Hubbard Glacier is one of about 50 tidewater glaciers in Alaska that advance and retreat over several tens of miles in a cycle dominated by the water depth at the glacier terminus. Glacier advances and retreats resulting from this cycle are asynchronous because all the glaciers do not advance and retreat together under the influence of common climate change; the cycle period (usually many centuries) is unique for each glacier.
Austin Post (1975) proposed the theory that ÒInstability results when a tidal glacier retreats even a short distance into a deep basin from a stable position on a terminal shoal.Ó Instability occurs because a glacier's calving rate of icebergs is directly proportional to the water depth at the terminus (Brown and others, 1982), and if for some reason the glacier retreats from the terminal moraine the calving rate in the deeper water behind the moraine becomes greater than can be supported by ice flow (Fig. 17a). This theory has been used to explain why Columbia Glacier, 270 miles west of Hubbard Glacier, recently began a retreat that is expected to continue for decades, ultimately resulting in a retreat of the terminus of the glacier by 25 miles (Meier and others, 1985; Meier, 1986).
A glacier entering water, or remaining in water following retreat, such as Hubbard Glacier, can remain stable and again advance a great distance if the glacier deposits a protective submarine moraine shoal at the calving terminus (Fig. 17b). The buildup of the moraine reduces the iceberg calving rate. If a tidal glacier has a large excess of snow and ice in its accumulation area and a small ablation area, and it carries sufficient rock debris to fill the fiord or lake with a moraine at the glacier terminus, the glacier can thicken behind the moraine, erode the glacier side of the moraine, and begin to advance again as it redeposits the moraine debris into the water at the glacier's terminus. Thus protected, a glacier can move the terminal moraine forward and advance into a body of deep water, a process that may take centuries to complete.
Eventually, however, glacier growth ceases because accumulation of snow and ice in the accumulation area is balanced by the melting of ice in the newly enlarged ablation area. Such a tidal glacier can remain at an advanced, stable position for an indeterminate length of time. However, it will survive only so long as the snow accumulation rate in the accumulation areas exceeds both the snow and ice melting and the iceberg calving rate in the ablation area. If the glacier recedes only a short distance from its terminal moraine, an unstable retreat ensues (Fig. 17a) as Post described, and the Òcalving glacier cycleÓ is complete. This process is responsible for major glacial advances and retreats of not only tidewater glaciers but also of glaciers that calve into lakes, such as Portage Glacier near Anchorage (Mayo and others, 1977).
The recession of a tidewater glacier can be stopped in either of two ways: when a glacier retreats onto land, or when moraine debris accumulates at the terminus to form a sand and gravel deposit between the water body and the glacier terminus.
During the last century, Hubbard Glacier has deposited and moved a protective terminal moraine shoal into Disenchantment Bay. The accumulation area of the glacier represents 95 percent of the total glacier area; therefore, the terminus undoubtedly will continue to advance in the near future. A warming of climate is not expected to reverse this process because moderate climatic warming in Alaska is associated with substantially increased snowfall in winter, which results in observed glacier growth (Mayo and Trabant, 1984). Hubbard Glacier is robust and, by January 1987, it had advanced entirely over most of Osier Island; it is expected that within the next few years the glacier will have advanced across the outburst channel to form a new ice dam.
The next ice dam at Hubbard Glacier probably will be wider and stronger than the 1986 dam because of the continued advance of the glacier and because thickening has occurred along the entire terminus of the glacier. However, the ice dam may lack a substantial terminal moraine because the underlying unconsolidated marine sediments undoubtedly were eroded during the previous outburst.
Russell Fiord also could be dammed by a landslide from the fiord wall into its entrance. A major landslide could occur because the mountain was undercut by the outburst flood, the bedrock is highly fractured from previous large earthquakes, the area receives great amounts of precipitation, and earthquakes within the region are frequent and large.
The major uncertainty is whether the re-formed lake will rise higher than it rose in 1986 and thus threaten the stability of the entire terminal lobe of Hubbard Glacier. Should this happen, it could take several years for the glacier to redeposit a stabilizing terminal moraine.
Russell Lake after it forms again could potentially trigger an instability at the glacier front resulting in rapid calving and a major retreat. Russell Lake would break out through the main part of the glacier, rather than being restrained between the fiord wall and Osier Island. An outburst flood through the glacier could carry away a large part of the terminus of Hubbard Glacier, therefore, exposing the terminus to deep water behind the terminal moraine. As a consequence, an immediate retreat of Hubbard Glacier could occur.
