Notes and Discussions A. THOMAS OVENSHINE U.S. Geological Sumy, Menlo Par/(, California 94025 Observations of Iceberg Rafting in Glacier Bay, Alaska, and the Identification of Ancient Ice-Rafted Deposits ABSTRACT Observations of icebergs in a modern glacial originally disseminated fine sediment. Thus when marine environment indicate that ancient rocks freed by melting and deposited by iceberg rafting, that received iceberg-rafted material should con- the pellets would reliably indicate the presence of tain: (1) local concentrations of stones that origi- glacial ice in an ancient environment. In the nated when icebergs overturned, and (2) small Gowganda Formation, a Precambnan glacial de- pellets of till that were originally sediment filling posit, strata that contain outsized, presumably the spaces between clear ice crystals. iceberg-rafted stones also contain abundant small The till pellets are especially significant in iden- flattened clasts of unsorted graywacke interpreted tifying an ancient glacial setting because they as the lithified counterparts of the till pellets ob- originate through a process unique to glaciers—the served on modern icebergs. flow- and recrystallization-mduced segregation of INTRODUCTION THE ICEBERG-RAFTING PROCESS In recent years, geologists have discussed Icebergs up to several hundred feet long which features of sedimentary rocks reliably form when tidal glaciers calve into the fiord indicate that glaciers existed in or near an waterways of Glacier Bay. For several hours ancient depositional basin (see Harland and after calving, the bergs fragment and turn over others, 1966, for a review). Some have con- in the water until the resulting solid chunks of cluded that finely laminated sediments con- glacier ice attain a stable position. Typically an taining outsized stones that could have been iceberg maintains one position for several days iceberg-rafted are the best evidence for a until the more rapid melting below the water- glacial setting (Pettijohn, 1957, p. 274; line changes the center of gravity; the iceberg Crowell, 1964, p. 97; Harland and others, 1966, then undergoes a second episode of overturning p. 251). Iceberg-rafting in ancient sediments or fragmentation, or both. In bergs less than is identified by "... numerous large boulders 25 ft long, this process begins suddenly and penetrating and deforming a series of host commonly is complete within a minute or two; strata . ." (Harland and others, 1966, p. 251). larger bergs fragment suddenly, but the over- Unfortunately, penetration and deformation turning is usually less rapid. of strata below an iceberg-rafted stone may be Sediment having the texture of till is con- very slight (Hardy and Legget, 1960), absent tained in about 1 percent of the icebergs and is (Boutakoff, 1948, p. 53; Schenk, 1965, p. 317), released continuously from the submerged part or indistinguishable from the later effects of as it melts while drifting through the water. In compaction (Ovenshine, 1965, p. 168). Thus contrast, stones and mud in the portion of the there are ambiguities in determining whether iceberg above the waterline are not released iceberg-rafted components occur in a par- continuously but accumulate on the upper ticular sedimentary section. Observations in surface of the iceberg. Most debris-laden ice- Glacier Bay (long. 136°00', lat. 58°30'), in bergs develop flat upper surfaces; on these, southeastern Alaska, on the ways iceberg- large quantities ol mud, sand, and boulders rafting occurs and on the structure of sediment- can accumulate (Fig. 1, A and B). This sedi- laden ice suggest two criteria that may aid in ment leaves the iceberg in three ways: (1) identifying ancient iceberg-rafted deposits. Tilting, fragmentation, or overturning of the Geological Society of America Bulletin, v. 81. p. 891-894, 1 fig., March f970 891 Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/81/3/891/3432949/i0016-7606-81-3-891.pdf by guest on 26 September 2021 892 A. T. OVENSHINE—OBSERVATIONS OF ICEBERG RAFTING, ALASKA iceberg. These are the principal mechanisms as a group may be expected during settling, the observed in Glacier Bay. (2) Mudflow and resulting deposit will be a cluster of rafted slumping. Watery, almost fluid mixtures of rock stones. The sudden release of poorly sorted flour and gravel sometimes accumulate in sediment containing abundant rock flour as conical, ice-cored piles on the surfaces of well as stones might form vertical density bergs. As melting increases relief, the piles currents like those studied experimentally by become unstable and slump or flow over the Bradley (1965). If the settling debris does not side of the berg. I have watched piles of un- disperse because it is confined in a vertical sorted pebbly mud 2 ft high flow from an density current or because the water is rela- iceberg in less than two minutes. (3) Meltwater tively shallow, the resulting heap of stones rivulets. Although much fine sediment remains could mimic the "carapace" structure con- in ablation hollows on the iceberg upper surface sisting of a lenticular mass of stones enclosed until overturning occurs, near the edges small by laminated shale illustrated by Boutakoff channels funnel a more or less continuous flow (1948, p. 52, Fig. 19) from glacial deposits in of meltwater from the berg. A few of these the Carboniferous and Permian Karoo Series rivulets transport suspended sediment and roll of the Belgian Congo. sand grains, small pebbles, and till pellets over These processes in which large quantities of the side of the berg. stones are suddenly released to settle to the The processes of overturning, mudflow, and bottom are commonly observed in Glacier Bay slumping provide a mechanism for the sudden and must also have been common during the release of relatively large quantities of stones deposition of ancient lacustrine and marine that will settle rapidly to the bottom. Al- iceberg-rafted sequences. I therefore suggest though some dispersion of the clasts released that ancient rafted sequences will contain some -- -- ---- -- ------- ------- - - ---- - - Figure 1. A. Distant view of debris-laden glacial iceberg, Glacier Bay, Alaska. Iceberg is approximately 40 ft long. B. Close view of boulder-bearing ablation till on the surface of the iceberg shown in A. Largest boulder is about 2 ft long. C. Laminated argillite with iceberg-rafted clasts, Gowganda Formation, Canada. R = crystalline rock fragments, P = graywacke clasts that were unconsolidated when deposited. Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/81/3/891/3432949/i0016-7606-81-3-891.pdf by guest on 26 September 2021 NOTES AND DISCUSSIONS 893 beds or bedding planes with clusters of rafted feature of the graywacke clasts is their flatten- stones. ing and elongation in the bedding direction, whereas the crystalline rock clasts do not show THE STRUCTURE OF this feature; probably, therefore, the graywacke SEDIMENT IMBEDDED IN ICE clasts were unconsolidated when deposited. Observations on icebergs stranded by low Seen in thin section, the graywacke clasts tide suggest a second criterion that may aid in exhibit diffuse margins, whereas those of the the identification of ancient glacial deposits. crystalline clasts are distinct; this also suggests Sediment in the Glacier Bay ice is not generally that the graywacke clasts were unconsolidated. disseminated through the ice but is localised in The presence of graywacke clasts in these foliation bands, where, as seen in detail, it is Gowganda strata considerably strengthens the lodged in the interstices between ice crystals. interpretation that glaciers were nearby and These interstices are approximately equidi- that icebergs were dropping stones and till mensional and range up to 5 mm in size, pellets into the accumulating mud. averaging 2 to 3 mm. The sediment filling is The formation of the till pellets observed similar to the matrix of tills in the Glacier Bay- unconsolidated on modern glacier icebergs in region and is an unsorted mixture ot rock Glacier Bay and lithified in the Precambrian flour, angular silt, and sand. The watery, semi- Gowganda Formation is the result of processes liquid mud that accumulates on the upper that are unique to glaciers. The fine sediment surfaces of icebergs through ablation inherits entrained by a glacier, initially disseminated something of its original disposition in ice through the ice, is progressively collected and because the sediment exhibits a pelletoidal segregated into the interstices between clear texture. The pellets result from freeing of the ice crystals by processes of flow and recrystal- sediment masses between clear ice crystals by lization analogous to the segregation of mineral melting of the enclosing ice. Individual pellets phases during regional metamorphism. The are soft and plastic when wet, and friable when small pellets thus formed have sufficient co- dry, but in both states they maintain a herence to survive melting of the enclosing palpable coherence. ice and transportation and deposition into a Evidence that these till pellets can survive new environment. Organic and shore ice- transportation and deposition into another rafting, and other types of rafting, do not in- environment is provided by their occurrence volve processes that necessarily yield small along the shorelines of Glacier Bay, where they pellets of till-like sediment. Thus to the extent have been dropped by stranded icebergs and that till pellets can be recognized in ancient withstood one or more tidal cycles. Thus it is rocks, they uniquely identify the existence of probable that pellets dropped from floating glacier ice near the environment of deposition. icebergs will maintain their integrity and I therefore suggest that sediments in which identifiable character as they are incorporated part of the deposition resulted from iceberg- into the bottom sediment presently accumulat- rafting will contain small pellets of till-like, ing in the fiords of Glacier Bay. The pelletoidal unsorted graywacke. texture of bottom sediment from the iceberg- rafting environment of the Ross Ice Shelf of CONCLUSIONS Antarctica (Stetson and Upson, 1937, p.