates between 8 and 15 km 3 of fresh water yearly, which flows directly into the Gulf of Alaska via the Seal River, potentially affecting its circulation and ecosystem. Elevation measurements from 1957 compared with our measurements made in 2004, combined with bed topography from ice penetrating radar, show that the Bering Lobe has lost ~13% of its total mass.

A 25–30 yr surge cycle anticipated by Post (1972) was confirmed by the 1993–1995 surge, although the advance culminated more than a kilometer short of the 1965–1967 surge limit. During the initial 6 mo. of the 1993–1995 surge the eastern terminus of the Bering Glacier Piedmont Lobe advanced 1.0–1.5 km at a rate that varied between 1.0–7.4 m/d, and thickened by an estimated 125–150 m. One year after the surge began an outburst of pressured subglacial water temporarily interrupted basal sliding and slowed ice front advance. Within days gravel and blocks of ice transported and deposited by that flood partially filled an ice-contact lake, forming a 1.5 km 2 sandur. During the next few months a second outburst nearly dissected a foreland island with the resulting construction of two additional sandar, each nearly 1 km 2 . Both outburst sites coincided with a subglacial conduit system that has persisted for decades and survived two surges. When the surge resumed, advance was intermittent and slower. A prominent push moraine marks the limit of ice advance on the eastern sector. Although basal sliding across a saturated substrate was a major contributor to surge-related changes along the eastern sector, the most profound foreland alteration was the result of outburst-related erosion, deposition, and drainage modification associated with outburst floods. The dominant modification of overridden terrain was subglacial hydraulic scouring of sub-kilometer scale basins, 15–20 m deep, and outburst-related proglacial sandur development. Only after a decade of retreat was it possible to assess the limited direct effects of overriding ice, which were confined to deposition of a sub-meter-thick deformation till, decameter-scale flutes, and drumlinized topography accompanied by truncation of subglacial strata.