ABSTRACT CONTRIBUTION TO THE LATE NEOGLACIAL HISTORY OF THE LYNN CANAL AND TAKU VALLEY SECTOR OF THE ALASKA BOUNDARY RANGE BY Christopher Paul Egan The sensitivity of ice to changes in the climatic factors which allow its existence enables its distribution to act as an integrator of substantial climatic informa- tion. In the morphology of existing glaciers, ice not only provides a link to past climatic events, but actively shapes and transports the denser solid materials of the earth's crust. In this way it forms a connection between glacially-produced landforms and the climatic causal fac- tors which frequently operated at levels of intensity different from those of the present. The disequilibrium of large numbers of glaciers at present attests to clima- tic change in the past few centuries. Terminus-recorded fluctuations of glacier volume and dimensions, although largely the complex result of changes in climate, are further complicated by non-clima- tic factors. These factors include geomorphic elements which originate with significant frequency in mountain Christopher Paul Egan areas formerly glaciated at widely variable levels of intensity. The diversity and effect of these factors should increase for glaciers of successively greater size and morphologic complexity. The late Neoglacial activities of a number of cirque and valley glaciers in the Juneau Icefield region are interpreted on the basis of air photo analysis, dendrochronological and radiocarbon dating, morphologic and stratigraphic relationships, and historical records. The fluctuational characteristics of these glaciers are evaluated to determine which kinds of glacier activity and which glaciers should prove to be most illuminating with respect to climatic interpretations, and which pro- cesses and glaciers have such a multifactored genesis that climatic effects are difficult to separate from non-cli- matic factors. Regionally, most Juneau Icefield valley glaciers expanded and advanced to maximum recent positions culmi- nating in the mid-18th century, indicating a previous climatic event of major significance. However, the maxi- mum recent advances of individual Juneau Icefield valley glaciers range over a period of at least three centuries (from ca. 1590-1600 to about 1910-16), and probably do not reflect the same climatic event. In contrast, large numbers of morphologically simple Juneau Icefield cirque glaciers typically exhibit a relatively simple pattern of Christopher Paul Egan three major advances in recent centuries, indicating that these glaciers may be used as more interpretable recorders or detectors of specific short-term climatic fluctuations. CONTRIBUTION TO THE LATE NEOGLACIAL HISTORY OF THE LYNN CANAL AND TAKU VALLEY SECTOR OF THE ALASKAN BOUNDARY RANGE BY Christopher Paul Egan A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Geology 1971 <7£ ,- ‘ 5.‘ if ACKNOWLEDGMENTS This study was undertaken as a part of the field program of the Glaciological Institute, Department of Geology, Michigan State University, in c00peration with the Juneau Icefield Research Program of the Foundation for Glacier and Environmental Research, Seattle, Washing- ton. Much of the field work was supported as part of the National Geographic Society Glacier Commemorative Project, carried out in the field by the FGER. The author is indebted to the Glaciological Institute and the Foundation for Glacier and Environmental Research for providing the logistic and other support by which this study was made possible. The author also wishes to acknowledge his debt to earlier workers on the Juneau Icefield, whose published and unpublished data aided the present study. The author greatly appreciates the generous assistance given by members of the 1964, 1965, 1966, and 1967 field programs. Sincere thanks are given to Louis Acker, Gregg Barwis, Harte Bressler, Patrick Egan, Roger Frame, Scott Hulse, Ernest Kuhlman, Ross Mack, Andy Miller, Norton Miller, Darrell Mintz, ii William Patzert, Fritz Rohrman, William Savage, Don Thomas, Bert Useem, Walt Vennum, Steve Walasek, Patrick Welsh, and Grant Yakopatz. The interest of Dr. Andy Broscoe and the late Dr. R. E. Beschel in certain aspects of this study is gratefully acknowledged. Special thanks are also given to Arlo Livingston of Livingston Copters, Inc., and to Ken Loken of Channel Flying Service, Inc. The author wishes to express deep appreciation to his chairman, Dr. Maynard M. Miller, for arranging support of the field work and for guidance in the pre- paration of this manuscript. The author is especially grateful for the help and encouragement generously given by Dr. Robert Ehrlich, and to Dr. Chilton E. Prouty and Dr. Aureal T. Cross for serving on his committee. iii TABLE OF CONTENTS Page ACKNOWLEWMNTS O O O O O O O O O O O O O O i i LIST OF TABLES . . . . . . . . . . . . . . vii LIST OF FIGURES O O O O O I O O O O O O O .Viii LIST OF APPENDICES . . . . . . . . . . . . . Xi Chapter I. INTRODUCTION AND GENERAL OBJECTIVES . . . . 1 II. REGIONAL SETTING . . . . . . . . . . . 5 III. THE CONTROL OF FLUVIAL AND GLACIAL VALLEY MORPHOLOGY BY BEDROCK STRUCTURAL ELEMENTS: LEMON CREEK DRAINAGE BASIN . . . . . . . 9 Introduction . . . . . . . . . . . 9 Location and General Description of the Lemon Creek Drainage Basin . . . 10 Bedrock Structural Elements . . . . . . 15 Drainage Basin Parameters . . . . . . 18 Relationships Between Bedrock Structural Elements and Drainage Basin Development. 20 Development of Glacial Valley Morphology. . 24 IV. JUNEAU ICEFIELD CIRQUE GLACIATION, WITH SPECIAL REFERENCE TO THE ANTLER AND GILKEY RIVER DRAINAGE BASINS . . . . . . . 31 Introduction and Objectives . . . . . . 31 Morphologic Effects of Bedrock Structural Elements . . . . . . . 35 Present and Former Terminal Positions of Glaciers in Cirques and Tributary Valleys . . . . . . . . 40 Summary of Glacio-Geomorphic Observations . 47 iv Chapter Page V. GILKEY GLACIER. . . . . . . . . . . . 52 Location and Description . . . . . . . 52 Lateral Moraines and Trimlines . . . . . 54 Terminal Zone. . . . . . . . . . . 55 Proglacial Lake and Valley Train Relationships . . . . . . . . . 58 Summary. . . . . . . . . . . . . 60 VI. ANTLER GLACIER. . . . . . . . . . . . 62 Location and Description . . . . . . . 62 Map and Aerial Photo Coverage . . . . . 63 Aerial Photo Interpretations. . . . . . 64 Field Results in the Terminal Moraine Zone. . . . . . . . . . 68 Recommendations for Future Work. . . . . 74 VII 0 BUCHER GLACIER. O O C O O O O O O O C 75 Location and Description . . . . . . . 75 Trimline Observations . . . . . . . . 76 Late l6th-Century Moraines . . . . . . 76 VIII. NORRIS AND TAKU GLACIERS . . . . . . . . 79 Location and Description . . . . . . . 79 Early Historical Observations of Norris Glacier . . . . . . . . 81 Interpretations of Pre-1900 Activity of Norris Glacier . . . . . . . . 83 Taku Glacier . . . . . . . . . . . 89 Maximum 18th Century Advance Limit of Taku Glacier. . . . . . . . . 91 IX. DAVIDSON GLACIER . . . . . . . . . . . 94 Location and Description . . . . . . 94 Map Coverage and Aerial Photography . . . 95 Late Neoglacial Activity of Davidson Glacier . . . . . . . . 96 Dendrochronological Evidence of the Mid-18th Century Maximum Recent Advance and Subsequent Recession of the Davidson Glacier . . . . . . 111 Chapter Page X. GENERAL MODEL OF JUNEAU ICEFIELD VALLEY GLACIER ACTIVITY. . . . . . . . . 116 Equilibrium State . . . . . . . . . 116 Advancing Condition. . . . . . . . . 119 Recessional Conditions. . . . . . . . 123 XI. CONCLUSIONS. . . . . . . . . . . . . 128 REFERENCES. . . . . . . . . . . . . . . . 164 APPENDICES O O O O O O C O O O C O O O O O 1 6 9 vi Table 1. LIST OF TABLES Page Principal joint surface orientations in the Ptarmigan Glacier area (listed in approximate decreasing order of frequency). . . . . . . . . . . . 16 Principal low-order stream segment orientations in the Antler-Gilkey drainage basin (listed in order of decreasing frequency) . . . . . . . . 36 Thickness and description of trenched section near Antler Glacier (Site AG-66-ex la) . . 73 Thickness and description of trenched section near Antler Glacier (Site AG-66-ex 1b) . . 73 vii Figure l. 2. 10. 11. 12. 13. 14. 15. 16. 17. 18. Taku River LIST OF FIGURES Juneau 1:250,000 topographic map . Skagway 1:250,000 topographic map. . Taku River l:250,000 topographic map. Juneau B-l, 1:63,360 quadrangle . . Juneau B-2, 1:63,360 quadrangle . . Juneau C-l, 1:63,360 quadrangle . . Juneau C-2, 1:63,360 quadrangle . . Juneau C-3, 1:63,360 quadrangle . . Juneau D-l, 1:63,360 quadrangle . . Juneau D-2, 1:63,360 quadrangle . . Juneau D-3, 1:63,360 quadrangle . . Skagway A-2, 1:63,360 quadrangle . . Taku River B-S, 1:63,360 quadrangle . B-6, 1:63,360 quadrangle . Taku River C-S, 1:63,360 quadrangle . Taku River C-6, 1:63,360 quadrangle . Vertical air photo of Gilkey Glacier terminal zone (July 4, 1962) . . Vertical air photo of Antler Glacier terminal zone (July 4, 1962), approximate 1948 ice front position also indicated. . . . . . . viii In In In In In In In In In In In In In In In In Page pocket pocket pocket pocket pocket pocket pocket pocket pocket pocket pocket pocket pocket pocket pocket pocket 132 134 Figure Page 19. Vertical air photo mosaic of the Norris and Taku Glacier terminal zones (July 17, 1962) . . . . . . . . 136 20. Vertical air photo mosaic of Davidson Glacier terminal zone (June 30, 1963). . . 138 21. Vertical air photo mosaic of Wright Glacier terminal zone (July 17, 1962), approximate 1948 ice front position also indicated . . . . . . . . . . 140 22. Longitudinal profile of Lemon Creek and principal tributaries . . . . . . . . 142 23. Joint orientation diagrams for selected Juneau Icefield locations: (a) Camp 18 nunatak area; (b) Camp 10 nunatak area; (c) Ptarmigan Glacier area; and (d) Glory Lake outlet/lower Norris Glacier area . . . . . . . . . . . 144 24. Orientation and cumulative length of straight channel segments in Lemon Creek drainage basin: North side of Lemon Creek valley (a) first-order channel segments; (b) second-order challen segments; (c) third- order channel segments; and South Side of Lemon Creek valley: (d) first-order channel segments; (e) second-order channel segments; and (f) third-order channel segments . . . 146 25. Orientation of geomorphic elements in the Antler and Gilkey drainage basin: (a) orientation and cumulative lengths of low-order straight stream channels; (b) orientation and cumulative lengths of large, straight glaciated valley segments; (c) orientation of well- developed glaciated and deglaciated cirques; and (d) orientation of active cirque and cliff glaciers (Juneau C-2, C-3, D-2, and D-3 sheets). . . . . . . 148 ix Figure Page 26. Terminal elevation variability of active cirque and cliff glaciers (Juneau C-2, C-3, D-2, and D-3 topographic sheets): (a) data plot for individual glaciers; and (b) mean terminal eleva- tions plotted for 900 arcs centered at the north, east, south, and west . . . . 150 27. Tree growth rate curves (a) Picca stitchensis, Antler Glacier terminal zone; (bT—Tsuga heterophylla, Antler Glacier terminal zone; (c) Tsuga heterophylla, Antler Glacier terminal zone; (d) Tsuga heterophylla, Antler Glacier terminal zone; and (E) Tsuga mertensiana, Bucher Glacier terminal zone . . . . . . . . 152 28. Stratigraphic sections along south side of Davidson Glacier proglacial lake outlet stream. . . . . . . . . . . 154 29. General view of Davidson Glacier buried forest . . . . . . . . . . . . . 156 30. Stumps overridden by Davidson Glacier during the early 12th century, and exhumed by proglacial drainage in recent decades . . . . . . . . . . 158 31. Trenched sections showing upper till, 12th- century forest floor litter, and lower till exposed on south side of Davidson Glacier proglacial lake outlet stream. . . 160 32. Upper till, containing logs, overlying 12th- century forest floor and lower till, and currently being exhumed by Davidson Glacier proglacial lake outlet stream. . . 162 LIST OF APPENDICES Appendix A. Stream parameter data for the currently nonglaciated part of the Lemon Creek drainage basin and for selected basin components. . . . . . . . . B. Cirque and cliff glacier orientation data. . C. Descriptions of trenched sections in Norris Glacier Outwash. . . . . . . D. Descriptions of trenched sections along Davidson Glacier proglacial stream. . . B. List of valley glaciers in the vicinity of the Juneau Icefield and dendrochronolo- gically dated late Neoglacial advances . F. Wright Glacier reconnaissance study. . . . xi Page 170 177 179 182 191 193 CHAPTER I INTRODUCTION AND GENERAL OBJECTIVES Ice, as a result of its low density and melting point relative to other solid earth materials, occurs on the planetary surface with a dynamic distribution closely related to climatic factors. The relationship between climate and ice is consequently of great significance as it provides a means by which an understanding of one enhances the interpretation of the other. The distribution of ice, sensitive to changes in the climatic factors which allow its temporary existence, acts as an integrator of substantial climatic information. In the morphology of existing glaciers, ice not only pro- vides a link to past climatic events, but actively shapes and transports the denser solid materials of the earth's crust. In this way it forms a connection between glaci- ally-produced landforms and the climatic causal factors which frequently operated at levels of intensity differ- ent from those of the present. The disequilibrium of large numbers of glaciers at present attests to climatic ' . D. g u . Q. . -~ change in the past few centuries. Understanding the nature of current glacier disequilibrium, which appears with special clarity throughout middle and high-middle lati- tude mountain areas, is complicated by the great dimen- sional and morphologic range of existing glaciers as well as by various local and regional factors. Our knowledge of past climatic conditions through an interpretation of glacial landforms and deposits will always be limited to the extent that glacial processes are not strictly linked to climatic change. It is likely that some glacial and glaciofluvial processes (and their related landscapes) are more directly sensitive to widespread climatic change than others. It is one of the principal purposes of this thesis to evaluate the characteristics of a spectrum of glaciers in order to determine which kinds of glacial activity and which glaciers should prove to be most illuminating with respect to climatic interpretation and, contrariwise, which processes and glaciers have such a multifactored genesis that climatic effects are difficult to separate from those arising from a multiplicity of other causes. The growth and decline of glaciers is most con- spicuously revealed by changes in surface profile at the terminus (Meier, 1965). In addition to terminal advance and retreat, these manifestations include concomitant development of characteristic landforms. The interpretation of paleoclimatic factors based on terminal fluctuations of valley glaciers is greatly complicated by series-connected intermediary processes which generally are incompletely known for individual glaciers. The principal processes involved in this cli- mate change/glacier terminal reSponse problem have been discussed critically by M. F. Meier (1965), who lists them in sequence as follows: (a) general meteorologic environment (b) local mass and energy exchange at glacier surface (c) net mass budget of glacier (d) dynamic response of glacier (e) lasting evidence of position of glacier margin. That the detailed geologic and geographic environ- ment of a glacier affects the response of that glacier to climate is implicit in these factors. With this in mind, it is evident that the value of large valley glaciers for paleoclimatic interpretations is not only affected by a wider geologic and geographic terrain, but is also con- trolled to a great extent by a set of tributary glaciers. The value of glacier terminal fluctuations for paleoclimatic interpretations should be further reduced significantly in the case of large valley glaciers. In complex valley glacier systems, the geomorphic and glaci- ologic variables of these individual tributaries are trans- mitted to and partially absorbed by the resultant main ice stream. Thus the response of large valley glaciers may be thought of as the sum of a great complex of terms; in contrast, cirque glaciers occupy smaller and more uniform terrain, and consequently should exhibit a simpler and more direct relationship to climatic change. ‘ Because of the facility with which ice changes state, investigations of present or former glaciation necessarily involve more than the phenomenon of glacier ice and must include an assessment of the closely related fluvial effects which often extend far beyond areas invaded by ice. The primary objective is to examine the relative value in the Juneau Icefield area of high-elevation cirque glaciers with that of larger valley glaciers in terms of reliability in response to changing climatic factors. For this purpose, it is necessary to develop general models for the behavior of both glacier types. The first of these models may be developed by synthesizing observations from a variety of cirque gla- ciers. Factors to be considered in this synthesis include specific elements such as bedrock control, geographic position, orientation, glacier length and terminal eleva- tion. These will then be compared to glaciers of suc- cessively larger dimensions in order to assess the rela- tive importance of extra-climatic factors in complicating glacier response. In addition, this will result in a general response model for large valley glaciers. CHAPTER II REGIONAL SETTING The Juneau Icefield region of southeast Alaska (Figures 1, 2, and 3) is well suited for this particular type of study because of its coastal mountain position, placing it in a zone of frequent interaction between maritime and continental climatic influences. As a con- sequency of high precipitation rates and generally moderate temperatures, active glacier regimens are pro- moted. The substantial elevation range of this segment of the Coast Range, with peaks reaching to the 2000 to 2600 meter level and with broad accumulation areas at up to 2000 meters, allows the development of glaciers having great morphological and dimensional variability. The assemblage of glacier models at different scales is further facilitated by the fact that during the past two centuries, most of these glaciers have been in a condi- tion of predominant thinning and terminal recession, and consequently exhibit a wide spectrum of recessional modes and stages at the present time. In addition, the Juneau Icefield glacier system includes one advancing glacier (Taku Glacier, plus its distributary branch, Hole-in- the-Wall Glacier), of particular value in revealing the special results of such activity. The bedrock geology of much of the study area remains unmapped except on a reconnaissance basis. 'Part of the western periphery of the icefield and the coastal zone has been mapped by Knopf (1911, 1912), Lathram, gt 31. (1959), and Spencer and Wright (1906), largely because of the presence of gold and other mineral deposits. Areas east of the icefield have been mapped by Aitkin (1955, 1959), Cairnes (1913), Gwillim (1901), Kerr (1948), and Souther (1960). The southern part of the icefield is currently being mapped at a scale of 1:63,360 by the U. S. Geologi- cal Survey (A. Ford and D. Brew, personal communication, 1968). From the coast eastward, the general bedrock setting consists of a variety of folded and faulted meta- sediments generally striking NNW-SSE and increasingly metamorphosed toward the east. Much of the crestal axis of the Coast Range is dominated by granodiorite. The eastern periphery is dominantly metasedimentary, with somewhat less structural complexity than is present on the western side of the range. The importance of faulting, jointing, and the strike of alternating resistant and nonresistant beds in controlling drainage patterns and other large-scale geo- morphic features is evident in Figures 1, 2, and 3. Most of the central part of the icefield con- sists of broad, low-gradient snow accumulation areas locally punctuated by nunataks. This central area is drained by valley glaciers arranged in a radial pattern. These glaciers generally are longer and extend to lower elevations on the western and southern (windward) sides of the range. Most of these glaciers are currently receding from a maximum recent advance which characteristically culmi- nated in the mid-18th century. This pattern is complicated by numerous earlier and later advances by different gla- ciers at different times. Because of a favorable combination of topographic relief and cirque abundance, the cirque glaciers to be discussed in this report are located on the western peri- phery of the icefield (Figures 7, 8, 9, 10, and 11). Valley glaciers to be discussed do not include all of those emanating from the icefield, but were sel- ected to provide a realistic variety of sizes, configura- tions, and recent terminal response patterns. These glaciers include the Antler--Bucher--Gilkey valley glacier complex on the western side of the icefield (Figures 9, 10, 11, 17, and 18), and the Norris--Taku system (Figures 4 and 19) on the south side. Reference is also made to Davidson Glacier (Figures 12 and 20), and Wright Glacier (Figures l3, 14, 15, 16, and 21), which emanate from separate icefields located respectively northwest and southeast of the Juneau Icefield. With these glaciers it is possible to examine the factors responsible for their individual behavior patterns, and to compare these patterns with those exhibited by cirque glaciers. CHAPTER III THE CONTROL OF FLUVIAL AND GLACIAL VALLEY MORPHOLOGY BY BEDROCK STRUCTURAL ELEMENTS: LEMON CREEK DRAINAGE BASIN Introduction Expanding valley glacier systems--whether attri- butable to regional uplift or to climatic deterioration, or both--are strongly guided in their subsequent morpholo- gical development by the existance of valleys developed initially by streams. In view of this, the present paper begins with an evaluation of the effects of certain bed- rock structural factors on the fluvial morphology of a small drainage basin in the study area. The effects of these local structural elements and of stream valleys on the morphology of former and existing glaciers will then be assessed. The Lemon Creek drainage basin was selected for this phase of the present study for several reasons, including (a) its location in an area displaying obvious structural control of stream patterns and valley asymmetry, (b) the presence, on the south side of the valley, of a 10 sequence of tributary valleys having successively higher floor elevations and displaying progressively more recent and intense effects of glaciation, (c) the possible rele- vance of these results to continuing hydrological studies of Lemon and Ptarmigan Creeks and glaciological studies of Lemon Creek and Ptarmigan Glaciers, and (d) unusual logistical advantages which make the area readily accessi- ble for future, more comprehensive work along the lines of the present study. The obvious logistical advantages of Lemon Creek valley's location with respect to Juneau and ,the Douglas heliport are further enhanced by an access trail along the trunk stream, and by the presence of one main and two subsidiary field stations of the Foundation for Glacier and Environmental Research (v. map, Figure 5) in the upper part of the basin. Location and General Description of the Lemon Creek Drainage Basin Overall Lemon Creek Basin The Lemon Creek drainage basin is located near the southern periphery of the Juneau Icefield, less than 10 km north of Juneau and about the same distance east of Juneau Airport (Figure 1). The complete basin area is shown on the 1:63,360, 100 ft. (30.48 m.) contour interval Juneau B-1 and B-2 quadrangles (Figures 4 and 5, respec- tively). A large scale map (1:10,000) of Lemon Creek Glacier is also available (American Geographical Society, 11 1960). Aerial photo coverage of the overall basin area was taken for the U. S. Geological Survey on July 5 and August 25, 1948, for the American Geographical Society on September 7, 1955 and September 18, 1957, and for the U. S. Forest Service on July 16, 1962. The drainage basin area is strongly asymmetri- cal, with a steep north side characterized by short, steep-gradient streams having a distinctly parallel pat- tern, and a south side occupied by a series of five main tributaries: (a) Sawmill Creek, (b) unnamed, (c) Canyon Creek, (d) Ptarmigan Glacier and its proglacial stream, and (e) Lemon Creek Glacier. As a group, these five val- leys may be regarded as a small-scale chronosequence of valley development by small glaciers, and of subsequent "reclamation" by streams as the glaciers recede. This reclamation is especially evident in the lower and middle reaches of the unnamed valley and Canyon Creek valley, where the broad glaciated valley floors are being con- spicuously trenched in a headward direction by their trunk streams. Drainage patterns in the first three of these small valleys tend to be primarily dendritic, but with elements of parallelism. The longitudinal profile of these principal streams is depicted in Figure 22. The three glaciers (Ptarmigan, Lemon Creek, and Thomas Claciers) which dominate the high eastern end of the Lemon Creek basin together occupy 18.42 km2 (7.11 miz), 12 or ca. 29.5 percent, of the total basin area of 62.34 km2 (24.07 miz). All three glaciers are receding at present. Lemon Creek Glacier: Dimensional CharacteriStics and Historical Fluctuations Lemon Creek Glacier is about 7 km (4.3 mi) in length, 2 km (1.25 mi) in width, and has a surface area of 12.64 (km2 or 4.88 mi2 (measured by planimeter from the l/63,360 Juneau B-1 and B-2 quadrangles, 1962 ed., based on 1948 aerial photography). The accumulation area is aligned north-south, and is connected at several minor snow divides with small glaciers contributing to the Norris Glacier system lying to the east. Another snow divide exists near the southern end of Lemon Creek Gla- cier, from which ice moves southward toward the Salmon Creek basin. This end of the glacier has constructed two small but significant moraines (Pierce, 1953). Lemon Creek Glacier's northern end curves westward, moves down a steep icefall, and terminates at about the 860 m (1500 ft) elevation. The terminal zone of this glacier was not examined by the writer. Heusser and Marcus (1964b) have tenta- tively dated the recession from a maximum recent position as having begun ca. 1750, and have summarized the known fluctuations of this glacier to its 1957-58 position. Heusser and Marcus also reported a radiocarbon date of 13 10,3001600 yr for basal peat in a muskeg ca. 375 m (1230 ft) below the 1750 terminal position. Glaciological studies with various objectives have been conducted on this glacier by Andress (1962), Heusser and Marcus (1964a), Marcus (1964), and Wilson (1959). The thickness and subglacial configuration of Lemon Creek Glacier along several profiles has been deter- mined by Thiel, La Chapelle, and Behrendt (1957). Ptarmigan Glacier: Dimensional Characteristics and Historical Fluctuations Ptarmigan Glacier is about 3.06 km (1.9 mi) in length, generally 0.8 km (0.5 mi) in width, and has a total surface area of 2.10 km2 or 0.81 mi2 (measured by planimeter from the 1/63,360 Juneau B-2 quadrangle). Ptarmigan valley lies above treeline, and as a result precise dating of its fluctuations is not possible. Pierce (1953) conducted a reconnaissance study of bedrock geology and soil and vegetation development at several high trimline positions above the west edge of the gla- cier, and concluded that "the condition of soil and vege- tation development point to a much more recent maximum than the 200 year old maximum." Heusser and Marcus (1964b) depict this glacier as having been connected with Lemon Creek Glacier at the time of its maximum of about 1750. This may have been the case, but during a brief examination of the lower Ptarmigan valley in 1967, the writer's provisional 14 interpretation of soil development and bedrock weathering at Camp 17A and of lichen development on talus slopes to the east of Camp 17A was that the ice maximum of ca. 1750 closely approached, but stopped short of, the 17A site. Erratics of massive gneiss identical with that composing the main part of the east wall of Ptarmigan valley are present on the site of Camp 17A, but these blocks may have reached this area by sliding down snow banks at the base of the east side of the valley, or they may be erratics from a pre-18th century advance. Several conspicuous end moraines are present within a few hundred meters of Camp 17A. As a result of erosion by Ptarmigan Creek, the outermost definite ice limit is marked by morainal deposits at only a few loca- tions, but several isolated boulders preserve the general trend of the former moraine near the 17A stream gaging station. (Subtle indications of a slightly more extensive ice limit are present 18 to 25 meters downvalley from this definite position). Lateral moraines attributable to this ice position are obscure, particularly on the east side of the valley, where talus is accumulating. The second definite end moraine is located ca. 58 m (190 ft) upvalley, and although obscured in places by slumping and stream erosion, is clearly defined. This moraine diverges in places into two small ridges up to 11 m (35 ft) between crests, but these may be closely related in age. 15 When examined on September 10, 1967, the terminal margin of Ptarmigan Glacier was obscured by well-compacted old snow or firn still remaining from at least two preced- ing winters. The presence of this snow on the glacier terminus near the close of a strongly negative budget year (the 1966-67 snowpack had melted from nearly the entire glacier surface by this time) is attributed to a combination of minor avalanching and effective shading at this locality. Inspection of the terminal zone disclosed numerous structural features indicating that substantial downwasting had been in progress for many years. These criteria include the poor development of overthrusts and shear crevasses, the low longitudinal ice surface gradi- ent, and a concave transverse profile. Several tens of meters upvalley from the terminus, numerous small moulins and stream channels had been incised completely through the glacier, revealing ice thicknesses on the order of only 2 to 5 meters. Continued rapid recession is clearly unavoidable under such conditions. Bedrock Structural Elements The writer's bedrock observations in this area were limited to a brief reconnaissance of rock type, structural attitude, and joint surface orientations at 18 locations along the ridges flanking Ptarmigan Glacier. These locations were distributed along 1 km of the west margin and 3 km of the east margin of Ptarmigan Glacier. Additional data from Pierce (1953) were utilized to extend 16 the coverage for another 2 km along the west side of the upper Ptarmigan Glacier. In the vicinity of Ptarmigan Glacier and the west edge of Lemon Creek Glacier, bedrock consists pre- dominantly of metasediments, including (in approximate order of decreasing resistance to weathering and erosion) various undifferentiated gneisses, ferruginous quartzite, quartz-mica schist, hornblende schist, and hornblendite. The principal bedrock structural elements in the locality of Ptarmigan Glacier include (a) the bedrock strike (generally NW-SE to N-S), (b) the steep dips (generally between 400 to 650 toward the NE or E), and (c) the presence of morphologically important joint sets (v. Table 1 below): TABLE l.--Principal joint surface orientations in the Ptarmigan Glacier area (listed in approximate decreasing order of frequency). Joint Characteristic Percent orientation dip Number (based on 150) N 40 - 70 E 65° to 85° NW 62 41.3 N o - 30 w 60° to 80° wsw 34 22.6 N 40 - 60 w 50° to 75° sw 17 11.3 N 20 - 30 E 45° to 85° NW 10 6.7 Regionally, a fourth main element is faulting, with dominant trends striking N 35 - 40 W, N 40 - 45 E, and N 5 W to N 10 E (data from Lathram, et al., 1959). The l7 ridge north of Vesper Peak is cut by several normal faults striking rough E - W or NE - SW. The writer has not had the opportunity to examine other ridges in the Lemon Creek basin, but examination of bedrock at the Mendenhall Glacier terminus 12 km (7.5 mi) to the northwest of Camp 17, and along the Glacier Highway and Thane Road to the south and southeast conform generally with this structural attitude (v. map, Figure 5). This general structural background is also reflected by the strongly developed asymmetry of nearby valleys (e.g., Steep Creek, Nugget Creek, Fall Creek, and Goat Creek valleys to the northwest, and of Salmon Creek, Gold Creek, and Gastineau Channel to the south and southeast). The dominant joint directions observed near Ptarmigan Glacier, as well as the strike of major faults mapped farther west and northwest by Lathram, et_al. (1959), are closely paralleled by unusually straight stream channel segments over much of the area shown on the Juneau B-2 quadrangle (Figure 5). For this reason, much of the emphasis in the following sections will be on the specific effects of jointing and faulting on channel orientations and valley morphology. The structural elements outlined above integrate closely with those expected in an area formerly subjected to compression along an approximately NE - SW direction: fold axes and thrust faults (but very few joints) trending NW - SE; normal faults and large numbers of joints, quartz 18 veins, and dikes trending NE - SW; strike-slip faults trending roughly N - S; and a minor joint set following directions of maximum shearing (v. Figure 23c). Drainage Basin Parameters The presence of active glaciers and large snow- covered areas in the eastern third of the Lemon Creek drainage basin precludes a conventional total-basin col- lection of stream parameter data, but by the expedience of excluding the uppermost drainage area not containing 2 or 9.84 mi2 area glacier ice (specifically, the 25.5 km draining to any part of Lemon Creek Glacier), the remain- ing area may be treated independently, although certain characteristics of the main trunk stream (e.g., its higher order, discharge, and erosive capability) are no longer comparable. The stream parameter data (Appendix A) were derived from the Juneau B-2, l/63,360, 100 ft. 0.1. quadrangle (Figure 5). After drainage basin limits were defined, basin areas were measured with a standard polar planimeter. On the basis of contour line configura- tion, probable stream channel segments not mapped as permanent streams were inferred. All obvious and inferred channel segments were assigned a stream order and measured to the nearest 0.08 km (0.05 mi). Because of the Lemon Creek valley's asymmetry, stream parameters from both sides of the master stream were compiled for comparison (Appendix A-2 and A-3). An additional compilation was also made for the three 19 successively higher and more recently deglaciated valleys in the south half of Lemon Creek valley (Appendix A-4, A-5, and A-6). Comparison of stream parameter data from the north and south sides of Lemon Creek valley reveals few signi- ficant differences except for mean lengths of orders. On the steep north side, mean lengths of first- and second- order channels are longer, while third—order channels are much shorter than corresponding orders on the lower- gradient south side. This contrast is also indicated by the unusually low length ratio between second- and third- order channels, and appears to be merely a reflection of the generally short valley length on the north side, where third-order segments become tributaries of Lemon Creek near their origins. Comparison of stream parameters for the three suc- cessively larger, less steep, and more recently deglaciated valleys south of Lemon Creek (Appendix A-4, A-5, and A-6), also does not reveal obvious, systematic change of the parameters which might be of general application. An unusually high and increasing bifurcation ratio between second- and third-order channels does appear, and suggests the possibility that the "U-shaped" or parabolic main valley cross-section of deglaciated valleys may strongly raise low-order bifurcation ratios. Mean channel length of low—order channels may be generally minimal in such valleys. 20 Relationships Between Bedrock Structural Elements and Drainage Basin Develgpment Relationships between jointing and stream channel orientation were assessed for channel segments of first, second, and third orders on both the north and south sides of Lemon Creek valley. The orientation (in terms of 100 sectors, starting at 00) and cumulative frequency of 150 joints measured on Ptarmigan valley ridges are plotted in Figure 23c. The length (to the nearest 0.08 km, or 0.05 mi) and orientation (in terms of 100 sectors) of all straight channel segments greater than 0.16 km (0.10 mi) are plotted cumulatively in Figure 24 a-f, according to order, for channels on each side of Lemon Creek. First-Order Channel Segment Orientation The flow directions of straight first—order chan- nel segments (total measured length = 18.5 km, or 11.5 mi) as plotted for the north side of Lemon Creek valley (Figure 24a) do not initially appear to relate to those of the south side (Figure 24d). Channels on the morpholo- gically simple north wall are strongly orientated toward the southeast, as would be expected for consequent streams on the northwest side of a SW-trending deglaciated valley. The large percentage (71.2%) of total measured channel length in this.40O arc strongly suggests a bedrock con- trol factor, and coincides generally with the orientation of the second most frequent joint direction (Figure 23c). However, the most strongly preferred stream orientation 21 (S 300 - 400 E) is seldom coincident with joint surfaces, although many joints lie to either side of this direction. Although in part attributable to the lack of direct joint measurements north of Lemon Creek, the regional bedrock strike in approximately the same direction as the channels implicates this as a principal control. Straight channels trending S 00 - 100 W are closely matched by joints trend- ing N 100 - 200 E. Also, the few north-trending channels all follow the predominant ENE joint orientation. The first-order channel orientation plot for the south side of the main valley is based on.a total of 19.0 km (11.8 mi) of measured straight segments (Figure 24d). This plot, strikingly unlike that of the north wall, clearly reflects the presence of the large tributary valleys south of Lemon Creek and a more.uniform exploita— tion of joint orientations. Joints oriented N 30 W to N 10 E, and N 30 to 40 E are exploited, respectively, by about 34 percent and 14 percent of the total channel length. The 21.2 percent of channel length between N 70 W and S 80 W is, however, obviously not relatable to joint- ing. In the absence of more adequately distributed field observations, this orientation is provisionally ascribed to control by bedrock strike. Second-Order Channel Segment Or1entat1ons Second-order channel orientations on each side of Lemon Creek valley (Figure 24b and c) closely parallel 22 those of the first-order channel on respective sides of the valley. Second-order channels on the north wall follow the N 0 - 30 W joint orientation in a general way, but show less dispersion than these joints, suggesting that an additional control factor, less variable over short distances, might be involved. The regional bed- rock strike is probably the principal control factor here. South of Lemon Creek, the N 0 - 30 W, N 40 - 70 E, and N 20 - 30 E joint orientations appear to be the most utilized. Third-Order Straight Channel Segment Orientations At the third order, the small number and lengths of straight channels obviously limits their significance. Nevertheless, the third-order straight channel length north of Lemon Creek continues and intensifies the pattern set by lower orders and is entirely restricted to S 30 - 40 E. Significantly, this closely parallels the regional bedrock strike. The third-order plot for straight channel segments south of Lemon Creek strongly contrasts with lower-order plots, and instead shows a tendency for NW orientations, a direction rarely taken by first- or second-order channels. This may result partly from exploitation of the N 0 - 30 W joint orientation, but the general NW strike of bedrock appears to be the main fac- tor in the small valleys south of Lemon Creek at the third- order level. It is not as significantly a factor at lower orders. 23 Channel Segment Orientations: Summary The plots described above reveal significant general tendencies as different orders are compared. The reduced dispersion of preferred orientations with increas- ing channel order is best exemplified on the north side of the valley, where the S 30 - 40 E orientation is taken by 31.7 percent of the first-order, 46.7 percent of the second-order, and 100 percent of the third-order straight segment lengths. The conclusion appears unavoidable that there is strong structural control on the fluvial drainage pattern. Similar control presumably existed before the onset of glaciation, causing fluvial systems to react in generally the same way in which the modern streams are reacting. During a time of climatic change in which glaciers are being initially developed or existing glaciers are expanding to lower elevations, it is reasonable to expect that the orientation and morphology of these glaciers would be guided by the previously-developed stream-carved valleys. The role of bedrock type and attitude usually is less clear than that of the joints, and cannot be ade- quately evaluated from the limited data available. A more fundamental difficulty results from the great variability of rock type and structural attitude in local areas (e.g., -1 10 to 100 kmz), particularly in areas that have been tectonically active. Such regions may display a much 24 greater degree of uniformity in joint systems over areas at least up to 102 or 103 kmz. This uniformity is illustrated by the similarity of joint spectra for Camp 10 nunatak, Glory Lake outlet, Camp 18 nunatak, and the Ptarmigan Clacier areas (Figure 23a, b, c, and d). Stated another way, in the mosaic of jointed and faulted bedrock in southeast Alaska, a greater degree of regional order or conformity is found in the joint and fault orientations than in the bedrock structural attitude. Development of Glacial Valley Morphology Initiation of Cirque Glaciation The initial incisions probably included consequent stream valleys developed on a surface of unknown relief (and speculative configuration), but the principal orientating factors of the present time--jointing, faults, and the strike of less-resistant strata--probably were exploited by headwaters of ancestral Lemon Creek early in the uplift cycle. In particular, the regionally consis- tent joint and fault direction (NW-SE) which has been shown to exert a strong control on the orientation of low- order stream channels in the Lemon Creek basin, has largely determined the overall morphologic grain of southeast Alaska up to the scale of the major fiords. The subsidiary joint (and fault) directions (NE-SW and NNE-SSW) similarly affect low-order stream orientations, often acting significantly on second- and third-order 25 basin development by intersecting the high-frequency first-order NW-SE channel segments. The initial geometry of cirque glaciation at high elevations, although restrained by favorable orientations with respect to prevailing accumulation season storm dir- ections and by shading during the ablation season, would generally be further guided in orientation by second- order stream valleys. First-order streams are relatively less significant at this initial stage of glaciation because of their small valley size. Expansion of incipi- ent cirque glaciers into small valley glaciers (ca. 1.5 to 3.0 km lengths) generally will involve valleys of third-order streams. Glacier lengths on the order of ca. 3.0 to 10.0 km (or more) would generally occupy valleys of fourth- (or higher-) order streams. Orientation of Ridges Flanking Ptarmigan Glacier The ridges forming the divide around the Ptarmigan drainage basin relate closely in orientation to structural factors. Two main ridge orientations occur on each side of this basin. On the west side, from Cairn Peak north- ward, the first several hundred meters of ridgeline trends N 55 W, conforming closely both to the local bedrock strike (N 30 W to N 50 W) and to the N 40 - 60 W joint orientation. The remaining part of this ridge trends irregularly toward the north and north-northwest (much of this irregularity is due to enlargement of cirques along 26 the ridge) and is attributable to the strongly developed N 0 - 20 W joint orientation. Locally, NE-facing slopes on this ridge are essentially bedrock dip slopes, with relatively simple morphology compared with the irregular opposite part of the ridge facing Canyon Creek valley. Many of the cirque walls and first-order stream-carved valleys facing toward Canyon Creek are in part locally orientated by jointing. On the east side of Ptarmigan Glacier, the Cairn Peak-Vesper Peak ridge segment does not closely relate to the bedrock strike, but closely parallels the NW- dipping N 20 - 30 E orientation of joints which are well represented along this ridge. The conspicuous northwest- facing slope of much of the upper Ptarmigan Glacier sur- face also is primarily attributable to this joint orienta- tion. From Vesper Peak northward, the ridge closely follows the local bedrock strike as both curve gradually toward the NNW. The massive gneiss which forms much of the northern end of this ridge frequently is "blocked out" by the N 0 - 20 W and the dominant N 50 - 70 E joints; the resulting fragments readily separate from underlying bedrock on relict bedding planes or low-angle unloading fractures. 27 Overall Lemon Creek Drainage Basin Morphology The overall asymmetry of Lemon Creek valley appears to be most reasonably accounted for by the steep NNW dip of the joints trending N tpO to 700 E, and probably to a lesser extent to the steep NE dip of the bedrock. The tributary valleys entering from the south owe their indi- vidual asymmetry mainly to bedrock structural attitude. The general S 500 W trend of Lemon Creek approxi- mates that of the principal joint orientation cited above, but Lemon Creek's channel pattern nevertheless shows significant deviations from this general direction. These deviations coincide with the ends of massive ridges which impose a right-angle bend (from west to south) at the end of each ridge. Four such bends occur, and together with connecting east-to-west segments--usually located below main tributary junctions--constitute the primary channel pattern of Lemon Creek. The two major gradient irregulari- ties in Lemon Creek's longitudinal profile (e.g., below Canyon Creek and the Lemon Creek Glacier terminus; v. Figure 22) also occur in association with the ends of ridges formed by resistant strata. This strong lithologic control significantly disrupts the otherwise important influence of jointing on straight channel orientations of Lemon Creek: although the overall direction of the trunk stream is S 500 W, individual straight segments rarely follow this direction (Figure 5) and instead form an angu- lar pattern consisting of a number of approximately straight segments. 28 Erosional modification of Lemon Creek valley by an extensive and thick icesheet during the Wisconsin stage of glaciation did not produce a classic smoothly-curving longitudinal axis, truncated spur ends, or a simple, symmetrical U-shaped cross profile, but the effects of a former sinuous valley glacier along the length of Lemon Creek are nevertheless evidence, especially at lower levels of the main valley, above which hanging tributary valleys occur (v. map, Figure 5). A significant type of erosional feature which fur- ther indicates the influence of bedrock structure on valley shape consists of several large, NW-facing "facets" on the south valley walls above the south-trending reaches of Lemon Creek. These approximately planar cliffs strike between N 15 - 30 E, dip at angles of about 450 NW, and closely coincide in attitude with one of the subsidiary joint directions. Future Geomorphic Development Future geomorphic development of the Lemon Creek drainage basin probably will include: 1. Capture of the head of Lemon Creek Glacier by Ptarmigan Glacier, or its fluvial descendent, by headward erosion of the divide between Cairn Peak and Vesper Peak. 2. Enlargement of Ptarmigan Glacier at the expense of Lemon Creek Glacier (thickening of Ptarmigan Glacier will be partly damped down because of mass loss by overflow westward into Canyon Creek valley, ultimately 29 this could lead to a significant "self-induced" diversion into Canyon Creek valley). 3. A second significant capture of Lemon Creek Glacier by Ptarmigan Glacier (or their fluvial counter- parts) at the low point 0.8 km north of Vesper Peak. 4. Lemon Creek Glacier's eastern divide will probably be increasingly lowered and shifted westward as future headward erosion by steep-gradient glaciers at the head of the west branches of Norris Glacier proceeds. In the event of a future climatic shift favoring glacier expansion, the first glacier reactions in the Lemon Creek drainage basin will probably be exhibited by Thomas Glacier, a small, steep-gradient glacier having a substantial part of its surface above 1200 meters. An early response should also be expected at the relatively high extreme south end of Lemon Creek Glacier, which now terminates at the ridge between Cairn Peak and Observa- tion Peak (Figure 5). An expansion of Lemon Creek Glacier which is sufficient to spill ice over the divide into Ptarmigan valley would not only aid the advance of Ptarmigan Glacier, but probably would also be accompanied by the overflow of some Ptarmigan Glacier ice westward into Canyon Creek (both of these events occurred during the last major expansion of these two glaciers some two centuries ago. The headwaters of Canyon Creek flow from cirques which are empty at present, but under colder climatic 30 conditions cirque glaciers could again develop in these locations. Such conditions, if maintained sufficiently long, would lead to the breaching of the bedrock ridge between Canyon Creek and Ptarmigan valley, and the ridge between Ptarmigan Glacier and Lemon Creek Glacier. East- ward expansion of Canyon Creek at the expense of Ptarmigan valley and of Ptarmigan Glacier at the expense of Lemon Creek Glacier would be consistent with the present asym— metry of these valleys (Figure 5), which clearly reflects the steep northeastward bedrock dips and certain strongly supplementing joint surface orientations of this area. th CHAPTER IV JUNEAU ICEFIELD CIRQUE GLACIATION, WITH SPECIAL REFERENCE TO THE ANTLER AND GILKEY RIVER DRAINAGE BASINS Introduction and Objectives In view of the importance of cirque and cliff glaciers in regional landscape modification, their status as numerous and relatively simple models of glaciers as climatically sensitive geomorphic phenomena, and their possible relevance in amplifying the interpretation of changes in large valley glacier activity, certain aspects of Juneau Icefield cirque glaciation are examined and summarized in this section. The aerial photo and topo- graphic map analyses discussed below are intended (a) to assess the possibility that the morphologically simple cirque glaciers might reveal a clearer and more consistent record of climate-induced advance/retreat cycles than large valley glaciers, and (b) to provide a basis for extending the preliminary geomorphic observations made in the preceding section to the interpretation of termi- nal and proglacial zones of large valley glaciers to be discussed later in this report. 31 32 The first of these objectives is approached in a way intended (a) to detect specific patterns of present cirque and cliff glacier activity, particularly with respect to variations in present terminal elevations as a function of orientation and geographic location; (b) to delineate present trends and the former extent of these glaciers in selected parts of the study area; (c) to examine the relevance and applicability of currently observable or recently documented glacier distribution patterns to the interpretation of evidence of former activity; and (d) to compare the behavioral variability of cirque glaciers with that of valley glaciers in this region. The second main objective of this cirque analysis involves an empirical interpretation of existing glacier activity and the related interpretation of former acti- vity-~at both higher and lower levels of intensity than that of the present--as factors in regional landscape development. The selection of the west-central part of the Juneau Icefield (with special emphasis on the lower Antler and Gilkey valleys) for an evaluation of cirque glaciers was based on (a) the unusually large elevation range of this sector (on the order of 2000 m), (b) the presence of numerous variously developed and differently orientated cirque and cliff glaciers, and (c) the excellent coverage 33 of this area by vertical aerial photographs taken in 1948 and 1962. The relative importance of specific processes in the development of cirques is currently unresolved, and is not one of the objectives of the present study. Processes proposed for cirque formation include (a) headwall shat- tering by freeze-thaw action in association with bergschrund development and on exposures above the glacier limit, or in association with conditions at the boundary of subfreezing ice with 00 C isothermal ice, (b) abrasion by basal slid- ing, and (c) joint-block removal. Arguments and evidence for some of these processes are reviewed by Embleton and King (1968). The significance in various mountain areas of other basic factors related to cirque morphology such as (a) rock type, (b) rock structure, (c) preglacial morphology, and (d) duration of glaciation are also dis- cussed by Embleton and King (1968), and by Flint, (1957, 1971). Cirque orientation is generally attributed to a combination of (a) shading from insolation, (b) prevailing wind directions during the accumulation season, and (c) rock structure (v. discussion in Embleton and King, 1968). The mean elevation of cirque glaciers is usually regarded as being closely related to a mean level of the local firn line, with cirque floors a few hundred meters below this. A relationship between the two is noted by Embleton and King (1968, p. 198), who observe that ". . . 34 the firn-line on a cirque glacier usually occurs about three-fifths of the way between the snout and the upper limit of the ice." Fisher (1953), however, has suggested that because freeze-thaw activity is most intense near the 00 C boundary in cold glaciers, the prevailing long- term mean atmospheric temperature may account for the uniformity in elevation of cirque floors in some areas. Miller (1961) has conducted a distribution study of 218 abandoned cirques in a larger part of the Juneau Icefield region, but with different objectives than those of the present paper. Miller's study was focused on the vertical distribution of cirques as an approach to the delineation of former long-term average levels of regional névé-lines. In this he considered the elevation of the bedrock floors of abandoned cirques to approximate these significant mean elevations of late summer transient snow- lines or névé-lines over periods of time approximating Wisconsinan stage intervals (Miller, personal communica- tion). He reported the repeated occurrence in many valleys of five principal cirque systems characterized by closely accordant mean elevations (at 107, 336, 534, 748, and 961 meters, or 350, 1100, 1750, 2450, and 3150 ft, respectively), with currently-glaciated broad basins lying in this region at 1068 m or 3500 ft. The present study, in contrast, is based largely on other distributional and morphologic aspects of the currently active glaciers in the highest cirques and on upper valley walls. 35 Morphologic Effects of Bedrock Structural Elements The bedrock geology of the Antler—Gilkey drainage basin has not been studied in detail, but limited ground and aerial observations indicate that the structural trends and patterns mapped near Juneau continue northward into this area, with strongly folded and faulted metasedi- ments predominant to the west or in the coastal valley sector, with granodiorite more common in the highland area to the east. Field observations of bedrock structural elements in this large drainage basin are currently limited to joint orientation data from the Camp 18 nunatak near the eastern edge of the basin (Figure 23a), plus a few bedrock attitude measurements made along the lower Gilkey Glacier. In spite of this inadequacy, the general trends of the geomorphically significant linear structural ele- ments are apparent through their effects on components of the landscape, including the cirques. Examination of aerial photographs reveals an abun- dance of various structural elements such as faults, joints, and metasedimentary bedrock attitude trends which impose an unusual degree of control on straightness and orienta- tion of minor low-order stream channels, major proglacial stream valleys, and valley glaciers. Low-Order Stream Channel Orientations The strong effect of these structural elements is evident in Figure 25a, a plot of the flow direction (plotted to the nearest multiple of 100) and cumulative 36 lengths of 70 structure-controlled first- or second- order straight stream channel segments. These channels vary from 0.06 to 0.34 km (0.2 to 1.1 mi) in length, and have a mean length of 0.832 km (0.517 mi). The principal orientations are listed below: TABLE 2.--Principal low-order stream segment orientations in the Antler-Gilkey drainage basin (listed in order of decreasing frequency): (a) S to S 20 W (with a well-developed reciprocal trend); (b) S 30 - 60 W (with a weakly developed reciprocal trend); (c) S 80 W - W (with a nearly equal reciprocal trend at N 80 E to S 80 E); (d) N 40 - 60 W (with a very weak reciprocal trend); and (e) N 10 - 30 W (with an approximately equal recip- rocal trend). Comparison with the Camp 18 nunatak joint pattern (Figure 23d) reveals a broad similarity, but with several discrepancies of ca. 100 between principal orientations, probably reflecting the fact that much of the data in Figure 25a was derived from the western half of the drain- age basin (16 to 32 km from Camp 18), where the predominant bedrock types are different and where stream channels are nmch more abundant. 37 The dominant channel direction is toward the south- west quadrant, both in number of channel segments (45.3%), and in total channel length (49%). Glacial Valley Orientations The direction and cumulative lengths of straight segments of large valley glaciers, main tributary glaciers, deglaciated tributary valleys, and large proglacial river valleys in the overall basin are plotted in Figure 25b. A close relationship in channel orientation clearly exists between the low-order stream channels (Figure 25a) and the large valleys (Figure 25b), including the strong preference for certain southwesterly flow directions. As may be expected, the large main valleys display a greater tendency than small valleys to lead westward toward sea- level (66.8% of the large valleys, and 60.6% of the small channels trend westward). Cirque Orientations and Glaciation Intensity The orientation of 91 cirques located in the Antler-Gilkey drainage basin are plotted in Figure 250. These criques range from classic cirque forms with tarns to moderately-developed examples having steep headwalls, sidewalls, and distinct axial deepening. Unusually irre- gular cirques and vaguely defined cliff glaciers are not included. Figure 25c is subdivided to indicate (a) cirques containing active glaciers, (b) cirques containing greatly diminished glaciers or with evidence of recent glacier 38 disappearance, and (c) cirques containing no glaciers at present, and which lack definite evidence of glacier acti- vity in approximately the past century. In this figure, tarns are indicated when present. The orientation patterns of Figure 25c indicate a close genetic relationship with bedrock structural factors. This is assumed to be a result of the initial development of cirque glaciers in structurally-controlled low-order preglacial stream channels, or the result of topographic irregularities produced by differential weathering and mass wastage processes which have been guided by structural elements. The orientation patterns also differ in several significant ways from the two preceding plots (Figures 25a and 25b). All of the south-facing orientations are reduced in frequency, while north-facing reciprocal directions are conspicuous (for example, in Figure 25c, 67% of the cirques face northward). In addition, the northwest, northeast, and east-west directions commonly taken by small channels and large valleys are, in the case of cirques, offset several degrees toward the north. This shift of orienta— tion is assumed to result from maximum frost wedging on north-facing surfaces, tending to rotate cirques toward a north-facing direction. The distribution pattern of glaciation intensity for present and former levels of intensity reveals a sig- nificant trend when the three cirque categories plotted 39 in Figure 25c are compared (v. Appendix B). In terms of the percentage of east- and west-facing cirques, the three categories of glaciation intensities do not differ greatly from the 9l-cirque mean. A similar comparison of north- and south-facing cirques, however, reveals that as the intensity of glaciation decreases, the percentage of deglaciated cirques facing northward increases relative to those facing southward. A related observation is that as the elapsed time since the latest cirque deglaciation increases (to some extent a function of cirque elevation), the percentage of northward-facing cirques in this cate- gory also increases, reinforcing the empirical observa- tion that most of the well-developed cirques face northerly directions. It is also evident that a dispro- portionate number of the large, maturely-developed cirques with tarns are aligned in one of the dominant structure- induced orientations. If these observations are region— ally representative, it is clear that a collection of currently active cirque and cliff glaciers should reveal a lower percentage of north-facing glaciers than those plotted in Figure 25c. Such a collection (v. Figure 25d and Appendix B-2), although assembled for other objectives, includes only 55 northward—facing glaciers out of a total of 98. Cirque and Cliff Glacier Orientations The orientations of an assemblage of 98 active cirque and cliff glaciers shown on the Juneau C-2, C-3, -1 40 D-2 and D-3 quadrangles (not restricted to the Antler- Gilkey basin) are plotted in Figure 25d. These glaciers were selected for use in an analysis of present terminal elevations to be discussed in the succeeding section of this report. Valley glaciers or strongly channeled cirque glaciers, which may extend down to much lower ele- vations, have been excluded from this assemblage. Because of the inclusion of a large number of weakly channeled cliff glaciers, Figure 25d does not reveal significant agreement in orientation with the various features plotted in Figures 25a, b, and c. Instead, other orientations are more uniformly followed. The only conspicuous exception is a preference for the N 10 to 40 W arc, which represents the leeward direction with respect to prevailing storm winds. Other high-frequency directions stem from the location of many of the cliff glaciers on the upper walls of large valleys, and consequently are ultimately reflect- ing, at a 900 phase change, the integrated structural elements which initially guided the trend of these main valleys. Present and Former Terminal Positions of Glaciers in cirques and Tributary Valleys Effect of Cirque Orientation on Glacier Terminal Elevations To assess the variability in terminal elevation of existing cirque glaciers with respect to orientation, a total of 98 small, morphologically simple cirque and cliff o r o *— .,‘ lv 4 u- I.‘ I l. 41 glaciers were examined on the 100 ft (ca. 30 m) contour interval Juneau C-2, C-3, D-2 and D-3 1:63,360 sheets (Figures 7, 8, 10, and 11). Glaciers greatly elongated downvalley or occupying unusually sheltered positions were omitted. The lowest terminal elevation was estimated to the nearest 15 m (50 ft). For long, roughly horizontal irregular cliff glacier ice fronts, an upper and lower terminal limit was recorded. An orientation representing the general direction of cirque trend or the longitudinal axis of present ice flow was estimated to the nearest 50 for each glacier. These data are plotted in Figure 26a for each glacier. The strong effect of orientation is evident in Figure 26, especially when mean terminal elevations facing the north, south, east, and west quadrants are compared. The mean elevation difference between east- and west- facing glaciers is negligible, in that east-facing gla- ciers terminate only 4 m, or 13 ft, lower than west-facing glaciers. On the other hand, glaciers facing the north quadrant terminate on the average 236 m (774 ft) lgwer than those facing south. Terminal Elevation Variability at Selected Orientations The composite plot (Figure 26a) reveals that at any specific orientation, the terminal elevation of cirque or cliff glaciers is likely to lie within a vertical range of ca. 460 m (1500 ft). Both upper and lower limits are bud O. 42 generally defined for specific orientations. This verti- cal variation could doubtless be decreased if the selection were to be limited to more nearly identical glaciers, or to those with closely similar exposure factors. In view of this, Figure 26 could also be used in the selection of "representative" glaciers, although the value of such gla- ciers as representative phenomena would presumably change at higher or lower levels of glaciation intensity than that of the present. Effect of Geographic Location on Glacier Terminal Elevations The mean terminal elevations of all glaciers facing 900 sectors centered at the north, east, south, and west, are plotted in Figure 26b for the Juneau C-2, C-3, D-2, and D-3 topographic sheets. Although all four plots exhibit a predictable displacement due to the lower terminal elevations facing northward, additional large systematic differences in terminal elevation are also evident and appear to be related to geographic location. This systematic change occurs both in north-south and in east-west shifts. North-south variations are evident when the C-2 (southern map) and D-2 (adjacent northern map) data are compared. On the C-2 sheet, the south- facing glaciers terminate ca. 128 m (389 ft) lgwe£_than those of the D-2 sheet. North-facing glaciers, however, terminate ca. 33 m (109 ft) higher on the C-2 sheet. The same relationship holds when C-3 and D-3 data are compared: n" m... .o. ,.u .u 1... ...~. ~...§. n‘ 4 ... :4. .4... . .n» |\ .w‘ 43 south-facing glaciers terminate ca. 8 m (25 ft) lower on the C-3 sheet, but north-facing glaciers terminate ca. 55 m (167 ft) higher on this sheet than on the adjacent D-3 sheet. East-west variations involve an increase of termi- nal elevation in a west to east direction for east-facing glaciers, but a decrease in elevation (also in the west to east direction) for west-facing glaciers. It is evident that if systematic vertical displace— ments of modern glacier termini of up to 245 m (800 ft) are involved in the limited area of four 1:63,360 sheets, larger areas are liable to exhibit even greater amplitudes in modern terminal elevations. This imposes a significant complexity on the regional interpretation of present or former cirque glaciation in terms of elevation. Former Terminal Positions of Cirque and Cliff Glaciers At the 30 m (100 ft) contour interval of the 1:63,360 Juneau C-2, C—3, D-2, and D-3 topographic maps, moraines are rarely indicated, and consequently the mor- aine positions are based on examination of aerial photo- graphs. The August 1948 vertical photos (from which the U. S. G. S. topographic sheets were prepared) proved to be more suitable for viewing ground detail above treeline than the 1962 photography which was taken early in the ablation season. The elevation estimates for the lower parts of the end moraines are believed to be on the order 44 of i 30 m (100 ft), but may exceed that in unusually steep terrain conditions. The end moraine sequences in these cirques exhibit a tendency to cluster into three main groups based on relative differences in the degree of moraine preserva- tion, rock surface weathering, lichen cover, and overall vegetation cover. The youngest of these moraine groups consists typically of a distinct, steep-sided, fresh-appearing moraine commonly associated with a slightly younger mor— aine nearly parallel with and a few tens of meters behind the outer moraine. The close association and similarity in trend of the two moraines and younger minor ridges indicates that in the time period between their formation, no major realignment of the ice front occurred. The aerial photos generally reveal little or no difference in the tone of these moraines, although the distal side of the older moraine may appear darker as a result of lichen growth or a young vegetation cover. These characteris- tics suggest that the time interval involved may not have been more than a few decades, and that a major recession and subsequent readvance probably did not occur. The sub- stantial volume of these moraines, when considered in terms of the small dimensions of the glaciers involved, indicates that this latest advance position may have been maintained in approximately equilibrium for many decades prior to the current state of recession. These youngest 45 moraines characteristically lie ca. 38 m (125 ft) lower than the terminus for glaciers facing the north, and 46 m (150 ft) for those facing the south quadrant. Outer moraines, trimlines, and ground moraines of the intermediate advance generally are at least partially covered with vegetation, in many places including small conifers. Where bare rock and talus is visible, weather- ing and a lichen cover have darkened the surface, which contrasts sharply with adjacent recently deglaciated sur- faces. The intermediate advance limit generally lies between 100 and 225 m (325 to 740 ft) below that of the youngest moraine group. The earliest of the three main advances, in nearly all cases, probably occurred substantially before and was much more extensive than the intermediate advance, extending down to elevations ca. 45 to 210 m (150 to 700 ft) lower than those reached by the intermediate advance. Moraines recording the earliest advance are intermittent, subdued, and generally well vegetated. Below treeline, these moraines may be covered by mature conifers. An age of at least two to three centuries is provisionally pro- posed for this earliest advance, but significantly older dates cannot be ruled out, as this currently undifferentiated advance may later prove to be multiple, providing a sequence of more than three occurrences of significant Neoglacial cirque glacier expansion. V'- 46 Comparative Behavior of Small Valley Glaciers The small valley glaciers present in the Antler- Gilkey drainage basin have produced moraines which may be related to the climatic factors responsible for specific fluctuations of the small cirque glaciers. Because of flow-lag relationships, these fluctuations are presumably recorded later by the valley glaciers than by the shorter and steeper cirque glaciers. A major additional factor acting to increase the complexity of small valley glacier reactions can be attri- buted to the morphology of the accumulation zones of these glaciers. As developed throughout the Juneau Icefield, small valley glaciers characteristically originate by the coalescence of steep-gradient cliff (or cirque) gla- ciers arcuately flanking the upper end of the main valley. At times of high-elevation glacier growth, these upper glaciers spill onto the floor of the much deeper main valley, forming a low-elevation linear glacier tongue extremely susceptible to subsequent climatic changes which reduce or terminate the high-elevation ice supply. Under strongly negative mass budget conditions, such an ice tongue is likely to stagnate along much of its length after an initial phase of terminal recession. The moraine record of the cycle briefly outlined above is not likely to provide much detail regarding the later recessional conditions. Also, the floors of deep and narrow glacial troughs are generally the sites of 47 temporary accumulation of talus, colluvium, and alluvial fan deposits which obscure the valley wall and valley floor morainic deposits. The presence of a tarn or moraine-dammed lake in some cases serves as another complicating element tending to retard an advance and greatly accelerate recession by iceberg calving. Summary of Glacio-Geomorphic Observations Present Cirque Glacier Status Throughout the Juneau Icefield district, the cur- rent status of cirque glaciers relative to the morphology of their respective cirques is clearly one of disequili- brium. Inspection of the 1948 and 1962 aerial photos indicates that all of the cirque glaciers in the study area were in retreat at those times. In some instances, cirque glaciers have entirely disappeared since forming end moraines within the last one to two centuries. On the other hand, the well-documented post-1940 cooling trend, which has promoted glacier advances in other parts of Alaska and the Pacific Northwest, may similarly cause an expansion of favorably located and orientated high-level cirque glaciers in the Juneau Icefield--if it has not already done so. tn 48 General Fluctuation Pattern ofiCirque Glaciers For morphologically simple cirque glaciers, at least three significant main advances, with numerous minor fluctuations, are recorded. The oldest of these advances should be dendrochronologically datable, and in unusually favorable locations this may be possible for the two younger advances. Such data will greatly clarify the interpretation of regional trends, particularly in the case of the oldest moraines. These represent a record of the earliest response by glaciers to the climatic changes which caused the first regionally significant valley glacier advances of the 18th-century. Morphology of High-Elevation Cirques Many of the cirques still containing active gla- ciers display a step or convex rock shoulder on their headwall or upper sidewalls. This break in slope divides many formerly single cirque glaciers into a small lower glacier at the base of the step, and a second higher gla- cier (or snowbank) on the step and against the basal con- cavity of the main cirque headwall. In many instances, only the higher of the two glacier segments remains active at present. In some localities-~particularly where cirques face south, southwest, or west--active glaciers are now absent and only snowbanks remain. Crevasse pat- terns and changes in surface gradient of active glaciers-- 49 generally in northward-facing cirques--also reveal the presence of a stepped headwall beneath the ice. The elevation of the inflectionpoint at the top part of the step varies with cirque orientation. Char- acteristically for cirques in geographically limited areas, it is some 200 to 250 m (650 to 825 ft) lower for north-facing than for south-facing cirques. The rela- tionship between cirque orientation and step elevation suggests that a climatic factor may be involved in the step formation, and would implicate climatic conditions less favorable to glaciers than present conditions--par- ticularly in the instances where the step is currently under glacier ice. The formation of the step may be related provisionally to climatic conditions prevalent during the hypsithermal interval (ca. 8000-3500 B.P.), a period of higher snowlines, diminished glacier volumes, and redistributions of glaciers with respect to elevation. Valley Glacier Morphology and Fluctuational Record Inspection of the 1948 and 1962 aerial photography of this region reveals that in general the small glaciers which occupy well-developed deep glacial troughs have been in recession for at least several decades, indicating strong disequilibrium with the climatic conditions of the last century. Their moraine and present terminal eleva- tions are, as noted previously, much more variable than those of the simpler cirque glaciers. This variability ..-u A~ '.v.yi . "V" - - s...“ . In .1! '~ tn- 0 . I u. I ~ 1...' ”‘0.- . III‘ ...- hu$ q I l I]! "» I 4 I. n: I 50 obviously results in part from the greater differences in length, gradient, shape, ice thickness, and response times of valley glaciers. Another significant reason for dif- ferences arises from the complex morphology of the com— pound cirques and cliff glaciers from which most of the valley glacier mass is derived: simple valley glaciers without large mass-contributing sidewall or headwall cliff (or cirque) glaciers rarely exist in the Juneau Icefield. During a cycle of cliff glacier growth, once a substantial mass is poured by several cliff glaciers into a main valley, the focusing of ice flow promotes the rapid advance of a narrow tongue of ice. At times of glacier shrinkage, many of these contributory glaciers com- pletely withdraw from the low-elevation main valley floor. Although less favorable climatic conditions may permit the perpetuation of protected high-elevation glaciers, the lower-elevation valley glacier tongue acquires a catastro- phically negative mass budget under such conditions—-an instability factor which may lead not only to fontal recession, but to complete stagnation along much of its length. This phenomenon has repeatedly occurred on the periphery of the Juneau Icefield, where the upper ends of nwny deep U-valleys contain stagnating glacier tongues or are floored with morainic debris characteristically deposited under stagnant-ice conditions, while distinct morainic arcs formed by active ice are absent. In the case of large valley glaciers draining the icefield, such as the an". . "4"- Heidi 'Co. ‘uu ug, 51 Norris, Mendenhall, and Llewellyn Glaciers, which have broad, extensive accumulation zones and are not confined to the upper ends of deep canyons, this condition does not necessarily pertain, and arcuate patterns of multiple end moraines commonly occur. Effects of Tarns on Proglacial Drainage The presence of one or more tarns in a glaciated valley has a strong effect on the flow regime and on the erosional and depositional capability of the tarn's out- let stream. Retention of clastics in tarns results in minimal aggradation below the outlet, both on the valley floor and, significantly, at the mouth of the valley where alluvial fan development is minimal. These conditions also promote development of a stabilizing plant cover. On the other hand, cirques lacking tarns may contribute a sub- stantial sedimentary load to proglacial streams, with con- sequent landscape effects including alluvial fan growth at the valley mouth where gradients abruptly decrease. These changes in stream activity periodically inhibit and are recorded by the vegetation cover on valley floors and particularly on alluvial fans. In and below tarnless gla- ciated valleys, vegetation patches of different composi- tion and stage commonly occur, but are rare or absent below tarns. A close parallel to this situation can be found in the effects on down-valley outwash surfaces and vegetation cover by proglacial lakes associated with large valley glaciers. These effects are discussed below. CHAPTER V GILKEY GLACIER Location and Description Gilkey Glacier is a deeply entrenched valley glacier located about 55 km (34 mi) north of Juneau (v. Figures 1, 7, 9, 10, and 17). The length of this westward-flowing glacier is more than 34 km (21 mi). For nearly its full length, the mean glacier width in the main canyon remains at about 1.6 km (1.0 mi). Several long tributary glaciers join Gilkey Glacier, but with the exception of Vaughan Lewis Glacier, these tributaries do not contribute greatly to the main trunk ice stream, which remains dominated by Gilkey Glacier. In the Gilkey Glacier area, maximum local relief is on the order of 1830 m (6,000 ft), with a high percent- age of the bedrock exposures in the drainage basin con- sisting of steep slopes. The lower parts of these slopes, up to about 760 m (2,500 ft) or more, are in many places heavily forested, mainly by Sitka spruce (Picea sitchensis), western hemlock (Tsuga heterophylla), and mountain hemlock (Tsuga mertensiana). The forest cover becomes continuously 52 53 developed on the valley walls downstream (west) from the Gilkey terminus. In the upglacier direction, the vegeta- tion cover on the valley walls becomes increasingly inter— ndttent because of steeper slopes, higher elevations, and a longer duration of the seasonal snow cover. At the present time, Gilkey Glacier is rapidly receding as a result of several factors which have caused it to be in a strongly negative mass budget condition for most of the past two centuries, and particularly for the past several decades. In part, this condition is attri- butable to the general climatic amelioration of recent centuries to which all mountain glacier systems have responded. An additional factor making Gilkey Glacier's present status strongly negative is.its unusually deep entrenched valley. For most of its total length, the present glacier surface lies below the 1070 to 1220 m (3,500 to 4,000 ft) elevation range within which the sea- sonal firn line usually occurs under the present climatic conditions. The high-elevation areas draining into the Gilkey Glacier system are small in relation to the large, low-elevation area below the mean seasonal firn line. For most of its length, the glacier ice surface has been fur- ther lowered approximately 200 meters in the past two centuries. A related reason for the present unfavorable ratio of accumulation area to ablation area is that at many places around this entrenched system, the quantity of ice that spills into it from the adjacent main icefield 54 plateau on the south and east has been greatly reduced, particularly in recent decades. These particular climatic and geomorphic circumstances have created a situation in which a relatively large climatic change is necessary for a readvance, while relatively small opposite changes are sufficient to promote rapid and accelerated recession. Lateral Moraines and Trimlines Lateral moraines and trimlines along the valley walls above the present surface of Gilkey Glacier and its tributaries confirm that this glacier system participated in the regional pattern of Neoglacial thickening and advance. A series of fluctuations in ice thickness which accompanied the subsequent trend of thinning are recorded by a sequence of subsidiary trimlines below the uppermost trimline. Because of the absence of vegetation on some steep-gradient bedrock slopes, all of the trimlines are» discontinuous, but the composite pattern comprises three major trimlines widely separated in age, with three to five poorly developed intermediate trimlines. The vegeta- tion below the highest trimline consists mainly of older (Alnus), willow (Salix), and devils club (Oplopanax). In the uppermost part of the scour zone, young conifers are locally abundant. The trimlines further reveal that the present dominance of Gilkey Glacier with respect to its tributaries was also characteristic during its recent maximum and sub- sequent lesser advances. This is clearly indicated by the uy‘ “a on 55 trimline configuration at the confluence of Bucher and Gilkey Glaciers, as well as at the confluence of the large unnamed glacier east of Bucher Glacier and entering Gilkey valley from the north. In both of these cases, all of the Gilkey trimlines curve downward from the Gilkey valley into the mouth of the tributary valley. Maximum Neogla- cial ice levels in the two side valleys never equalled that of the Gilkey surface in the sectors where these glaciers joined. Terminal Zone On the basis of field work in 1958, Heusser and Marcus (1964b) reported that the recession of Gilkey Gla- cier began about 1783. Aerial photographs taken in 1929 show the ice front ca. 0.25 km (0.15 mi) behind the distal base of the largest end moraine. The arcuate shape and crevassing of the 1929 front suggests that near-equilibrium conditions and possibly a slight readvance may have occurred in the few decades prior to 1929. The lowest of the lateral trimlines probably correlates with this slight readvance. Additional evidence of a post-1783 readvance consists of a discontinuity in the till exposed in a stream-dissected cross section of the proximal part of the end moraine com- plex. This section is about 20 m (65 ft) in thickness, and includes a roughly horizontal zone of large, angular boulders which may represent ablation till deposited dur— ing recession from the 1783 (or earlier) advance limit. u.» 56 In comparison with the till below the boulder zone, the upper till contains a higher content of gray silt which may represent older proglacial lake deposits incorporated during this suggested readvance. Recession from 1929 to the position shown on the August 14, 1948 aerial photos was ca. 0.25 to 0.32 km (0.15 to 0.20 mi). The 1948 photography shows that the terminal zone had thinned greatly since 1929, thrust cre- vasses were weakly developed, and that most of this zone was approaching stagnation (v. photos SEA - 126 - 140, 141; SEA - 126 - 209, 210). The single proglacial channel currently in use was occupied in 1948, but probably had been initiated only a few years previously, as other channels appeared freshly abandoned at that time. Numerous pre-l948 pro- glacial drainage channels were cut through the terminal moraine complex and destroyed the entire distal portion of the outermost end moraine. This erosion was aided by drainage from a hanging tributary valley north of the Gilkey terminus (but subsequently diverted into a progla- cial lake). Because of this dissection, recessional moraines in the terminal zone are discontinuous. One major recessional moraine is represented by an irregular belt 30 to 60 m (100 to 200 ft) within the outermost mor- aine limit. The local relief of this inner hummocky belt is greater than that of the immediately adjacent surface, but less than that of remaining portions of the outermost II 1 a 57 moraine. Discontinuous till ridges closer to the ice front may represent perhaps a dozen minor oscillations, but these segments are difficult to trace except in the extreme proximal zone of the end moraine complex. The terminal condition of 1948 was clearly approach- ing the point at which a proglacial lake would begin to form, and accelerated recession would occur. The 1962 photography (e.g., EKX - 3 - 167, 168, and 169; see Figure 17) shows the middle part of the irre- gular ice front ca. 65 to 80 m (200 to 250 ft) behind the 1948 position. Proglacial lake development was well underway at this time but was restricted to the sides of the glacier, where long but narrow marginal lakes about 0.97 km (0.6 mi) were present. The initial appearance of the proglacial lake at the sides probably results from the lower movement rates and lesser thickness of marginal ice compared with the central part of the ice front. A mar— ginal lake position is further prOmoted by the fact that most englacial drainage leaves the glacier in marginal locations, possibly because of guidance by splaying crevasses in the terminal zone. Absorption of long wave radiation from the valley walls probably aids marginal ablation sub- stantially. Late in the 1964 ablation season, the last ice touching the central end moraine zone calved into the lake. The entire front has subsequently retreated rapidly by calving icebergs into a steadily lengthening proglacial I 58 lake. The other principal mode of recession, vertical downwastage, has thinned the lowermost several kilometers of Gilkey Glacier to the point of near-stagnancy. Under these conditions, continued rapid recession of the Gilkey terminus, amounting to at least several kilometers in the next decade, is to be expected. Proglacial Lake and Vallenyrain Relationships In contrast to the Antler River valley train sur- face to be discussed in the following section of this report, the Gilkey River valley train is covered by much less vegetation, most of which is currently restricted to abandoned channel beds. Comparison of the 1948 and 1962 aerial photos, along with the writer's own aerial obser- vations in 1964, 1965, and 1966 reveals that the vegetation cover has made substantial progress in this period of less than two decades. This is largely a result of a tendency of Gilkey River to occupy fewer channels, and to shift or modify these channels less frequently than in former years. In turn, this change in river behavior may be attributed to the presence of a sediment-retaining progla- cial lake in the past decade. The level of the Gilkey proglacial lake is con- trolled mainly by the gradient and threshold elevation of the outlet stream, which in turn is closely approximated by the gradient of the valley train surface. This surface is in turn affected by the presence of the lake, which acts as an effective sink for much of the silt and nearly .l 'q CI '44 pa o rt. 59 all of the sand-size and larger clastic particles, leading to greatly reduced aggradation and transport along the stream bed. An additional factor promoting outwash sur- face stability which is significant in the Gilkey system consists of large glacier marginal lakes which periodically fill and empty in the two tributary canyons near the Vaughan Lewis Glacier terminal zone, located 10 to 12 km upvalley from the Gilkey terminus. The discharge from these lakes was observed as a significant glacier burst or j6ku1hlaup on September 28, 1969, by M. M. Miller, who at that time also noted much erosion of end moraines as well as in the outlet channel at the south side of the Gilkey Glacier terminus (personal communication). With continued glacier surface lowering and with decreasing englacial discharge, the effect of these periodic discharges on the proglacial valley surface can only decrease. It is improbable that end moraines generally act significantly to establish or control proglacial lake levels, except in the case of local ponding of water at the ice margin. Even in the earliest stages of lake forma- tion, the lake outlet will normally be the previously established main proglacial drainage channel. The main proglacial stream channel generally may be regarded as either antecedent or contemporaneous with respect to moraine formation, rather than superposed across moraines. Instances of superposition across moraines probably occur, but the resulting gradient irregularity would be quickly removed, 60 with the proglacial lake level again controlled by the overall gradient of the outlet stream at the point of intersection with the glacier-excavated lake basin. Summary The dendrochronological and geomorphic evidence indicates a strong advance culminating in the mid to late 18th century, followed by recession and a second but slightly weaker advance. No evidence of a pre-l8th cen- turn advance has been found. Evidence of minor oscilla- tions is fragmentary, possibly because of destruction of moraines during glacier bursts or jokulhlaups. A few small fluctuations are recorded, however. The large number of well-rounded pebbles in the terminal moraines suggests the expected overriding of a former valley train surface. Of more significance is the presence in the end moraines of a large amount of fine- grained material which may have been derived from progla- cial lake deposits overridden by the 18th century advance. As noted earlier, the increased amount of fines in upper parts of the large inner moraine strongly suggests incor- poration of proglacial lake deposits by a later main advance. The presence of deeply weathered bedrock in contact with the present ice margin of Gilkey Glacier, several kilometers from the present terminus, corroborates the conclusion that the pre-18th century Gilkey Glacier was much thinner than it is at present, and implies that 61 the terminus at that time was a substantial distance, pro- bably many kilometers, upvalley from the early 20th cen- tury position. CHAPTER VI ANTLER GLACIER Location and Description Antler Glacier is located about 64 km (40 mi) NNW of Juneau, Alaska, and about 32 km (20 mi) east of Lynn Canal (v. Figures 1, 10, 11, and 18). This currently receding glacier originates as a westward-flowing distri- butary or "overflow" branch of Bucher Glacier. A large icefall occurs where the Bucher and Antler Glaciers separate. From the base of this icefall, Antler Glacier continues westward in a deep glacial trough about 3.2 km (2.0 mi) to an elevation of less than 30 m (100 ft), where it terminates in a proglacial lake. In common with other branches of the Gilkey Gla- cier system, Antler Glacier occupies only a small percent- age of the volume and a fraction of the length of a deeply- trenched, steep-walled canyon carved by substantially thicker ice during the Wisconsin Stage (and possibly earlier stages) of glaciation. Local relief resulting from this deep treching is on the order of 1500 m (5,000 ft). 62 63 Below the Antler Glacier icefall, the present elevation of the ice surface lies well below the current firnline, and in consequence summer ablation rates are high. A lowering of the upper Bucher Glacier accumula- tion area surface has greatly reduced the discharge of ice over the Antler Glacier icefall. This is well shown by a substantial decrease in icefall width on the 1948 and 1962 aerial photos. By 1966, the icefall had become further reduced to half its 1948 width. The great reduc- tion in discharge, particularly on the south side of the icefall, has also led to significant changes in medial moraine shapes and flow patterns (Hashimoto, et al., 1966). Map and Aerial Photo Coverage The Antler Glacier area is depicted on the 1:63,360 Juneau D-2 and D-3 quadrangles (Figures 10 and 11). Aerial photography was taken in 1929 for the U. S. Geological Survey, but is no longer available. Vertical photography which includes the Antler Glacier drainage basin and sur— rounding area was taken on July 5 and 11, and on August 14, 1948, for use by the U. 8. Geological Survey in topo- graphic map preparation (v. SEA - 125 - 100, 101, and 102; SEA - 126 - 207 and 208). Additional regional photography, but which includes only the Antler Glacier terminal zone, was taken on July 4, 1962, for the U. S. Forest Service (e.g., EKX - 2 - 35, 36, and 37; EKX - 3 - 177 and 178; see Figure 18). 64 Aerial Photo Interpretations Terminal Zone Examination of the 1948 and 1962 aerial photos reveal a sequence of at least seven distinct end moraines on the valley floor downstream from the proglacial lake outlet. The crest-to-crest spacing of these seven mor- aines, starting at the outermost moraine, is 65, 145, 48, 192, 210, and 80 meters. The outermost distinct moraine lies 0.76 km (0.47 mi) downstream from the lake outlet, 2.37 km (1.47 mi) from the 1948 ice front, and 3.57 km (2.22 mi) from the 1962 glacier margin. All seven of the moraines are narrow and seldom exceed more than a few meters in height. In relation to the other moraines, the outermost and the fifth moraine (450 m upvalley) are the largest. A dense vegetation cover on nearly all of the proglacial zone obscures several additional lesser moraine ridges. The present terminus ends in a proglacial lake which probably began to develop in the early decades of the 1900's. In 1948, this lake was more than 1.6 km (1.0 mi) in length and about 0.8 km (0.5 mi) in width. By 1962, additional recession of 1.1 to 1.3 km (0.7 to 0.8 mi) had occurred, mainly by iceberg calving. Between 1962 and 1966, about 0.5 to 0.6 km (0.3 to 0.4 mi) of recession took place. The depth of this lake is unknown, but the fact that large bergs calved from the terminus drift nearly to the lake outlet indicates that the depth 65 in large areas is on the order of at least 10 m (33 ft), and of course may be substantially greater. The 1948 and 1962 photographs indicate the possi- bility of a substantially older and more extensive advance position ca. 0.4 km (0.25 mi) beyond the outermost of the seven distinct moraines. The evidence for this more extended position consists of a subtle convex-downvalley arcuate pattern on the valley floor delineated by a small number of trees which are larger than any on the valley floor immediately upvalley or downvalley from this posi- tion. These trees may be rooted in an old, partially buried moraine system. Most of the valley floor in this area consists of abandoned proglacial drainage channels. Examination of the valley walls in a few favorable locations further upvalley reveals a narrow zone of trees above the highest distinct trimline which are taller than those still further above--a phenomenon which also occurs in other areas of the icefield periphery on which lateral noraines have been deposited in recent centuries. (An excellent example is found on the east slopes of the southern end of Norris Ridge, where trees on and below the ca. 1750 position of Taku Glacier are much larger than the older trees above and outside of the 1750 limit). valley Train Surface Characteristics Although the valley floor extending from the Antler Glacier end moraines to the Gilkey River was not examined on the ground, significant features of this area are 66 relevant to a general discussion of the effects of glaciation. From the outermost distinct glacier limit down- valley to its junction with the Gilkey River, the Antler River valley train surface consists of a complex network of anastamozing channels on which a dense but low vegeta- tion cover (generally not exceeding a few meters in height) has developed. Vegetation patterns on the valley floor closely reflect the pattern of abandoned stream channels. The only areas not densely vegetated are channel bars and point bars associated with the single active channel. Many valley train surfaces in other parts of the study area have vegetation covers in dis- tinctly different stages of development in comparison with the Antler Glacier area. Disequilibrium of the valley train surface, reflected by varied stages of vege- tation development, may result from a combination of sev- eral factors. These include changes in discharge related to cleatic trends, which in a currently-glaciated area involves not only the absolute quantities and seasonal distribution patterns, but also the amount of water impeded in its passage through the fluvial system by the maintenance of a snow or ice cover. A related climatic factor involves lateral shifting of broad snow divides in accumulation areas. Glacio-geomorphic factors such as glacier piracy and the widespread occurrence of ice 67 overflow through bedrock gaps also contribute to large variations in river discharge. The quantity and caliber of sediment load provided by glaciers to their proglacial streams is also subject to change as glaciers themselves are subject to wide behavioral ranges of advance, possible surges, equili- brium, and recession or stagnation. An additional significant element is the formation of a proglacial lake during the recessional phase, which greatly affects proglacial stream activity by retaining clastics too coarse to remain in suspension between the ice front and the lake outlet. The stage of vegetation development on the Antler valley train appears to be more advanced on the 1948 aerial photos than can be reasonably attributed to pro- glacial lake development in the early decades of the 1900's, and instead probably reflects a gradual decrease in discharge resulting from the long-term reduction of ice overflowing from the Bucher Glacier source. This would involve a minimal time base on the order of many decades to a few centuries. Proglacial lake formation in the early 1900's would also--but more abruptly--promote the stability of the valley train surface. For this, a time period of only one or two years might be involved. Most of the circumstances described above for the Antler-Bucher glacier system are of course essentially the same as those outlined for small, cirque-headed valley 68 glaciers in an earlier section of this report. However, the much greater dimensions and involvement of a greater number of ice sources obviously leads to substantial com- plexities in Antler Glacier's response to specific cli- matic factors. Field Results in the Terminal Moraine Zone The main objectives of the Antler Glacier field study were: 1. To establish the time of the maximum recent advance by dating the oldest trees at the trimline; 2. To determine, if possible, the dates of more recent fluctuations recorded below the main trimline; 3. To date the initial development of the pro- glacial lake, by establishing the date at which the innermost end moraines were formed at the lake outlet. These objectives were only partially achieved in the 3% days of field work. With regard to dating the maximum.recent advance, study of the aerial photos and observations from the air prior to the party's landing indicated that the terminal moraine position on the val- ley floor lacked a forest cover sufficiently mature to provide representative and reliable dates. Dating trees on the trimline also proved less satisfactory than anti- cipated, as distinct lateral moraines and trimline posi- tions could not be located at the time available. All I transects made near the terminus revealed instead a 69 gradational sequence of progressively older trees across the highest trimline. Lack of distinct lateral moraines and trimlines also precluded dating the more recent fluctuations, despite the fact that a sequence of at least seven small end moraine arcs is present on the valley floor below the foot of the lake. Some information was obtained, however, from several trees above and below the probable trimline position which were cored, using a short increment borer. On the south wall of Antler valley above the pro- glacial lake, five distinct trimlines are present, which may correlate with the five outermost end moraines. What appears to be a higher, rather indistinct, and presumably much older trimline is also present, which may relate either to the outermost distinct end moraine or to the earlier and more extensive possible ice limit described earlier. Upvalley from the terminus, lower and younger ice limits are marked by at least one kame terrace. Dendrochronology Dendrochronological data were collected from three localities near Antler Glacier. The first area examined was on the valley floor and extended from the proglacial lake outlet to the lowermost slopes of the north wall of the va-ley, and was restricted to a zone within 60 m (200 ft) of the lake. The second area was located on the lower slopes of the valley wall above the northwest end of the proglacial lake, and west of a stream which enters 70 the lake in this area. All of this area was below the highest former ice level. The third area was located farther to the northwest, and is mostly, if not entirely, above the highest ice limit of recent centuries. Because of the steepness of the valley walls in this last area, no lateral moraines have been preserved. Below the gen- eral trimline position, a few boulders are present, but many of these have probably fallen from the cliffs higher on the valley wall. Vegetationally, in this sector there is no obvious indication of a sharp demarcation marking the trimline in this part of the valley. In the first sector, ten trees were examined on the valley floor north of the lake outlet and within 60 m of the lake. The oldest trees generally were Sitka alders, six of which (based on annual ring count near the stem base) were 17, 19, 24, 24, 25, and 27 years of age. Three willows were 15, 17, and 19 years, and a single Sitka spruce was eight years of age. On the basis of the 27-year old alder, and allowing for a five—year establishment time and two years to grow to the cutting height, a minimum deglaciation date of ca. 1932 is obtained for the ground moraine-outwash area adjacent to the pro- glacial lake. In the second sampling area, i.e., on the lower valley wall, seven trees were cored. The oldest trees dated were a mountain hemlock (54 years by ring count to the center) and a Sitka spruce (51 years). For the older 71 of these trees, adding 10 years for establishment on this slope and seven years to reach the 30 cm (12 in) coring height, the minimum deglaciation date is ca. 1895. The presence of a few scattered boulders and a large, decom— posed stump a few meters above this point, but still well below the apparent highest ice level of the last two cen- turies, suggests that this is an important, though indis- tinct, trimline position. ' The ten-year period allocated for pioneer estab— lishment of these trees is considered minimal in View of the steepness of the valley wall at this locality; the actual period could well be on the order of several decades. Bray and Struik (1963), working much further south in British Columbia, found that ecesis periods on lateral moraines were from 23 to 43 years, versus about 20 to 26 years for terminal or recessional moraines. Four trees dated above the highest probable trim- line limit yielded ages of 222 to 399 years (Figure 27). These ages are minimal because none of the centers could be reached with the increment borer. The oldest tree, a mountain hemlock, was unusually difficult to core and nearly resulted in the loss of the increment borer. Only one core was obtained, but this probably closely approached the center, and the ring count age of 399 years may be only a few decades too low. With ten years added for growth to the 61 cm (24 in) coring height, and ten years for establishment, a minimal date of 1547 is indicated. o‘cuu. 72 Relatively rapid growth during this tree's early decades suggests the possibility that the terrain may have been deglaciated a few decades earlier--i.e., in the early 1500's, strengthening the possibility of a substantially older advance than is generally documented in the Juneau icefield region. Stratigraphic Evidence of Multiple Glaciation Evidence of possible multiple advance/retreat cycles of Antler Glacier was noted at the base of a steep granodiorite outcrop slightly above the second collection of trees discussed in the preceding section, and close to the suspected intermediate trimline position described earlier. Most of the sediment at this site consists of angular fragments, mostly feldspar crystals, which have fallen or washed down from the overhanging outcrop. Interruptions to this pattern include two horizons of gray silt which may have been deposited in glacier-mar- ginal lakes. These silt layers are underlain by thin organic horizons. A brief description of two trenched exposures spaced ca. 5 meters laterally apart are listed in Table 3 and Table 4. Provisionally, this deposit is interpreted as representing two advances of Antler Glacier separated by a time interval possibly on the order of one to three cen- turies in order to allow for the accumulation of weathered 73 TABLE 3.--Thickness and description of trenched section near Antler Glacier (Site AG-66-ex la). Thickness Description Variable 2.5 to 5.0 cm 1.3 to 2.5 cm 20.4 to 30.6 cm 0.7 to 4.0 cm 1.3 to 2.5 cm Thickness unknown; minimum 60 cm Surface debris: modern forest litter with a small amount of weathered rock debris (grus) from overhanging grano- diorite outcrop. Gray silt (probable deposition in a fosse—marginal lake). Dark zone of organic matter; includes some small undecomposed logs. Weathered rock debris, apparently from overhanging cliff. Gray silt. Organic horizon. Weathered rock debris, closely similar to that above the lower silt, but stained slightly brown. Base not exposed. TABLE 4.--Thickness and description of trenched section near Antler Glacier (Site AG-66-ex lb). Thickness Description Variable 2.5 to 5.0 cm 2.5 to 5.0 cm 30.0 i 10.0 cm 7.5 to 17.8 cm Thickness unknown Surface debris: modern forest litter, mosses, etc., or gray silt. Gray silt. Dark organic matter. Weathered rock debris, predominantly gray color. Lowermost 2 to 5 cm stained brown. Gray silt. Upper surface strongly stained brown-orange. Weathered rock debris, predominantly stained brown. Contains a few decom- posed pebbles. Base not exposed. 74 rock debris between the two silt horizons. The shallow depth of the organic horizons makes postdepositional con- tamination by modern rootlets appear likely, precluding reliable radiocarbon dating. Recommendations for Future Work l. The maximum recent advance date could probably be determined by locating ice-tilted trees on the trimline one or two km upvalley from the areas examined in 1966, particularly on the south wall of the valley above the proglacial lake and at the junction of the large tributary valley from the south. 2. The stratigraphic evidence of a substantially earlier major advance should be re-examined and a search should be made for uncontaminated organ material suitable for radiocarbon dating. 3. The valley floor and flanking slopes down- valley from the outermost distinct end moraine should be examined for tOpographic and stratigraphic evidences of the suspected pre-l700 advances. CHAPTER VII BUCHER GLACIER Location and Description Bucher Glacier (v. Juneau D—2 sheet, Figure 10) is the largest tributary to Gilkey Glacier, joining it at about 11.3 km (7 mi) above the Gilkey terminus, at an elevation of ca. 700 m (2300 ft). At this point, Bucher Glacier's ice discharge is much less than that of Gilkey Glacier, and consequently Bucher Glacier terminates close to the junction as a minor lateral part of Gilkey Glacier. Most of Bucher Glacier's accumulation area lies between 1220 and 1830 m (4000 and 6000 ft). The area- elevation relationships of Bucher Glacier--and other adjacent glacier systems on the west edge of the icefield—- are difficult to quantify because of the fact-that these are diffluent elements of a large icefield, with accumula- tion areas broadly connected by low-gradient ice divides. The Bucher Glacier accumulation basin periphery adjoins those of the upper Gilkey lying to the east, and of a large, unnamed westward-flowing valley glacier on the north and west. 75 76 As noted in the preceding section, part of Bucher Glacier discharges across a gap leading into a deep, west- ward-trending trough forming Antler Glacier. For at least the last several decades, and less continuously over the last one or two centuries, thinning of Bucher Glacier has led to a significant reduction in this overflow to Antler Glacier. Trimline Observations A sequence of at least seven distinct trimlines was noted on the valley walls above Bucher Glacier. On the basis of rock weathering and vegetation development, the highest of these trimlines appears to be distinctly older than the six (or more) lower trimlines. Because nearly all of Bucher Glacier is near or above treeline, opportuni- ties to date these past fluctuations are limited. Late l6th-Century Moraines The most favorable locality for dating the recent maximum expansion of Bucher Glacier was identified on aerial photos prior to the 1964 field season, and results from the existence of a glacial trough which formerly con- tained a distributary branch of Bucher Glacier (v. photos SEA - 125 - 103 and 104). Although this trough has not been completely pene- trated by the Bucher Glacier distributary branch for at least several centuries (and possibly for many thousands of years), its low floor gradient has promoted the pre- servation of moraine arcs produced by former thickening 77 and lateral extension of Bucher Glacier. Two of these arcs are clearly visible on the 1948 aerial photos. Much older arcs, if present, may lie under proglacial debris from a cliff glacier on the west wall of the valley. This material, and a large amount of talus, covers much of the valley floor. During a reconnaissance of the Gilkey Glacier sys- tem in 1964, this site was examined when much of the area was still under a snow cover. The outermost moraine is 4.6 to 6.1 m (15 to 20 ft) in height, and in places appears to bifurcate slightly. The distal side is dis- tinctly steeper than the proximal. The main inner moraine arc is slightly smaller, and lies 61 to 69 m (200 to 225 ft) behind the outermost. On the basis of vegetation development, this main inner moraine is substantially younger than the outermost moraine. Two minor moraines are present between the two main arcs. Between the inner large moraine and the present ice edge, one distinct and several discontinuous small moraine ridges are present. None of these moraines was breached by proglacial stream action, probably because of the limited quantity of meltwater at this elevation (790 m, or 2600 ft), the short distance of glacier penetration into this valley, and possibly because of rapid recession from these moraines. Several small mountain hemlocks on the outer mor- aine crest and adjacent slopes were photographed, but could not be dated, as increment borers or saws were not 78 available. This very significant site was revisited dur- ing the 1965 field season by a field party which included Dr. R. E. Beschel, who recognized, on the basis of crustose lichen diameters, the possibility that the outer moraine may have been formed 300 to 350 years earlier. This was confirmed by a disc from the largest of several mountain hemlocks growing on the moraine crest. This disc contained ca. 330 growth rings, but at least another 30 were estimated to have once been present in the decomposed central part of the disc (Beschel, personal communication). The growth rate of this tree (per decade) is plotted in Figure 27d. CHAPTER VIII NORRIS AND TAKU GLACIERS Location and Description The Antler and Gilkey Glaciers described in pre— ceding sections terminate in narrow confining canyons which closely restrict both the development of moraines and the character of proglacial deposits. The Norris and Taku Glaciers were also studied to provide a greater variety of morphology and its effect on glacier terminal phenomena. The present section is based on the more com- plex situation offered by the interconnected Norris-Taku Glacier system at the southern part of the icefield. Norris and Taku Glaciers terminate on the north side of Taku Inlet about 25 km southeast of Juneau (Figures 1, 4, and 19). Although terminating within 1.3 km of each other, these glaciers present a striking con- trast in that Norris Glacier has been in a condition of thinning and terminal recession since ca. 1916, while Taku Glacier has been vigorously advancing from ca. 1895 to the present time (Field, 1954; Miller, 1963). 79 80 Selection of the Norris-Taku terminal zone for field study was based largely on the complex variety of proglacial features at the receding Norris terminus, many of which are readily observable in various developmental stages. In addition, the close proximity of the currently- advancing Taku Glacier and its distributary branch, Hole- in-the-Wall Glacier, offers an unusual opportunity to observe and contrast the special character and activity of advancing ice fronts, adding substantially to the construc- tion of an overall valley glacier model. An additional reason for including these glaciers in this study arose from the fact that certain earlier workers (Lawrence, 1950; Muntz, 1953 and 1955) had not been able to reach agreement on several basic aspects of the past fluctuation patterns and interactions of these glaciers, and it was clearly necessary that these differ- ences be res-lved before either glacier could be included in any regional synthesis or model. Fundamentally, the present contrasting behavior of these glaciers originates in their great differences in morphology and dimensions, which lead to different reac- tions to the climatic trends of recent centuries. Norris Glacier's accumulation area consists of three short branches generally too low to accumulate sufficient snow to main- tain an equilibrium profile at present. On the other hand, Taku Glacier has several large main branches, at least two of which drain from some of the highest broad accumulation 81 zones of the icefield, where net accumulation rates are generally high under present climatic conditions (Andress, 1962; Egan, 1965; Miller, 1963). Tributary lengths also differ significantly between the two glaciers, with the main Taku branches generally three to five times as long as those of Norris Glacier, leading to a substantially greater lag in response to climatic change. These differ- ing morphologic parameters have been extensively discussed by Miller (1963). The differences briefly outlined above are sufficiently large to cause greatly contrasting termi- nal behavior. Additional factors acting to further com- plicate the individual behavior of these glaciers will be discussed in the succeeding sections. Early Historical Observations of Norris Glacier The earliest recorded observations specifically relating to Norris Glacier date from the late 1800's and indicate that a strong terminal advance was underway at that time. An undated photograph of Norris Glacier is included in F. F. Wright's account of a trip to southeast Alaska in 1886 (1889, opp. p. 29), and shows the ice sur- face in very close proximity with the highest trimline on both sides of the valley. Undated photographs by Winter and Pond, probably taken during or shortly following the first decade of the 1900's, reveal slight marginal thin— ning l to 2 kilometers upvalley since the 1880's, but it 82 is possible that the ice front continued to move forward during that interval. Early accounts also document the occurrence of large discharges of water from the terminus as a result of Norris Glacier's blockage of a side valley less than 2 km above the terminus of that time. (This valley is now occupied by Glory Lake, which is bedrock--dammed). For example, Reid (1901) observed in 1899 that this glacier, ". . . which ends on gravel-deposits in Taku inlet, is reported to have suffered the loss of a large part of its end, due apparently to the washing out of the supporting gravels." These outbreaks apparently occurred for several decades, as Field (1954) notes that "periodic outbreaks under the ice of a lake along the flank of the glacier above its terminus, which have been occurring at least since 1916, have kept much of the outwash fan so unstable that forest development has not advanced beyond pioneer stages except in a few small areas." D. B. Lawrence's (1950) report stated that ". . . in 1906 the Wrights found it advancing, and this culmi- nated between 1910 and 1916." Lawrence's report also includes photographs taken in 1916 and 1949 from Taku Point, showing that substantial thinning and recession had taken place in that period. Examination of the 1948 and 1962 aerial photography and the writer's own recent observations in subsequent seasons (1964 through 1968) 83 verify that Norris Glacier's terminus still continues the pattern of thinning and frontal recession it has maintained over the past half century. Interpretations of Pre-l900 Activity of Norris Glacier Pre-l900 fluctuations of Norris Glacier are poorly recorded. No end moraines are present beyond the 1916 ice front position, from which a broad pitted and channeled outwash fan extends 2.0 to 2.5 km to tide level. With the exception of two localities to be discussed shortly, this surface is contemporaneous with or postdates the latest ice advance. A second reason for the absence of a pre-l900 record is revealed by careful examination of the lateral margins on both sides of the valley, where it is apparent that the glacier height attained in the early 1900's not only exceeded and destroyed the trimline evidence of any 18th century advance, but exceeded any position reached for at least seven centuries prior to the latest advance (Muntz, 1955). Ironically, the main geomorphic circumstance responsible for the destruction of much of the proglacial record provides at least a limited amount of information on pre-l900 Norris Glacier activity. This circumstance-- the catastrophic outbreaks of large amounts of water from the glacier--blocked valley now occupied by Glory Lake-- is the essential factor to be examined before the character 84 of the terminal or proglacial deposits can be understood or interpreted. During the late 19th-century thickening of Norris Glacier's terminal zone, Glory Lake valley was entered a short distance by Norris Glacier, forming a large glacier- marginal lake with a surface ca. 60 to 90 meters above the present level of Glory Lake. Several small end moraines between Glory Lake and Norris Glacier indicate numerous glacier height fluctuations, as do a sequence of subsidiary lake strandlines above Glory Lake. The highest of these strandlines is conspicuously marked by an abrupt change from mature spruce, hemlock and cedar above the line, to older and young spruce below. Three less conspicuous strandlines are present in the zone between the present lake level and the highest strandline. In order to estimate the date by which Norris Glacier achieved its maximum thickness at the mouth of Glory Lake valley, the growth rate pattern of the largest tree growing immediately above the highest lake strand- line was examined by means of increment borer cores. This tree, a Sitka spruce 100 cm in diameter was located 6 meters above the highest strandline, and within 100 meters of the position of the ice dam. This location was definitely within the zone of katabatic air flow from Norris Glacier at the time of its latest thickening, and consequently the growth rate of this tree should reflect this latest advance, and possibly earlier advances, by a 85 reduction in growth rate. Unfortunately, the increment borer was incapable of reaching the tree's growth center, despite coring attempts from various directions (the geometric center was at 50 cm, while the longest core obtainable was only 27 cm). The oldest core contained 252 annual growth rings. Analysis of these rings revealed that growth of this tree was repressed from 1855 to 1890, and especially over the interval 1860 to 1885. The period of maximum repression was 1875-1878, which agrees closely with the early photographic record cited above, and sup- ports the assumption that this particular tree can provide an indication of glacier proximity. Further support for this interpretation is found in the absence of unusual repression from 1890 to 1966, a period during which photo- graphic records show predominant thinning of Norris Gla- cier at the mouth of Glory Lake valley. During the entire period recorded by this core (1714-1966), the only other repression is a slight reduction in growth from 1755 to 1765. This suggests that in the mid-18th century, Norris Glacier was not as thick or as capable of penetrating into Glory Lake valley as it was in the late 19th cen- tury. The generally high growth rates from 1767 to 1855 suggests further that the probable high ice level of the mid 1700's was not maintained, but instead decreased dur- ing the late 1700's and early 1800's. 86 Additional inferences can be drawn from the pre- sence of a large, decomposing stump located ca. 20 meters below the highest strandline above Glory Lake. Because of the decomposed condition of this stump, the annual growth rings could not be counted, but in view of its 1.5 meter diameter, an age of at least 300 to 400 years is inferred. Assuming that this stump represents a tree killed in the late 1800's (i.e., ca. 1875) by the flooding of Glory Lake valley caused by the advancing Norris Glacier, a period of some 300 to 400 years probably elapsed prior to 1875 during which the lower Norris Glacier could not have attained the thickness it achieved in the late 1800's. This conclusion is at variance with that reached by Muntz (1955), who maintained, largely on the basis of a lack of large stumps below the highest strandline, that Norris Glacier thickened in the 1700's and remained thick until the post-1916 recession. With regard to the 1910 l6-terminal advance of Norris Glacier, the evidence discussed above suggests that this advance may have been the greatest for a period of some three to four centuries. This conclusion does not conflict with Muntz's (1955) dating of trees older than 473 years immediately above the highest (i.e., late 1800-ear1y 1900) Norris trimline. A greater Norris Glacier advance in the 1875 to early 1900 interval than in the mid 1750's would of course account for the absence of a pre-1900 moraine 87 record at the Norris terminus. The effect of periodic high- energy discharges of the ice-dammed lake in Glory Lake valley over several decades would further reduce the pre- 1700 depositional record by its erosive action on nearly the entire proglacial zone. As cited above, only two main areas on the Norris outwash zone was pre-19l6 surfaces. These consist of (a) a small pitted outwash area adjacent to the 1910-16 end moraine and now out by Norris River, and (b) several larger remnant outwash surfaces adjacent to the tide flats, some 1.2 to 2.5 km southeast of the Norris terminal zone. The first of these areas is at least 90 years in age. This is based on a 77 year ring count of the oldest Sitka spruce cored in this locality, to which is added eight years for growth to the 39 cm coring height, and five years for establishment. The 80 and 5-year corrections are based on growth rate and ecesis estimates by Lawrence (1950). The survival of this surface from several decades prior to the 1910-16 maximum advance may be related to its location on the opposite side of the glacier front from Glory Lake valley, possibly giving it some degree of protection from the high discharges of water periodically released from that valley. The second group of remnant surfaces are located near the tide flats and consist of triangular areas bor- dered on all sides except toward the tide flats by abandoned drainage channels formed during or shortly 88 following the latest Norris Glacier advance cycle. These surfaces stand 1 to 3 meters higher than the adjacent younger outwash surface. The two surfaces contrast sharply on aerial photos (e.g., the 1948 SEA - 126 - 228, 229, 230, and the 1962 EKW - 6 - 127, 128, 129 photos), because the older areas are the only parts of the Norris outwash plain on which mature conifers (mainly Sitka spruce) occur at the present time. On the basis of tree- ring counts to their growth centers, many of these spruces approach ages of 160 to 180 years. Allowing five years for establishment of this first-generation forest, and 15 years for growth to the coring height, this surface is concluded to have become stabilized no later than 1765, or close to the date of possible growth of Norris Glacier (i.e., ca. 1755-1765), and to the 1755 date of Taku Gla- cier's major advance to Taku Point, as interpreted by Lawrence (1950). These older outwash surfaces are shal- lowly pitted in a few places, but generally display less relief than does the adjacent younger surface. Several shallow (l to 2 meter) excavations trenched along the margins of these older surfaces reveal that they consist of well—sorted water-deposited material, predominantly sand, and are not end moraines (v. Appendix C). The strong 19th-century divergence of Norris Gla- cier's terminal behavior from the regional pattern of dominant post-1800 shrinkage from 18th—century maximum positions suggests the effect of a major factor not present 89 in most other Juneau Icefield glacier systems. One such factor acting to complicate this glacier's mode of response consists of a large tributary valley ca. 7 km upglacier from the Norris Glacier terminus. This valley, now occu— pied by the "Dead Branch" of Norris Glacier (Figure 4), probably contained little or no ice at the onset of the 18th-century glacial expansion. Much of the ice of the expanding Norris Glacier was diverted into this large side valley as a distributary lobe, rather than as a tributary element, with the result that the advance of the lower Norris Glacier was substantially reduced. The subsequent more extensive advance of the late 1800's and early 1900's was then aided by the fact that the "Dead Branch" valley still contained a large amount of ice which reduced fur- ther inflow of Norris Glacier ice at that time. The deflection of medial moraine patterns clearly indicates the movement of Norris Glacier ice into this side valley. Thus, the advance culminating in the 1916 maximum could actually have been accomplished by a smaller climatic event than that involved in the earlier 18th-century buildup. Taku Glacier The advance of Taku Glacier has been documented in many sources (Field, 1954; Lawrence, 1950; Miller, 1963) and generally is interpreted as having begun between 1890 and 1895. Unlike the adjacent Norris Glacier, however, it has continued its advance to date. Advance rates have 9O varied somewhat over this period. Evaluations of the causal factors behind this advance have been presented by Lawrence (1950), Heusser, et_al. (1954), and Miller (1963, 1965). The Taku Glacier terminus was tidal from before 1890 until ca. 1930—40, when proglacial deposits began to predominate around the ice front. At present, the entire ice front at sea level is rimmed by prograding outwash sands and gravel and by finer sediments, some of which have been squeezed forward and upward by the increasingly thick ice resting on the bottom sediments of Taku Inlet. Miller (1963) has calculated the average net advance of Taku Glacier as 170 meters per year, and states that the advance was still proceeding at about the same rate in 1971 (personal communication). The observed forward advance is difficult to assess because of a large number of factors which may signifi- cantly affect advance rates. In addition to changes in net mass budget, and in the glacier's dynamic response to this budget, differential ablation occurs and important topographic effects are involved. The latter include the changing fiord depths and irregular valley wall configura- tion, which have been shown to have particularly strong effects on advance and retreat rates of tidal icefronts (Mercer, 1961). 91 Maximum 18th Century Advance Limit of Taku Glacier A major difficulty in delineating the past history of the Taku Glacier terminus stems from the fact that for most of the past several centuries this glacier has termi- nated in deep tidal water. Valley wall trimlines and lateral moraines, how- ever, comspicuously mark former marginal positions of this glacier, even though its former frontal position is not immediately evident from constructional or destructional features, with the exception of the known proximity of the mid-18th century ice front with Taku Point. The first significant ground study of the Norris- Taku terminal zone and the fluctuational pattern of these glaciers was carried out by D. B. Lawrence, W. O. Field and M. M. Miller in 1941, who collected tree-ring data from the Norris outwash apron. Lawrence later returned to this general area, focusing his dendrochronological study on the adjacent Taku Glacier. At Taku Point and on the east side of the glacier, Lawrence (1950) determined that the maximum recent advance of Taku Glacier occurred in 1755-1757. He also reported on the basis of multiple relief strandlines that a large part of the Taku River valley upvalley from Taku Point was the site of a large lake dammed by Taku Glacier, apparently during the 18th- century advance. Lawrence also maintained that Norris Glacier had strongly contributed to the Taku's extension westward down 92 Taku Inlet to a point some 3.5 kilometers from Taku Point. A. P. Muntz (1955), who collected tree-ring data from the lower Norris Glacier area in 1953, disagreed with several of Lawrence's interpretations. Muntz concluded that the 1910-16 Norris Glacier maximum advance was greater than any position reached in the 18th century, and consequently held that Lawrence's interpretation of an 18th century ice dam consisting of both the Taku and Norris Glaciers was incorrect. In addition, Muntz did not accept Lawrence's conclusion that the upper Taku River valley had been inundated in the mid-18th century as a result of the damming effect of Taku Glacier's advance at Taku Point. On the basis of field studies in this area in 1965, 1966, and 1967, the writer also considers it impossible for Norris Glacier to have advanced as far as Taku Point. However, with regard to Muntz's contention, the field evidence--in the form of several distinct strandlines-- clearly favors Lawrence's View that a large lake, or a succession of lakes, was impounded in Taku valley upstream from Taku Point. In the writer's View, the definite position of the 1755 Taku margin on Norris Ridge (v. Figures 4 and 19)-- only a few hundred meters from the present ice margin-- precludes an advance on the scale proposed by Lawrence, who regarded all of the Norris outwash area to have been under Taku (and Norris) ice at that time. Consideration 93 of the present surface gradient of the advancing Taku in relation to the 1755 margin makes a less extensive limit, only partially across the Norris outwash zone, appear much more probable. 'Another factor in this interpreta- tion is the erosive capability of water from the ice- dammed lake in Glory Lake valley, as discussed above, was probably capable of destroying Taku end moraines on the Norris outwash. However, one small remnant of such a moraine appears to have survived. This consists of a linear ridge of compacted till located in front of Norris Glacier, and lying between the 1910-16 limit and the present proglacial lake outlet. The till fabric of this feature is strongly orientated in a NE-SW direction, with long axes generally dipping slightly toward the NE (W. Savage, personal communication), in agreement with the interpretation of this feature as an end moraine of Taku Glacier. Its preservation is attributed to its having been subsequently covered and protected by the negligibly- eroding outer margin of Norris Glacier during its late- 1800 advance, preventing its destruction by glacier bursts from Glory Lake. CHAPTER IX DAVIDSON GLACIER Location and Description Davidson Glacier is a receding glacier located on the western side of upper Lynn Canal, about 16 km (10 miles) south of Haines, Alaska, and 104 km (65 miles) north-northeast of Juneau (Figures 2, 12, and 20). Davidson Glacier flows eastward from an extensive com- plex of tributaries lying in the Chilkat Range, east of the boundary of Glacier Bay National Monument. At sev- eral places, Davidson Glacier's accumulation area adjoins that of Casemen Glacier, which flows westward and south- ward to a receding terminus near Muir Inlet, Glacier Bay. The main connection between these glaciers consists of a braod divide at about 1265 meters (4150 ft). The total area of the Davidson Glacier drainage basin, including nunataks and bedrock slopes, is about 148 sq. km (ca. 57 sq. mi.). Davidson Glacier and its tributaries occupy about 115 sq. km (44.4 sq. mi), or about 78 percent of the overall drainage basin area. Four distinct main branches can be delineated. These range 94 95 from about 16 to 19 km (10 to 12 miles) in length, and vary from 1 km to over 3.2 km (0.6 to over 2 miles) in width. Each of these main branches is fed by numerous cirque glaciers, most of which lie between 1372 and 1676 meters (4500 to 5500 ft) in elevation. Only one large cliff glacier and a few small cirque glaciers exist as discrete units within the overall drainage basin. The highest point in the drainage basin is an unnamed nunatuk summit at 2085 meters (6841 ft). The present terminus lies less than 30 meters (100 ft) above sea level. The lowermost 8.5 km (5.3 miles) of Davidson Gla— cier flows through a steep, narrow canyon. In this dis— tance, the ice surface elevation drOps from 930 meters (3050 ft) to less than 10 meters (33 ft). The steepest surface gradient occurs at the lower end of the canyon, where the glacier drops 300 meters (1000 ft) in 1.6 km (1.0 mile), and narrows to 0.8 km (0.5 mile) in width. On emerging from this canyon Davidson Glacier forms an irregular piedmont lobe, ending in a proglacial lake. At present, the terminus is characterized by rapid thinning and marginal recession. The present ice margin lies more than 1.6 km (1.0 mile) behind a terminal moraine formed during the mid-18th century. Map Coverage and Aerial Photography The most detailed topographic map of the Davidson Glacier drainage area is the USGS Skagway A-2 quadrangle, at a scale of 1:63,360 (Figure 12). A small part of the 96 accumulation area is located on the adjointing Juneau D-5 sheet. Both of these maps were prepared from vertical aerial photographs taken in 1948, and employ a lOO-foot contour interval. Vertical aerial photography of the Davidson Gla— cier terminus is available for the following years: 1940 (U. S. Army Air Corps); 1948 (U. S. Navy); 1963 (U. 8. Forest Service); 1966 (Alaska Department of Highways). In addition, useful oblique aerial photography has been obtained by Dr. M. M. Miller (Michigan State Univer- sity) for the following years: 1948, 1951, 1958, 1964, 1967, and 1969. Late Neoglacial Activityfiof Davidson Glacier Early Descriptions Early references to Davidson Glacier were made by several individuals, including G. Davidson (1868), T. A. Blake (1868), T. Meehan (1883), and J. Muir (1893). One of the earliest descriptions of Davidson Gla- cier was given by G. F. Wright (1889), who traveled from Juneau to Chilkat by steamer in the summer of 1886. He states that along this section of the coast, ". . . nine- teen glaciers of large size are in full sight from the steamer's deck, but none of them come down far enough to break off into the water and give birth to icebergs. The 97 Davidson Glacier, however, comes down just to the water's edge, and has there built up an immense terminal moraine all along its front" (1889, p. 27). Wright's article includes an illustration of Davidson Glacier as seen from about two miles away. In this illustration (Wright's Figure 19), the glacier terminus is shown to be in a much thicker and expanded condition than it is in at the pre- sent time; a large piedmont lobe is depicted, but its limits are obscured by large trees growing on the terminal moraine. Wright's caption states that the terminal mor- aine is ". . . about two hundred and fifty feet high," but its actual height is much less than that. The magnitude of the error in Wright's estimate can be attributed to the large size of the trees growing on and in front of the terminal and outermost end moraines, and to the distance (ca. 1 to 2 miles) from which his observation was made. An additional possibility is that at that distance Wright may have mistaken large, drift-covered masses of stagnant ice for end moraines. Along much of its length, the terminal moraine is only about 2 or 3 meters (6 to 10 ft) in height. The highest end moraine crests near the termi- nal moraine are only about 15 meters (50 ft) above the proglacial flats. One of the first geologists to visit the terminus of Davidson Center was I. C. Russell, who compared the terminus to that of the Rhone Glacier, and described it as ". . . fan-shaped or semi-circular" (1890, p. 152). 98 Russell (1890, p. 152) also stated that ". . . the pre- sence of bare fields of debris about the extremities of many of the glaciers in the neighborhood of Lynn Canal, indicate that the ice streams of that region are receding. This is well illustrated by the bare and rugged piles of fine debris which encircle the expanded foot of the David- son Glacier." Recent Coastal Uplift in the Davidson Glacier Area Measurements of Recent Uplift.--Woodworth and Haight (1927, p. 73-76), of the U. S. Coast and Geodetic Survey, reported the measured uplift of several coastal benchmark locations both north and south of the Davidson terminus. Uplift data between 1890 and 1922 for the three locations nearest to Davidson Glacier are tabulated below: Measured uplift: (1890 to 1922) Haines (14.5 km (9 mi) N of Davidson Glacier) 0.77 m (2.53 ft) Pyramid Harbor (11.3 km (7 mi) NNW of Davidson Glacier) 0.73 m (2.37 ft) William Henry Bay (43.5 km (27 mi) SSE of Davidson Glacier) 0.73 m (2.37 ft) Measurements made in 1950 by the U. S. Coast and Geodetic Survey showed that coastal uplift at Haines amounted to 1.22 m (4.0 ft) between 1890 and 1950. 99 In 1952, W. S. Twenhofel compiled a substantial amount of coastal uplift data for Southeast Alaska. Twen- hofel (1952, p. 531-532) concluded that the measured uplift was not merely the result of the 1899 Yakutat earthquake, but that the upper Lynn Canal area was still rising. Pertinent sea level trend data compiled recently by Hicks and Shofnos (1965, p. 3317) are tabulated below: Period of meas. Sea level trend Skagway 1909-1959 «1.80 cm/yr Haines 1922—1959 -2.26 cm/yr William Henry Bay 1922-1959 -2.26 cm/yr According to the data cited above, the total uplift at Haines from 1890 to 1959 was 1.61 meters (5.38 ft). The close similarity between coastal uplift rates measured to the north of Davidson Glacier and those meas- ured to the south suggest that comparable uplift probably occurred at the intermediate location of Davidson Glacier. This 70-year record of rapid uplift is clearly a significant element in the recent geologic history of the Davidson Glacier terminus. Results of Recent Uplift.--Effects of the recent coastal uplift of the Davidson Glacier area include: 1. former bars and spits which now lie a few meters above the highest tide level. These 100 features are evident on the 1963 and 1966 vertical aerial photographs. 2. extensive deposits of beach shingle with marine shells, now located above the highest tide level. At several localities, a young first-generation forest has become estab— lished on these deposits. 3. the presence of a belt of progressively younger trees along the seaward periphery of the forested outwash area. Between the edge of this wooded area and the present highest storm beaches, only grasses and shrubs are present in any quantity. A few small isolated trees occur in this zone. Overridden Postglacial Gravels The stratigraphically lowest unconsolidated deposit exposed in the vicinity of the Davidson Glacier terminus is a well-stratified deltaic sand and gravel sequence located along the southern shore of the progla- cial lake (Figures 12 and 20). The exposure consists of a gravel bank a few hundred meters in length and up to 11 meters in height. The main circumstance which led to the exposure of this gravel was the presence of a glacier-marginal stream which until about 1965 flowed northeastward in a narrow gap between the glacier margin and the gravel deposit. Gradual thinning and retreat of the ice resulted in the 101 abandonment of this channel. When the area was first observed by the writer on August 7, 1966, this stream was no longer forced to flow against the gravel bank, but had migrated laterally with the retreating ice margin. By August 7, 1967, from 10 to 30 meters of additional marginal recession had occurred, and the stream channel along the base of the gravel bank had been abandoned. A "talus apron" of sand and gravel is not accumulating along the base of this exposure. During the latest advance/retreat cycle of David- son Glacier, these gravels were truncated by the overriding ice, which removed the tOpset beds and an unknown thickness of the foreset beds. The remaining gravel is not overlain by an unweathered gray till. Throughout the total observed thickness of 11 meters, the sands and gravels are conspicuously well- stratified. Along a lOO-meter segment of the exposure, the stratification of the foreset beds dips toward the northeast (generally N 60 E to N 70 E) at up to 240 to 300. Most of the particles are rounded to sub-rounded. The entire exposed thickness of these gravels is characterized by a rusty, reddish-brown staining. Many decomposed cobbles and pebbles are present. No marine shells or wood fragments have been found in the deposit. In the absence of material for radiocarbon dating, the age of this deposit remains unknown. The presence of decomposed pebbles and the pronounced staining of the entire 102 deposit suggests a duration of at least a few centuries prior to subsequent overriding by ice. These gravels may have accumulated during the Hypsithermal or Thermal Maxi- mum interval and be correlative with one of the gravel members of the Van Horn Formation described in Glacier Bay by Haselton (1966), but until datable material is recovered from this gravel exposure, this possibility remains unestablished. Main Buried Forest Horizon General Description of Exposure.--During the late 1950's or early 1960's, one of the two drainage channels leading from the Davidson Glacier proglacial zone was abandoned as a result of recession of glacier terminus. All proglacial drainage was diverted to the remaining channel, which subsequently exhumed a significant buried forest horizon. For a distance of about 800 meters (ca. 2,600 ft) along the present outlet of the Davidson Glacier progla- cial lake, a large number of glacially overridden logs and stumps protrude from the stream banks. This buried forest horizon generally lies between one to two meters above the normal summer water level of the stream, and is overlain by up to ten meters of gravel and till. At most places along the stream, slumping of the overlying till has obscured both the overridden horizon and the associated sediments. The only part of the buried forest horizon 103 which can be readily examined without extensive trenching lies along the south side of the stream immediately below the lake outlet. At this site, within an area of approxi- mately 20 by 120 meters (65 by 400 ft), part of the buried forest horizon has been exhumed by the stream to a level which very nearly coincides with the land surface on which the trees originally grew. A general View of the area (Figure 29), shows several dozen stumps, still standing in their original growth positions, and a large number of fallen logs, most of which are still embedded in the till. Most of the upright stumps are sheared off (v. Figure 30), and the mean direction of the attached splin- ters, measured on seven stumps, was found to be N 45 E. This is the approximate direction of the most recent ice movement over this area, but also is the same direction locally taken by the stream which uncovered the stumps. Although it is possible that some of this shearing may have been produced by advancing glacier ice, the writer attributes much of this "shearing" to the impact of large blocks of ice being swept downstream while the channel bed was being incised. While in the field, the writer observed several large ice blocks, up to 2 x 3 x 3 meters in size, being swept and tumbled down the proglacial lake outlet. In 1965 and 1966, several wood samples from this locality were collected for radiocarbon dating. These samples indicate that the ice advance which buried the logs 104 and stumps in this horizon occurred ca. A.D. 1210 - 1225 i 100 years (samples M - 1922 and M - 1924, respectively). Recession from this site may have occurred by ca. A. D. 1410 i 100 years (sample M - 1923), although alternate interpretations of this third sample are possible, as dis- cussed below. (All radiocarbon dates cited here were determined by the University of Michigan radiocarbon laboratory, with arrangements with the Foundation for Glacier and Environmental Research). Local Stratigraphic Relationships.--A sequence of eight stratigraphic sections (Figure 28; also see Figures 31 and 32) was measured on the south side of the Davidson Glacier proglacial lake outlet. The eight sections are located along a 200 meter (660 ft) reach of the outlet stream and lie in a nearly straight line which also is essentially parallel to the northeast direction of ice movement during the last glacial advance over this area. The top of each section ends at one of two main terrace levels which represent former beds of the outlet stream. Two of the sections were trenched to the level of the out- let stream. The remaining six sections were excavated to 0.5 to 1.0 meters above the water level of the stream. A description of each section is given in Appendices D-l through 8. Lower Till: A conspicuous overridden forest hori- zon is present in all eight sections, and generally is 105 underlain by a till which contains a greater proportion of pebbles and small boulders than the till overlying the forest horizon. Although a few angular particles are present in the lower till, nearly all of the particles in the pebble and larger size range are rounded or sub- rounded, strongly indicating a source in older valley train or outwash apron deposits. The lower till is weakly stained brown to orange-brown to a depth of about 60 cm. In many cases, staining is limited to the surfaces of resistant rock types. A few disintegrated mica schist fragments are present. Main Overridden Forest Horizon: On this till sur- face of undetermined age, a forest cover became estab- lished, only to be overridden by a glacier advance during the first decades of the 13th century. Radiocarbon dating of wood from the trunk of a small spruce buried during the ice advance indicates an age of 760 i 100 years B. P., or A. D. 1190; the normalized date is A. D. 1225 (Sample M- 1922). The outer rings of a nearby stump in growth posi- tion were dated at 880 i 100 years B. P., or A. D. 1070; the normalized date is A. D. 1210 (Sample M-l924). At the time of its destruction, this forest contained trees of substantial size and up to at least two centuries in age. The organic litter (compressed leaves, twigs, bark, and branches) on the former forest floor does not appear to have been greatly disturbed by the advancing ice. Many stumps are still in their growth positions, and the presence 106 of bark on stumps and many of the fallen logs indicates that they were killed either at, or slightly prior to, the time of burial. Several of the trenched sections include thin sand or silt strata between the forest floor material and the overlying till. These fluvial deposits seldom exceed 80 cm in thickness. The maximum extent of this main buried forest hori- zon in the direction toward the present ice front is not known, as no surface exposures of this horizon were noted west of the proglacial lake outlet. A subsurface extent of at least 200 meters west of the lake outlet is indi- cated by wood fragments encountered during drilling opera- tions conducted by an Alaska State Highway Department crew in 1967. These fragments were at a depth of 2.5 meters below the surface of the highest terrace of the proglacial lake. Dendrochronological Character of the Overridden Forest: The age of the forest prior to burial was, at a minimum, slightly more than two centuries. One of the largest overridden stumps was 200 i 2 years in age, by annual ring count; this stump was 2.29 meters (7 ft 6% in) in circumference at a height of 0.76 meters (2 ft 6 in). The largest overridden log observed at this site was 21.03 meters (69 ft 1 in) in length, and 2.24 meters (7 ft 4% in) in circumference near its larger end. Several meters of the upper and lower ends of this log were missing. 107 A conspicuous decrease in growth rates during a period of at least two to three decades prior to burial by the early 13th-century advance was observed on each of several stumps examined in the field. Examination of annual growth rings of sample DG65-l (C-l4 Sample M-l922; normal- ized date A. D. 1225), which was 81 i 3 years old at the time of burial, shows the onset of a five-year period of slow growth 47 years prior to burial, followed by a con- tinuous repression of growth 33 years before burial. The total thickness of the outermost 20 annual rings is only 3.0 millimeters. A second sample, DG66-3 (C-l4 Sample M-1924, normalized date A. D. 1210), was at least 131 years old (by annual ring count) at the time of burial. The absence of bark on this sample leads to some uncertainty as to whether the outermost observed ring represents the actual last-formed ring; a count along a different radius totaled 129 rings, and the possibility remains that a small number of the outermost rings are missing. The following periods of time prior to the death of the tree are therefore mini- mal. The earliest general decrease in growth rate occurred 99 years before death of the tree. The final period of repression began about 62 years before death; a total of only 21.5 mm of growth occurred during this time. This repression became increasingly severe as the ice approached, and in the last 20 years only 4.5 mm of wood was added. 108 A third sample consists of a complete section from a prostrate log (DG67 L-7) embedded in the till at the second stratigraphic section depicted in Figure 28. This log is 16.5 cm in diameter and contains 120 annual growth rings. These growth rings indicate that a sequence of increasingly severe repressions began at the following times before the death of the tree: 106 years, 81 years, 38 years, and 18 years. In the last 20 years of growth, only 4.0 mm of wood was added to the radius. Upper Till: The till overlying the main overridden forest level is light gray in color, and in all eight trenched sections examined it contains a higher proportion of silt than the underlying till. Much of this silt in the upper till may have been acquired as the ice advanced across proglacial lake deposits formed after the previous advance/retreat cycle. Summary Interpretations.and Recommendations.--The most complex and atypical of the eight sections is section number 6, in which three distinct horizons of organic material are present. The lowest of these levels consists of 1.0 to 3.0 cm of compressed leaves and twigs. No logs or stumps were noted along the narrowly trenched part of this horizon. This level is overlain by 2.5 cm of silt, 5 to 15 cm of sand, 33 to 43 cm of silt, and 8 to 13 cm of sand and fine gravel. These fluvial or lacustrine deposits are overlain by a second organic level about 1.0 to 3.0 cm 109 in thickness. This organic material also consists of matted and compressed leaves, needles, and twigs; a few small roots were noted beneath this level, but no large logs or stumps were present in this part of the section. This second thin organic horizon is overlain by 1.2 meters of predominantly finve-grained fluvial material, above which a conspicuous zone of plant fragments occurs. In this particular section, this highest organic level repre- sents part of a major forest level, in which small logs and roots are present. On the basis of physical appear- ance alone, this horizon most closely resembles the main overridden forest level described earlier. Tracing of this horizon back to section number 5 was not possible in the time available, because of the large volume of slumped material along the full length of the stream bank between these sections. The correlation of this uppermost forest level with the early 13th-century forest level previously des- cribed from sections 1 - 5 conflicts with the significantly younger radiocarbon age of the organic horizon lying 1.2 meters lower in this section. A flattened root from the intermediate organic level (v. Figure 28) was radiocarbon dated (Sample M-1923) at 560 i 100 years B. P., or A. D. 1390; the normalized date is A. D. 1410. Despite careful collection procedures, it is possible that this sample was contaminated with younger material. On the other hand, 110 if this radiocarbon date is approximately correct, the following provisional interpretations may be made: 1. The stratigraphically lowest organic horizon may represent the early 13th century level. 2. The stratigraphically highest organic horizon could represent the forest which was overridden during the ice advance which culminated in the mid-18th century. 3. The intermediate organic level may represent material which accumulated in a topographic depression following recession from the early 13th century advance. The presence of this intermediate organic level does not necessarily indicate a 15th-century ice advance, but its burial by silt, sand, and gravel does suggest the possibility of increased proglacial fluvial activity dur- ing the early 15th century. A large amount of additional trenching would be required to resolve the relationship of the three secion 6 organic levels to the single forest bed in sections 1 through 5. A series of shallow trenches to at least the depth of the upper organic level should be cut at close intervals between sections 5 and 6. Depending on which of the three organic levels actually correlates with the main overridden forest, at least one or two radiocarbon dates will still be required before a comprehensive evaluation of this section can be made. An additional recommendation for future work at this locality is that at least one complete section be 111 trenched from the present ground surface (deglaciated within the last five decades) downward to the 13th-century buried forest level. 'This may reveal a younger buried forest overridden during the late 17th--early 18th century maximum recent advance. If this possible forest horizon does exist, it could have important chronological value as a potential bridge between the oldest recently-cut progla- cial trees (i.e., ages of 360 to 425 years) and the actual year of ice advance at the trenched site. Dendrochronological Evidence of the Mid-18th Century Maximum Recent Advance and Subsequent Recession of of Davidson GIac1er Dendrochronological dating of the Davidson Glacier terminal fluctuations was complicated by logging operations that have been conducted on a large part of the progla- cial zone and on the three to five oldest end moraines. Within each logged tract, cutting dates of individual trees range over a period of several years, and in some instances, trees of significance to the present study were cut by homesteaders at least two to three decades prior to the 1966 and 1967 field work. On the proglacial zone in particular, it was not unusual to find, within a radius of a few tens of meters, stumps cut over a period of several decades. This range in cutting dates substantially reduced the usefulness of logging date information based on examination of aerial photographs, U. S. Forest Ser- vice records of timber sales, and interviews with logging 112 personnel in the field. Provisional estimates of elapsed time since cutting were made in the field for individual stumps, but were found to be subject to unacceptably large errors unless the cutting was done only one or two years prior to examination. An effective capability for determining the cut- ting date or final year of growth was provided by compar— ing annual growth curves of the undated trees with growth curves of trees either cut or cored at known dates. The assumption of detectable similarities in annual growth patterns of a geographically restricted population of trees, although supported by the Davidson Glacier dendro- chronological data, is in this instance further promoted by the restriction of the data to one species (Sitka spruce), and by the absence of strongly developed topo- graphic variations within the small area involved. Accord- ing to Lawrence (1950, p. 202), ". . . cross-dating between trees growing 10 to 25 miles apart is readily possible"; in the case of the Davidson Glacier end moraine and pro- glacial zone, only a few hundred meters separated the dated trees, and no difficulty was encountered in determin- ing the year of cutting by cross-dating logged trees with living trees. The Davidson Glacier terminal moraine was dated on the basis of four recently logged trees on the distal flank of the moraine. Each was located about one to two meters back from the distal base of the moraine, and at a 113 height of 0.3 to 1.5 meters above the adjacent proglacial zone. Their ages ranged from 241 to 259 years at the time of cutting (1962). All four trees had been severely repressed in the 1760's, but only two of the trees appear to have been tilted as a result of the maximum recent advance. Tree DGG6-S-7 was 249 years old at the time of cutting (1962), and provided the clearest evidence of having been tilted by ice. Recovery from the tilting began at about 1752. The second ice-tilted tree, DG66-S-l4, was located 10 meters north of tree S-7. S-14 was 259 years old when out (1962). Its growth rate began to decrease in 1751, and was extremely low from about 1755 to 1762, after which response to tilting began. The disagreement in the dates of response of these two trees (1752 and 1762, respectively) is probably due to (a) differences of a few years in the actual date of the tilting, (b) differences in the amount of tilting and earliest date of recovery, and (c) difficulty of estimat- ing the exact year in which recovery began. This last factor is probably responsible for an error on the order of three years or more, and stems from the fact that the change from pre-tilting "normal" growth to post-tilting asymmetrical growth may be transitional, occurring over an interval of several years. 114 The two untilted trees, DGG6-S-15 and -16, were 248 and 241 years in age, respectively. Both were located on the distal flank of the terminal moraine, at a point about 100 meters south of the two ice-tilted trees discussed above. S-15 was located about 2.0 meters back from the distal base, and about 1.5 meters up the distal slope. S-16 was located about 15 meters south of S-lS, and was 1.0 meter back from the base and 0.3 meter up the distal slope. The growth patterns of these two trees did not reveal any evidence of tilting by ice, but both trees had experienced three distinct growth repressions, as tabulated below: 0666-8-15 DG66-S-l6 1743-1750 1742-1749 Severe repression, particur larly severe from 1747-1751. 1767-1775 1762-1771 Repression. 1783-1788 1783-1793 Repression. The severe repression from 1747 to 1751 closely agrees with the most probable date of maximum recent advance indicated by the two ice-pushed trees discussed earlier in this section, i.e., 1752. The discrepancy of l to 5 years may reflect a slight difference in the time of the maximum recent advance at different parts of the piedmont ice front. The two repressions shown by the growth rings of trees S-15 and S-16 (1762-1775 and 1783- 115 1793) suggest that marginal fluctuations or minor read- vances did occur. Further evidence of this is found at several points along the terminal moraine where it is locally overlapped by an adjacent younger and larger end moraine. CHAPTER X GENERAL MODEL OF JUNEAU ICEFIELD VALLEY GLACIER ACTIVITY Equilibrium State The concept of equilibrium in a valley glacier involves a state of dynamic balance between a wide range of internal and external factors. In the series-connected processes outlined by Meier (1965) and cited at the out— set of this paper, external factors begin with the general meteorologic environment and local mass and energy exchange, which constitute a complex blend of quantities and dis- tributions of these quantities subject to substantial change over long as well as short time periods. In addi- tion to these atmospheric changes, the uplift of tectoni- cally active mountain areas imposes an additional slow shift in local climatic background, leading to tempera- ture decrease and, generally, to a precipitation increase. The presence of the mountains themselves further promotes drastic differences on a local, microclimatic scale. Not unexpectedly, glacier net mass balances showing wide year-to-year differences are abundantly documented. 116 117 The dynamic response by glaciers to changing mass balance quantities is conditioned not only by the complex reactions of ice itself, but also by variable glacier morphology and dimensions conditioned by the configuration of the bedrock surface on which the ice flows. The final activity of the terminus is never a simple response, and in view of the capability for change exhibited by the variables involved, long-term terminal stability is not a likely event. On the basis of the directly observed and indirectly interpreted fluctuational history of Juneau Icefield glaciers, terminal stability in the sense of an equilibrium condition has in fact seldom occurred. Of the glaciers discussed above, and strictly in terms of terminal position, Gilkey Glacier probably main- tained a slightly-changing terminal position for the longest period of time, i.e. from about the mid-1700's to the early 1900's. In part this is probably attributable to the generally low and nearly uniform gradient of the valley floor and to its nearly constant width in the terminal zone, which would encourage relatively uniform movement rates near the terminus. The apparent dominance of the single main glacier during the past several cen- turies, with limited major contributions by tributary glaciers, should have further encouraged this relatively simple pattern by minimizing higher (and more variable) velocities caused by increased ice thickness added by tributaries. The substantial length of this glacier (ca. 118 25 km from the mean seasonal firnline to the terminus) should also act to diffuse large positive changes in net accumulation at upper levels, and to produce a long response time. When viewed as a whole, however, Gilkey Glacier's terminal position stability loses much of its significance in comparison with the very large quantity of ice lost by overall glacier surface lowering and indi- cated by the high level of the 18th-century trimline. Although generally losing mass for nearly two centuries, negligible terminal recession occurred until recent decades, when glacier thinning led to velocity reduction to the point of near-stagnation. The current high frontal recession rate is greatly aided by the presence of a deep proglacial lake. In the event that climatic stability occurred over an interval sufficiently long to permit long—term mainten- ance of an equilibrium glacier profile, the subglacial erosive action continuously carried out upglacier would gradually lower the glacier bed and decrease the floor gradient. The first of these effects would lead to an increase in ablation rates at the new lower elevations, and would also increase the total area of the ablation zone. The second would reinforce the first because of the great effect of slight changes in slope on basal shear stress and flow velocity. Subglacial and proglacial deposition near the terminal zone would further lessen the glacier gradient and, consequently, its velocity. 119 Advancing Condition The expansion of morphologically simple glaciers, such as Ptarmigan Glacier (Figure 5), in response to climatic change may occur as a slow "normal" expansion or as a more rapid surge. Glacier surges have attracted con- siderable attention (Tarr and Martin, 1914; Miller, 1958; Post, 1960; 1969; Meier and Post, 1969; Bayrock, 1967). but these phenomena are not yet well understood. Glaciers in either a slowly advancing or a rapidly surging condi- tion would be expected to erode at greater rates, and to provide their outlet streams with rock fragments at greater rates, than they are capable of in a non-advancing condition. Detailed differences in the character of glacier-eroded and transported fragments caused by differ- ences in glacier behavior were not investigated for this study. If increasing ice velocity is largely the result of an increasing proportion of basal sliding with respect to internal flow, an increase in the proportion of fine grain sizes may occur in the subsequently-formed end moraines. This process would act at variance with the frequently-cited increase in the proportions of small particle sizes found in successively glggp till sheets deposited by Pleistocene continental glaciers in the con- terminus U. S., a phenomenon usually attributed to the advanced stage of weathering of the initially-glaciated surface. 120 The quantity of material deposited in moraines at the culmination of an advance cycle may depend on a com- bination of several factors, as listed below: (a) the duration of the maximum glacier extension, (b) the number of frontal advance/retreat oscillations, (c) high rates of forward movement of debris- transporting, (d) the quantity of entrained or superglacial debris, and (e) squeezing of saturated material forward and upward to the ice margin by the increasing downward pressure of thicker ice upvalley (this is facilitated under temperate glaciers by the presence of abundant water, and par- ticularly if clays or silts are abundant in the overridden material). It should be noted that in this latter case, these "mor- aines" may not consist of ice-transported material, but could be formed largely of lacustrine or marine sediments-- as is now occurring along much of the advancing Taku and Hole-in-the-Wall ice fronts. In some cases, the presence of a previously-formed moraine may retard forward flow sufficiently to cause additions of material on itself. This would be expected to occur most readily in front of small advancing cliff and cirque glaciers, which would be less capable of 121 advancing over such an obstacle. This may account for the frequent occurrence in many mountain areas of large, massive end moraines in front of small cirque and cliff glaciers. However, it is also possible that a particle size effect is at least partially responsible for the large moraines commonly amassed by small, steep-gradient glaciers. These glaciers generally are closely restricted in or near the zone of maximum freeze/thaw activity, and receive a high proportion of large and angular frost- shattered blocks. Short transport distances inhibit the production of finely-ground material at the glacier bed. The composition of the end moraines of small glaciers therefore should be richer in coarse material and poorer in fines than that of large valley glacier moraines. This implies the possibility of a correlation between valley glacier length and the proportion of fine particle sizes in the end moraines. An additional reason for a relatively small pro- portion of fine particle sizes in cirque glacier moraines may be related to the inability of their low-discharge proglacial streams to erode and transport the large par- ticle sizes predominant in these moraines. Removal of fines from the moraines would be further aided by the weak development of vegetation on these high-elevation morainic surfaces. Downvalley from an advancing glacier, the main effect usually cited in the literature is an increased 122 tendency for aggradation. This may generally occur, but in some instances at least part of the aggradation may be related to the increasing proximity or encroachment of the water-discharging ice front. Deposition of water- transported particles in the immediate vicinity of an ice front may be due to a change in channel shape (from deep and narrow to wide and flat), or to reduction of gradient, as streams leave glacier fronts (Embleton and King, 1968). The only currently-advancing Juneau Icefield glacier, Taku Glacier, is actively constructing a locally pitted out— wash fan along much of its periphery. This outwash zone is prograding onto that of Norris Glacier, burying trees and forcing Norris Glacier's outlet stream westward (v. Figures 4, 19). The extent of a glacier's maximum advance may be strongly controlled by factors other than climatic change or dynamic response characteristics. In the case of Norris Glacier, the presence of a large side valley appears to have acted to damp down and minimize the ultimate extent of an initial advance cycle by diverting ice from the main ice stream. This effect should be decreased in subsequent advance cycles, provided that a substantial amount of ice still remained in the side valley. The availability of such side valleys is great on the Juneau Icefield periphery, where present glacier dimensions are greatly reduced with respect to the dimensions 123 of those which have strongly modified the valleys in the past. Recessional Conditions The recession of a valley glacier terminus gener- ally occurs at an increasing rate as upvalley ice thick- ness, surface gradient, and velocity all decrease. Major differences in recession modes may occur, however, depend- ing on the topographic and subglacial character of the terminal zone, and the quantity and type of entrained debris. Fundamentally, valley glaciers may end in one or more of four major types of terminal situations: (a) at tide level, in contact with deep or shallow water; (b) on a broad, unconfining, low-gradient valley floor usually found at a canyon mouth; (c) on a valley train surface flooring a confining glacial trough; (d) on a bedrock valley floor. The recession of individual glaciers may, of course, occur according to any of these different modes in the course of a recession. Generally, some or all of the four main types of recession would be followed in the sequence listed. 124 Glacier Termination at Tide Level In recent centuries, the only Juneau Icefield glacier to retreat with a tidal icefront has been Taku Glacier. The special character of tidal icefronts has been discussed by Tarr (1897, 1909); Carey and Ahmed (1961); and Mercer (1961). Glacier Termination at Canyon Mouths Terminal recession on low-gradient surfaces at canyon mouths is best exhibited in the Juneau Icefield region by the East and West Twin Glaciers (Figure 16) and by Norris Glacier (Figures 4, 19). Other excellent examples are Davidson Glacier (Figure 12, 20), and Wright Glacier (Appendix F, and Figures 13, 14, 15, 16, and 21), draining from separate icefields lying respectively to the northwest and southeast of the Juneau Icefield. The lobate spreading of ice at canyon mouths greatly decreases thickness and flow velocity, and increases the ablation area. This results in a tendency for such glaciers to terminate shortly after leaving their canyons, and also results in a tendency for suc- cessively greater advances to require diprOportionately larger ice discharge rates. Rapid reduction in velocity toward the glacier margins, however, leads to strong sub- glacial depositional tendencies, as revealed by the varied depositional record of Davidson Glacier. 125 End and ground moraine morphology appears to change systematically as recession proceeds, beginning with a relatively simple outer moraine and leading to progres- sively more crenulated morainic patterns. After a very short initial recession, belts of small glacier-marginal ponds become abundant in an outer zone ca. 300 to 500 meters in width. The depressions occupied by these ponds usually are formed by the recession of ice locally thicker because of irregularity of the glacier subsurface. In a zone from approximately 500 to 1900 meters behind the advance limit, continued recession generally leads to a further increase in local marginal irregularity and to the development of larger glacier-marginal ponds. These are characteristically elongated in the direction of ice movement, and with a triangular shape pointing away from the glacier. Subglacial erosion of unconsolidated sediments at canyon mouths is often substantial (e.g., Twin Glacier Lake in Taku Valley is at least 135 meters in depth (Lawrence, 1950), and the Norris Glacier proglacial lake was at least 30 meters deep when measured in only a few places by the writer in 1967). These excavated basins characteristically become the sites of proglacial lakes after a total recession of some 400 to 600 meters from the advance limit. With initiation of a proglacial lake, a major transformation in recession rates and in proglacial stream 126 characteristics occurs, as discussed above for the Antler, Gilkey, and Norris Glaciers. The effectiveness of iceberg calving as an ablation process has been widely documented by Tarr (1897, 1909), Field (1947), Mercer (1961), and Miller (1964), and is largely responsible for the high recession rates in recent decades of many of the valley glaciers associated with the Juneau Icefield and neighbor- ing icefields. The great effectiveness of ablation in proglacial lakes may also act significantly as a means of obscuring or even preventing subsequent advances, although such advances could be detected and documented along the glacier margin farther upvalley. Glaciep Termination on Valley Train Survaces Glaciers terminating on valley train surfaces and confined in narrow canyons generally exhibit recessional behavior similar to that outlined above, although there are some differences. One such difference is the usual absence of arcuate belts of ponds formed along the ice front at certain stages of development. Glacier Termination on Bedrock This mode of recession may be regarded, in the case of most valley glaciers, as generally following one - or more of the other recessional modes. Large valley glaciers in the Juneau Icefield area do not generally terminate in this manner at present, although several have now receded to the point at which a transition to this 127 type of recession is occurring (e.g., Davidson, Eagle, Herbert, Mendenhall, Wright, and Llewellyn Glaciers). CHAPTER XI CONCLUSIONS The preceding discussion of the varied factors in glacier response leads to the conclusion that the number and complexity of the factors affecting individual gla- ciers virtually preclude the establishment of direct time correlations between glaciers in terms of climatic events, except in the case of nearly identical glaciers. The great configurational variety of large valley glacier systems sufficient complicates their individual responses to a given climatic change that correlations between them are only possible on a long-period or pattern basis. Regionally, most Juneau Icefield valley glaciers did expand and advance to recent maximum positions culmi- nating in the 18th century, indicating a previous climatic event of major significance (v. Appendix E). However, for individual valley glaciers of the Juneau Icefield, the maximum recent advance has been shown to range from ca. 1590-1600 (Bucher Glacier) to about 1910-l6 (Norris Gla- cier), and probably does not reflect the same past climatic event. Thus the concept of a "maximum recent advance" 128 129 must be viewed as not necessarily a reflection of a single major climatic oscillation, but as an overall growth of entire glacier systems, which may, however, be ambigu- ously represented by terminal response patterns. Optimal glaciers for climatic interpretations should be selected from a collection of small cirque or cliff glaciers which, in effect, "integrate" fewer and less variable factors than large glaciers. In addition, the great number and variety of these entities in and around mountain areas provide a greater opportunity for the selection of indi- vidual glaciers which may be expected to reveal a record of special sensitivity to particular environmental changes, such as temperature change, precipitation change, shifts in dominant storm tracks, etc. Cirque glaciers, however, may not necessarily provide a long historical record in comparison with large valley glaciers which terminate at low elevations, where glacial and fluvial depositional pro- cesses may predominate over long time intervals. Under the climatic conditions prevailing for the past two centuries, and particularly for the first half of this century, mass decrease and recession (at varying rates) has been characteristic of nearly all glaciers of the Juneau Icefield. Glaciers which are currently in rapid recession are, as has been shown in earlier sections of this paper, often doing so not only in response to unknown climatic changes many decades or centuries in the past, but also in response to purely non-climatic factors. 130 Small but deeply entrenched valley glaciers nourished by cliff glaciers (described in an earlier section of this paper) illustrate a behavior pattern which is sensitive to mass balance changes during an advance cycle and per- haps shortly afterward, but which loses this sensitivity in a recessional phase by nearly or completely stagnating. Glaciers which are in a condition of pronounced recession (Gilkey Glacier provides an extreme example) could be expected to retreat for many decades even if their upper ends were gaining mass. An additional negative factor arises from the progressive lowering of the glacier sur— face to elevations of decreased accumulation and increased ablation, necessitating increasingly large positive budget changes for an advance. The specific glacier characteristics which should enhance sensitivity to environmental factors such as temperature or snowfall increase or decrease are difficult to make without appropriate meteorological data from a mosaic of high-altitude ground sites. As a general guide, glaciers otherwise representative from a morphologic view- point should also be as nearly as possible in equilibrium at present to be useful as climatic monitors. With the large population of cirque glaciers present in the Juneau Icefield region, it should be possible to locate individual glaciers in or close to an equilibrium condition for use as specific climatic monitors to help clarify the broad climatic change/glacier reaction relationship. 131 Short response times should generally be best exhibited by glaciers having a moderately steep gradient, small size, and a simple configuration without present or incipient overflow gaps or periodic inflow from adjacent higher glaciers. The effect of orientation on the selec- tion of representative glaciers involves two main ele- ments: (a) exposure to or shading from solar radiation, and (b) the direction of prevailing winds. These two fac- tors are difficult to separate in southeast Alaska because of the dominant southerly and southeasterly winds. The strong contrast in glacier behavior indicated in Figure 26 suggests that the north- or northwest-facing glaciers range through fluctuations of significantly greater magni- tude than the south-facing glaciers, and lends support to the former as more sensitive climatic monitors. South- facing glaciers, however, may provide a shorter response time than their north-facing counterparts because of their generally smaller dimensions. The location of cirques with respect to broad areas from which wind-drifted snow can be obtained would also favor north-facing cirques, in View of the prevailing southerly wind. To monitor major lateral shifts in prevailing storm tracks, a wide north-south and east-west distribu- tion of reference glaciers should provide an interpretable regional response to accumulation season storms, and also to summer storms of great ablative capability. 132 Figure l7.--Vertical air photo of Gilkey Glacier terminal zone (July 4, 1962). 133 t. Gilkey Glacuer c. ' "hr ' ‘1 . .. . r, ,a (it, .\ 3&9”? -. ‘1 '- 1w “LEA-173143”? ' 1' ) ,1).. . u H 1,.- . ‘N J“ x , 7., , firm. , |.0 mi. u H 0.5 |.5 km. 134 Figure l8.--Vertical air photo of Antler Glacier terminal zone (July 4, 1962). Approximate 1948 ice front position also indicated. 135 ' -. AMI" 7’? Glacier O'H-o 0 U1 hi: 2. 0.5 1:0 136 Figure l9.--Vertical air photo mosaic of the Norris and Taku Glacier terminal zones (July 17, 1962). 137 I, I “ l l i - I :5" )‘F‘H‘W'E‘ 9431:: ‘1 .‘C’ - ‘ ' 3151?};(‘a‘1'85’l’1 "”L’f.’ .‘f? 1"“ ‘3‘" ( . [/4 '. [fifth], "hid“; ”'1 fivi Ir .5“ It‘s. . ".k 138 .Ammaa .om messy mcoN HMGHEuou uoflomao cowofl>mo mo badmoE ouonm Ham HMOHuHm>II.om musmfim 139 .2 / .6 \ ann-w-n:b w... {$2051. w§$ ovua nw;on-9 \ 140 Figure 21.--Vertica1 air photo mosaic of Wright Glacier terminal zone (July 17, 1962). Approximate 1948 ice front position also indicated. 141 EKW-B-bl EKW-8 11g; 0 142 Figure 22.--Longitudina1 profile of Lemon Creek and principal tributaries. 143 mUKg§888=§ . Inn .5 553 5.2.... _ . : :5 I.I.I”u...-..,.......j.._... a _..____.u_:...§s§$s 4A3£=s==ax . .:5.§=.S=“..e§“~\¢ t. A. .2 .s . u. :5 wwSSasfl r - - .3 838+ % .y 11 m $0 Inna-2% munm? 16391.“ I ~ 3. e In“ 9. . n.. --~N~s~9nvx I 16 III I III, \ \‘3- :\\\\\“\ 0 4r,” 1 MI 59%?”9” o“ #:‘\“:\ “‘\ “ x V \\ \\\ N“ \\\ ‘I\\\‘\“\ “§‘ é gs e s \\s Q \ \ ...w.......¢ s‘s\\\ I I 4,0 . \. \\‘ .\ x .\ .\ o 9 e\“.\\\\\“ “ \x‘ o“ o\\\‘\\ \ o“ ‘9‘“ \ \ 9 \\\ \ fi‘ “:2. I‘““ \ g d" u ““u up an! n a I mum \\\ \\ ‘9‘ \\ . W m“ \In‘ \ um‘fi“\\\:‘“\\\ unfik‘um ' ll fl“ “mm \ “Yum“! ‘ m“ ”w“ 5%“ m mu“ m» I I l I eI/II Ia II III §>8g2§88lx§ 148 Figure 25.-~Orientation of geomorphic elements in the Antler and Gilkey drainage basin: (a) orientation and cumulative lengths of low-order straight stream channels; (b) orientation and cumulative lengths of large, straight glaciated valley segments; (c) orientation of well-develOped glaciated and deglaciated cirques; and (d) orientation of active cirque and cliff glaciers (Juneau C—2, C-3, D-2, and D-3 sheets). 149 _. ._ %%I.$ . .‘t. . .. . .fiwflwfiwgo I _ w: I. veoflfl: «NW. ....I _ Mm. . WWII}... . .........u....¢........ . .IM....“.M:¢...N.I . . s c. .. \\ ¢ \\‘ If v :I % “In? awoken! . I I... ......... 9.....3 $5 I. saw I... «my»... fig... IWWII. i. I... III. «WM. Ififlfiw WM... \.\Mm.wwwfifiwuwwm H- .mMImmmmww IIII .flIIIhMMquI I”. .Iwwmmm 3:1... 33.4 “fi ,. .. i\“ I . I. I ,I II: mm Ill“ uh“ III “I“ | I m .m. .. \ II mmmmwmmm... wwwmmflmafi .. “figmwmfiwwfi ‘. x I! “W'- FIIIIII I. m‘ 5!. 6 &I I I JIM” I a. . I V»? eggmmm .Iw. \ 0.3% Janis... 2.. mm. I "III “a." *7 ”2". 4?: n., II is... u s . .1 . I. NIH... M§W§I . .3, II .. 9.. ~ .4 +4.}... «s. .3 5 I. IIIII. Ium I I 436...... I . .13 tI .IIIIJI w Iunm ..I 3...... . II 111me a 08.5.1.1...93 .III .IIIhIIII m1: ..... .e. III, .11 e3... 1.... I IIIIIIIII III I I... :2. . .IIIII1 III .13.... I}... . IVII...IIIIIIIIZw/111 1111.3...» . I .IIIIIIIIIIIMIIIII ......... . IIIIIII I1 I 11 ......................“.S... 3.... “Iva/IIIIIIIIIoa/I IIII111I1IIIIIII I III. .. .. .e... s III I III IIIIIIIIII I IIIIII. III. in ..... s ...... ... III/.12. I... III I II I II II I II II I II .....“.......“..n..n...55.. . I..“I....I....HII4III.II4IIII.III IIIIIIIIII I I... IIIII . . . : I I I IIIIIIII ... .11... s .I... IIIIIIIIIIIII II .IIIIII 1...: S... = S :3. I I III I IIII . _ anvil/III: .1. .........:= . s: i: I: III 0... I .IIIIIIIIIIIIII . . .. . . . I: III I. II II .......\.... .. IIIIIIIIIIIIIIIIIIIII 3...... .....«.. 5...: . .. :4uI:IIIIIIIIIIII . .. .. .. 3...... I IIIIIII 111.1. . . . .. =r. . .. .31... Ina“.......s.s..s I. I. IIII..,..I...I..II. ........ , IIIII .IIIIIIIIIIIIII . . ............ s s .I. IIII II II II II II II I . .2. . .. 3...... II II I.I II II III ..................... : IIIIIII ,IIIIIIII II IIIIII IIIII ... :53... .. IIIIIII..IIIIIIIIIIIIIII III I I ...: IIIIIIIIIIII .. . ..I II III. II III IIIIII . I... IIIIIII III .I . III III II II II III IIII III III II I I5— 54 I :IIIHIIInI III . II... II II 4?: .. II . a...“ ~— 5. . .2. . x... .. I .. I. I II a. 3.... t .1 I I . I. . g............§....¥. n a n I II... . .. =..§.....1.e2 I II III. IIIIIII... II... .3. . . .. . .. . .. . IIIIIIIIIIII 2.. .... IIIIIIImIIIIHIIIIqus .... .....M....\... 3...... Ian/.3... II. IIIIIII I Means»... I... IIIIIIIIVIIIII? . III III 0101 MI II .I..............I...... I... IIIIIIIIIIIIIIIIIII I/IIIIII: If IIIII. I... 3......I... In... IIIIIIIIIIIII.I (“VIII/IE... a a... . .I «III.» $335.... 5......“ II eat... I. I. 3...... u I3: I. I... I III II II I. II. III... 2...... . v . .... Q... I11III141 IIIIIIII IIIIII IIIIIIIIIII. “35...... I1 IIIIIIIII II II.III III, IIIIIIIIII I... .4“.an IIIIIIII n.9,, II... I... I .I II... 2.4.2,. I... $.31 a a I . III...» I... «A? I I .vu. M h .7 IIIMI III. m I»? III I 11. $me ”WWI WumehwquIw. ..I I II I. I . a I... a... III... I mummmmmwfiww .. I. «III...» I . I I :I . III... .. .3... AI 4 {II , II Kass ##3##. I mmm «mum mum. nu "III "In mmmmuu—zfiimw— mummammmmflfiuwfl I IIIIIIIWIIIIIIIII .. I .59.. 45% . . «I. fir ..I Inn 0... % n "I a$¢fi%.%¢«.é. , may. ......... .r IIII. a». an: . . . s I. III... MIMI. wanna. . I I. 3% 3.3%. I... ., It... I... Id“... fixer WIM III... IMIII I 3., .II, .7» Antakwnv MAW II... «In... .. III...» . . I. ... .. I I I I I . . III Mk“? age: I. o. o.................. qunIIIIIIIIIIIIII..IIIII . IMIIIIII .... . . i.e....IIIIIvI. £9035...» I I IIIIIIII III. «F9......I.../.I.II":I... .fiexs... .I,,.II.IIIIII 2.4%.“? I I4“: .. I. I IIIIIHIIIIIIII .5 .........._.._.............I..~w~.. ..I..“,I...,..IIII,.IIII , .I I IIIII .III II :IIIIII/IIIIII III IIIIIIII ....u....uv~”~ I I. InIII 150 Figure 26.-~Terminal elevation variability of active cirque and cliff glaciers (Juneau C-2, C—3, D-2, and D-3 topographic sheets): (a) data plot for individual glaciers; and (b) mean terminal elevations plotted for 900 arcs centered at the north, east, south, and west. 151 N10 3' 010 El «In 300.14. III DIG 30“. Ill "Duh—0: muszgsa =5 . ... ~ 8 ......w....... .....5......... a . ... .. ... . .5 5. 5 5 1. .. ... ... .... 5 .5 5 5 5 11.1 .. .1. ... ... 55 5 5. 5 5 1... .1 .. ... ... . .5 .. 5. 5 5 5 1 . 1. .. .. ... ... .... .. 5. 5. 5 5 5 s 1 1. 1. 1. .1 ... ... .... ... 5 5 5 V 5.. 1 1 1 .. .. .. ... ... .... .. ... ... 5 5 55 s s 11.1.11.... .1. ... .... ...: ... ...... 5. 5 ... 5 5 5 55 111111 1. .. 1. ... ... .... ...... ... ...... ... .5 .. 5 5 555\ 1 1 1 1 1 1 .. .. . ... ...: .....5 5 .5 5 5 5 5 s \ s5 1 111111. .... ... ..... ... ..... .. .... ... .. 5 5 5 s s s s ~ 1111 1 . 11 1 . .. .. ... =2... 5...... .... ... 5. 5 5 s 555\ 1 111.111...... .. ... .... .. .... ... .. .. 5 55~\\~s 11 1 11 1. 1 .1 .. .. ... .... .5... ...... ... .. .. s x x s s s \s \ 1111111... . .. .. ... .... ..... ... ... 5 ..5s5.\\\\ 11111111.... .. ... ... .=:_. .5... ... ... .. ......VVes \\\ 1111111111111 .1 .. ...... ... ....... .5. .... 5 .5 5 55 s s \\\\s\\ 11111111....1. .. ... ... .....\....5\\\\\\\s\\\\ 11111111.... ..x........s.\s\s\\\ 11111 1111.11.94. .5. .... ......ssV\9\V\‘ 3 11111111111104... .. 555191.130: 11111101101110.1144... ... ,§¢s\.\\\\\\\\\\\‘\\\ 3 111111111133... . . ... .....VV ......VVosox 111111111111 11111111 11.11... 5 4......» .e o\\\ 11 \\\\\\ 11111111111113}. .. , . . 5 5 .. \xkv‘fi‘xxvofi 11111111111111111111111111 .... 1.11.... .. .. . ... 5 5 ....\\\. \xavs \\\\\\\\\ 11111111111111.11 .../1 . .... . . .351 s 1. . \\ \\\\\\\.fl\\\\\.v 1111111111112... . .... ... . .... .. 11.1 Re .313: \\ 11111111111111.» 11,1 1/1 . . .....1 .. ...... 1.15. V . \\.v\\\\\\‘« 11111111 0111137111.,1/1/1111 .A .1. 1...... ... ....... 1 1.59., ....wvxs m \\ \uwx amnuuuWuMua 1&1, .1111, . %%%..g......_.___. 1.331,. $111»... W... «. www.mmmfiw .... .../... Vwfivv 31%. «m m... 1 11% 1 . . \ \ ..., ..uuS. mu, m 111.111.11.111? 1 V % Wm. "M11 .1». "1 wag/11111 Wm mmmu. m % .uunnlwmwaoofinz/ » $3 mu ... l uylnmwm WWWM: 17,111 g “MM”? ’l '1 1 11' 9N ‘ “I ‘ Q‘ 5:: ’4 ii 13.: E . .. 4.1.11111’1 . ...... 1’11 41 ’01 1 ’“Il MW 11, "'11 ”fl 5415 5.7 11 #3:: R .. @ I w «fir 1 111111111111 m n \Mmmwwwmwmmwfix.» .. ..ngufiwmq . WMMW: mm w". “v.“ \ ., 144$ ...... ....111111111uu . 01.11 1111 uwuw.u.\..“\\ 11o. 5...... .. .6. 111111111115? 13.: in. \\u«««\\\ @115. .35 e. ..........».....11111111 71114., 11111971 3.33.. 1... \VV. «1‘s .... ....» ... .. 111111.11. 1111111. 1111111 uh ‘\\\.v\\\\ .3 . . .... ... ......11111. 11771111111101} \\\¢\\\\\ .\ \\\\. ..... ... .11. 11 .11 11111111111 11111 \\‘\‘\¢ \\ .\\.. .. 1 1111. 1111111 11111111h11 \\\\\\ .\\ .\ 1111 11:1!!! 11 1 \$\\ \\.\\\ .\\~\ 11 1 1 1... 1 1 111111011111 \\\\0\\\. \\V\ ...s .111 .1... 11111.11111111111111111111 \\\\\\\\\\\\\\\\\\s.\\‘ 111.2... 61.11111 111111111111 :3. VV... . ..e.....7..Z.1/,1/111/1111 8 1111...... V n 1...... .../33131115111 8 §\\\\\\\\ . . n . ........ 2...... 111111111111 \\\\\\ V‘s ... . .... .... .../.... 111111111 \\\\\\\5 s... .... ...... ....V A . .. .. 111 1111111 V\\s.. . . .. ...... . .... 1111111111 ‘\\\\55. s ... ...: .. ...: ... .... 11.1 11111111 Voéssss 5 ...... .. . . ........... .... 11111111 s \s s x 5 5 5 .. ...: .. .... . 1 11111 11 51.5 5 ... .. .. .... ... .. .... 1. .... 111111 3...... .... .. ... .5. .. ...... ... .... w. .. .... .11 111 5‘ 9s 5 5 5 5. .... .... ...... 4r ... 1 1111111 11.55.111.55: ...: .... .....fl. 2.1.5.1... 5555.5. 5555....5...=. 1 ... .....111111 e 5 5 5 .5 5 .5 .. .. . .. .... .1 .1... s. .. ... ... ... .... .5... . ... ... ... ... 1. 55 5 5 5. .5 ... ....... .. . .111 1. 5 .5 555.. 5. ...... .. ..... . .1 5 5 5 .5 .... .. .. . .1 .55. 5. .5. .. . .... .. . 8 55...: .......... .. ... ........... ... 5...... .. «#20: H! gfi >44(U_0040:§02 kw b. lop—(E 3.32. H.023 320: m0. 330!— 8 20:.<>w._u 12.3w.— 5.30m Z. 32: mpg—X0“: .5 on ... . .... 8 5‘ o =~ : ..--.-.................. ...R.........a...§...~ 11.... .. 5 55s~ .1 .. .. .. ... .. . ... ... 5 5 5 .1.fl%.%.»%.-~u.-~...~........................ .. ..u............W.........“.uuw“..Nufiufiflsuwss 1. 1.11.... ... ......... ...... .....E: ....S... .5. ... .5. 5.... $1.... 1111 1111.11... .. .... ...... ......=.: ......5. .5. 55 .5 5. 5 5 ~ ~ \ 1111.11 1. .1 ...... ... ...: .... ...... .5. .5. ..5 .55 5 555 5 111111111 1.1.1 .. .... .... ....... ...... .... .5 ... 5 5 55505 1111 1 1 .1 . .. .. ... ... . ...... ...... ... ... .. 5 5 5 s s s \s 1111111. .1 .. .. ... .... .... 5 5 5 s ss5~ss ss 11 1 111 1 . 1. .. .. ... ... ...... ..... ... ... ... .. 5 ..5 V ‘5‘. 11111111... . .. .. ... ... ..... .... ... .. ..5sssx\\\\ 11 111.111.... .. ... .... .... ... ... .. .. . . .ssV \ss 11111111.... .. ... ... ... ...... 5. ... ..5555\5\ \\\ 1111111111111 . 1. 1.. .. .... ..... /. .... 5 .. x. 5 5 s V \ \\\\\\\ 11111111....1. ..... .... ... .... .... ....5..s..\\.s\¢\\\\\ 111111111111... .1 ... .. ... .2. .... .. .. V. .. .. . . s .e\\\\\ 111111 111111.144... .... c . ... 55...... 5\\\s\\\\\\\\\\ co 11111111111711.1312... ...... ... 5 .........$.......V\s.\\\\\\\\‘ S 11111111111111 .11....2... .. s 5 ... ......xxxxsxxxs\\\\\\‘\ 11111111111111.41. .1... .. ~ 5 5 V. stxeoxxssxx‘xv \ 11111001110110.1111 1.1 .. .4 5 1... ....xx\\\\\\\\\ \\ 11111 I 111/11... 114. ~ ~ ~ \\\\\\\\\\\\ \\\ 11111011111111111111111. 1 1 .. ... .5 5 .5...\.\\\\\\\\\\\\\\\\\\ 11111111 11 1 1 1 11 1.11.... .. .. . 5 ~ y u ....VVx \\\\\\\\\\\\\ 1111111111111111111111 1 1 .1 ... .5 5 . s s ..\\\\\\\\\\\\\\\\‘\\\\ 1.111111111111111111111111 111111. .. .. ... 5 s55. V\\\\\\.V\\\\\\\\\“\ 11011111111110/III/1 11 1 1 . . 5s 5 5‘ \‘§\\\\\\\\\\\\\\\\\\\|\ 111 1111/ 111111111 111111.. ... .. . ~5s s\\\\\\\\\\\\\\\\\‘\u\ I! h 110111111/11 111 1 . 5. 5 s \ ‘\ \\\\\\\\\\\\\\\\‘\| 11hihd1111110111/1/11/ 11111 .1. .. 5 555s§$‘\\\\\\\\\\\\\‘\\n\\\ 11111M1u111111 111111111 .111. . ..5...¢\\ss\v\\\\s\\\o\\\\\\\uu 1”me 111.11111/1111/ 111111 1 . ... .. 55\\\\\\\“\\\\\\\\\\u\t| 1 Ilhllfllll 111 11111 .5 ... 5 s \\ \\\$“\\“\\\\\\\\\‘\\I 1n nnnlnM001111111111 1111 ... .... sxsss \\“\\“\\“«\\““\\\u‘ nu WM." nun/11711111111111 5...».\\\\ s\\\ m“m\\n“«|un 1 111111111 1111,, 5 9.5V? \\\\ \\\“\\\ | mmwmflwmmm ......2. ......u....=....w.v§m “l... #11,: 111111 1 ._ ... . s \\\\\\\ \\...u\ I Int flu 111111111. . .. ss\‘\ \\ 9||nmm| mmimmmlmumn .unmlmumnmaaloowv». ..§§mwm%« “qwmmm numnuu firm/11.1 1111...”. .13.»? wmmmmmn mmmn. "L..." ”I! ,. n I nu IN“ I IHMfl" , mmmuflmm “Mltul 9. "n ll fin" nll 0“ mmmmu mmmwu mmmwmmmmummflm u 1 umflWMMMW M |I Mun H #1". Imuhl an W MWM1fl 1 I Ififi 1111 umanummmmw .1111 1,111 m H11“ 1 m u. 3..meme @311 ,, mm 119111- \\ 1 “\mfirmwmfiww: \ . . a. .1411416111fline I. fl \\ wfiflflflfififlyflwfl“ ¢\\\\\\\ ”...”...Wgéhllllfllwllnhflllllllllllh \ vi“ «Va. \\\s .....11 1 111 91111 \fl“fl«fl\\\$$\\¢ \5555. E: .... 11 11111 11 11110111 \\\\ ¢\\.v\\$.. \ s x s . .. .1. $1111 1 100/. I .111111 ‘\\\“\\\\\\\\\¢ \‘QQst 5 _ .. .1 1 1111 001/111 III hlhlllll \\\\$\\\\\\\\\\V . 9555.. . a. . .. .1. .1 ozoozooopqm/ 111111 ‘\‘\.\o\o\§\. .155 «5 . ... . 1. .../11111111111 111111 \\\\\ \\\ $.\ *5 5» s . .1 1 14. 1 10111001100011 o\\\\\\\o\\\\ x$55 .5 . a a. . .1111.1fi1/71111/11111~1011111 \\\\\\\\\~ . ..V. .6. . ~ . .... . .. .... 11.111111111111114 \\\\\\\\\\¢.\V\¢¢$§ . 5 . . ..... .. 9...... 10 11111111111111 3 3X... ...... .... . .. .. ... .. 9.1.1.1.:11/11/1/11 8 «‘1‘o\\o\.\\.\...\. .... V. .. . .../.... ..., 31111111111111 3 \\.\\V. 1.38.... «“5... .. ......a... .../31111111111111: ‘11:... ...\... .... .... .. .../1.111111 11111 ‘\\\\s\655\5 . ... 7... .. . $1. 11111111 ‘\ \\ s s s 55 .55 .5 ...... ... ..../... . 1. 1111111111 1111 65‘0115 ....\. ... .... . .... ... ... .. .... 1 1 11111 is. V .... .. ...... ... .... . . . ... ...... ... 91.111111 \s55.5.. .... ... .... ... ..... 6.1.1.111 \ssV5. 5\5.. .5 .... . .. .1 1111111111 V5555 s. 5. . 5.... ...... ... .. .. .. 1. . 1 1 1 s-ss.... ....5 ........1..1111 5555s 5... ... .... ... .. .. .... 111 5 s. 5.. ,5. ...... ... ... .... ... .... .11. ss...55.5.5 .. .. ....... ..5....... ......11111. 5 5 5 5. 5. .5 . .. ... ... .. .. 1 1 5 5 5 . .. .. .... ... ......11111 N. 5 ~55 ~ 5: . .... ... 1a '1 o 55 ~ 5. I. ... .. .. .1 . 5 5 5. .. .. ... .. .. 8.1%... «9.4....47... 8 ... ......a... 152 Figure 27.--Tree growth rate curves (a) Picca sitchensis, Antler Glacier terminal zone; (b) Tsu a heterophylla, Antler Glacier terminal zone; (cY’Tsuga heterophylla, Antler Glacier termi- nal zone; Yd) Tsuga heterophylla, Antler Glacier terminal zone; and (e) Tsuga mertensiana, Bucher Glacier terminal zone. 153 m1 nmgsuauaw M1 u‘ D m o H nl sgsuaqam SZElmeZ. 5.30:0 41o: «(g-m 154 Figure 28.--Stratigraphic sections along south side of Davidson Glacier proglacial lake outlet stream. ICE HOW «— omcno~ 06 “C. 7 155 is O «<21 P 4 0 Es 4‘ z .4 UN‘NEAYREED GIAY "LL SYAINED, MAYHND [Ill 0 a It > O U T o <1 § $3-0 ‘ E 7 ‘ SIC. 2 0 4.3!!!" A 33% DA Y! D SAMPLE - I924) - um) lump oacmsc nonuzo~ .' M A.D. |22$3|m .M c. u e )0 n. 9 A.Q. Illotlw KI ‘ C- N 6090- C‘ “ do"; I59. ‘ SEC l 9.5 . [mm swms, Loos — 4 . 4.9.1." 0 § 5 a 3 . , I’I l I . . .. - I II I I U . . I 1‘ i 9 {v (I <1 UV I I I g f 3 g E S o a 3 § E Jr SEC. his" z'é'o'fi 5'. O O 156 Figure 29.--General View of Davidson Glacier buried forest. 157 infirm; v «I. L? 93;... I. . ., 158 Figure 30.--Stumps overridden by Davidson Glacier during the early 12th century, and exhumed by pro— glacial drainage in recent decades. 160 Figure 31.--Trenched sections showing upper till, 12th- century forest floor litter, and lower till exposed on south side of Davidson Glacier proglacial lake outlet stream. 162 Figure 32.--Upper till, containing logs, overlying 12th- century forest floor and lower till, and currently being exhumed by Davidson Glacier proglacial lake outlet stream. REFERENCES 164 REFERENCES Aitkin, J. D., 1951, Atlin map-area, British Columbia: Geol. Survey of Canada Memoir 307, 89p. American Geographical Society, 1960, Nine glacier maps: Northwestern North America: Amer. Geogr. Soc. Special Publication No. 34. Andress, E. C., 1962, Névé studies on the Juneau Icefield, Alaska, with special reference to glacio-hydrology on the Lemon Glacier: unpublished Master's Thesis, Michigan State University, 174p. Bray, J. R., and Struik, G. J., 1963, Forest growth and glacial chronology in eastern British Columbia and their relation to climate trends: Can. Jour. Bot., v. 41, p. 1245-1271. Cairnes, D. D., 1913, Portions of Atlin District, British Columbia: with special reference to lode mining: Canada Department of Mines, Geol. Survey Branch, Memoir n. 37, Ottawa Government Printing Bureau, 129p. Christie, R. L., 1957, Bennett map-area, British Columbia: Geol. Survey of Canada Map 19-1957. Egan, C. P., 1965, Firn stratigraphy and névé regime trends on the Juneau Icefield, Alaska, 1925-65: unpublished Master's Thesis, Michigan State University, 61p. Embleton, C., and King, C. A. M., 1968, Glacial and peri- glacial geomorphology: St. Martin's Press, New York, 608p. Field, W. 0., 1947, Glacier recession in Muir Inlet, Glacier Bay, Alaska: Geog. Rev., v. 37, p. 369-399. , 1954, Notes on the advance of Taku Glacier: in Heusser, et al., GeOg. Rev. v. 44, p. 236-239. Field, W. O. and Miller, M. M., 1950, Studies of the Taku Glacier, Alaska: Jour. of Geology, v. 59, n. 6, p. 622-3. 165 166 Fisher, J. E., 1953, Two tunnels in cold ice at 4000 m on the Breithorn: Jour. Glaciology, V. 2, p. 513-520. Flint, R. F., 1957, Glacial and Pleistocene geology: John Wiley and Sons, New York, 553p. , 1971, Glacial and Quaternary geology: John Wiley and Sons, New York, 892p. Gwillim, J. C., 1901, Atlin Mining District, British Colum- bia: Geol. Survey of Canada Annual Report 1899, Part B, v. 12, 43p., map. Hashimoto, S., et al., 1966, Glaciological studies of the Antler Glacier, Alaska: Jour. Faculty of Science, Hokkaido Univ., Series IV, v. 13, n. 3, p. 237-256. Heusser, C. J., Schuster, R. L., and Gilkey, A. K., 1954, Geobotanical studies on the Taku Glacier anomaly: Geog. Rev., v. 44, p. 224-239. Heusser, C. J. and Marcus, M. G., 1964a, Surface movement, hydrological change and equilibrium flow on Lemon Creek Glacier, Alaska: Jour. Glaciology, v. 5, n. 37, p. 61-75. ' Heusser, C. J. and Marcus, M. G., 1964b, Historical varia- tions of Lemon Creek Glacier, Alaska, and their relationship to the climatic record: Jour. Glaciology, v. S, n. 37, p. 77-86. Kamb, Barclay, 1964, Glacier geophysics: Science, v. 146, n. 3642, p. 353-365. Kerr, F. A., 1936, Quaternary glaciation in the Coast Range, northern British Columbia and Alaska: Jour. Geology: v. 44, p. 681-700. , 1948, Taku River Map Area, British Columbia, Canada: Department of Mines and Resources, Geol. Survey, Memoir 248, p. 1-84. Klotz, O. J., 1899, Notes on glaciers of South-Eastern Alaska and adjoining territories: Geog. Jour., V. 14’ no 5' p. 523-524. Knopf, A., 1911, Geology and mineral resources of the Berner's Bay Region, Alaska: U. S. Geol. Survey, Bull. 446. , 1912, TheIEagle River Region, Southeastern Alaska: U. S. Geol. Survey, Bull. 502. 167 Lathram, E. H., Loney, R. A., Condon, W. H., and Berg, H. C., 1959, Progress map of the geology of the Juneau quadrangle, Alaska: U. S. Geol. Surv. Misc. Inv. Map. 1-303. Lawrence, D. B., 1950a, Estimating dates of recent glacier advances and recession rates by studying tree growth layers: Trans. Am. Geophys. Union, v. 31, p. 243-248. , 1950b, Glacier fluctuation for six centuries in southeastern Alaska and its relation to solar activity: Geog. Rev., v. 40, n. 2, p. 191-223. Marcus, M. G., 1964, Climate-glacier studies in the Juneau Ice Field Region, Alaska: Department of Geography Research Paper n. 88, University of Chicago, 128 p. McCall, J. G., 1952, The internal structure of a cirque glacier: Jour. Glaciology, v. 2, n. 12, p. 127. Meier, M. F., 1965, Glaciers and climate: The Quaternary of the United States, Princeton Univ. Press, p. 795-805. Mercer, J. H., 1961, The response of fjord glaciers to changes in the firn limit: Jour. Glaciology, V. 3’ no 29' p. 850-858. Miller, M. M., 1961, A distribution study of abandoned cirques in the Alaska-Canada Boundary Range: in Geology of the Arctic, Univ. of Toronto Press, p. 833-847. , 1963, Taku Glacier evaluation study: State of Alaska Department of Highways (Report). In cooperation with U. S. Dept. of Commerce, Bureau of Public Roads, and Foundation for Glacier Research, 200p. , 1964, Inventory of terminal position changes in Alaskan coastal glaciers since the 1700's: Pro- ceedings of the American Philosophical Society, v. 108, n. 3, p. 257-273. Muntz, A. P., 1953, Geobotanical work: i2_Neilsen, L. E., 1953, Progress Report, Juneau Ice Field Research Project, Alaska: Amer. Geog. Soc., 61p. , 1955, Recent glacier activity in the Taku Inlet area, Southeastern Alaska: Arctic, v. 8, n. 2, p. 83-95. 168 Neilsen, L. E., 1957, Preliminary study onthe regimen and movement of the Taku Glacier, Alaska: Geol. Soc. America Bull., v. 68, n. 2, p. 178. Paterson, W. S. B., 1969, The physics of glaciers: Perga- mon Press, New York, 250p. Pierce, R. L., 1953, Geobotanical work: in Neilsen, L. E., 1953, Progress Report, Juneau Ice Field Research Project, Alaska: Amer. Geog. Soc., 61p. Seddon, B., 1957, Late-glacial cwm glaciers in Wales: Jour. GlaciologYp v. 3, n. 22, p. 94-99. Souther, J. G., 1960, Tulsequah map-area, British Columbia: Geol. Survey of Canada map 6-1960. Spencer, A. C. and Wright, C. W., 1906, The Juneau Gold Belt, Alaska: U. S. Geol. Survey Bull. 287. Tarr, R. S., 1897, The Arctic sea ice as a geological agent: Amer. Jour. Sci., v. 153, p. 223-229. , 1909, The Yakutat Bay region, Alaska: U. S. Geol. Survey Prof. Paper 64, 183p. Tarr, R. S. and Martin, L., 1914, Alaskan glacier studies: Nat. Geog. Soc., Washington, 498p. Thiel, E., LaChapelle, E., and Behrendt, J., 1957, The thickness of Lemon Creek Glacier, Alaska, as determined by gravity measurements: Trans. Amer. Geophys. Union, v. 38, n. 5, p. 745-749. Twenhofel, W. S., 1952, Recent shore-line changes along the Pacific coast of Alaska: Amer. Jour. Sci., V. 250' p. 523-548. Viereck, Leslie A., 1967, Botanical dating of recent glacial activity in western North America: in Arctic and Alpine Environments, Indiana Univ. Press, p. 189-204. Wilson, C. R., 1959, Surface movement and its relationship to the average annual hydrological budget of Lemon Creek Glacier, Alaska: Jour. Glaciology, v. 3, n. 25, p. 355-361. Wright, H. E., and Frey, D. G., eds., 1965, The Quaternary of the United States: Princeton Univ. Press, 922p. Wright, G. F., 1891, The Ice Age in North America, and its bearing upon the antiquity of man: New York, 763p. APPENDICES 169 APPENDIX A STREAM PARAMETER DATA FOR THE CURRENTLY NONGLACIATED PART OF THE LEMON CREEK DRAINAGE BASIN AND FOR SELECTED BASIN COMPONENTS 170 .AmflE no.vmv ex vm.mm u mCOfluomw wowmflomam mcflcsHUCH mono aflmmn Hamnm>04 171 m H.vm m.moa mmH om.o om.oa m.m m.oa H v no.5 oo.m mm.o mv.a m.v v.5 m m mm.H oo.m mv.o mn.o m.ma m.Hm mm m mm.a Hm.v om.o m¢.o «.mm «.mm mHH H fie ex HE Ex owumm owumm no 2 Hmono gpmcmq cowpmomswflm be: Emmuum numcwq cmmz Aumcmq Hmuoa «Ha\ao.oa Nex\~H.¢ socmsvmnu ammunm «Hexom.¢ max\am.~ mhflmamc mmmcflmuo mus m~.va max mm.mm mmum cemmm «.Auoflomaw xmmno :oemq mo uHmm mam ou acacwmno mmHm hoaam> cam Ho>oo mow Hammond mo m>wm5aoxov gamma mmmcwmno xmmHU nOEmAII.H< mnmma 172 h.vm h.mm hm mm.o ov.o m.o m.o m m mv.o oo.m mm.o mm.o m.w H.OH NH m mN.H mm.m mv.o mw.o m.hH m.m~ mv H owumm HE Ex HE Ex OHDMm Hmono gumcmq GOHHMOHSMHm Hmoasz Emmuum numnmq cwm: HE . m. \om m a I. m.E\ma 4 HS . N. co m numnmq Hmuoe NEM\hw.m msx\mm.m N ex vm.mH mocmswmum Emmhum maHmcoo mmeHMHQ NORM £4...me .hoHHm> xmono nosmq mo wUHm ADHOZII.N< mamma 173 m.Nm m.Nm vm mm.H ON.N H.v m.m m m mm.N mm.m mv.o wh.o m.> h.HH mH N mm.H mm.v mN.o h¢.o m.HN w.vm mp H HE Ex HE Ex OHDMM 0Humm Ho 2 Hmouo Qumcmq QOHDMOHSMHm _e Emmhum zumcwH cwm: mfls\o¢.aa a. . ~.:\mm m HE . N. MN m sumcmq Hmuoe max\Hv.¢ Nex\mv.~ N Ex Nm.Hh 1 mocmswoum Emmupm mgHmcmo mmmchun monm chmm + .NmHHm> xmmHU :OEmH mo mon Ausomll.m4 wands 174 mm.m mm.m NH mm.o mm.o mm.o mm.o H m mN.H o.N mv.o mm.o mm.o hm.H N N Nh.H m.v mN.o ov.o mN.N Nm.m m H HE EM HE Ex 0Humm 0Humm no 2 Hmouo numcmq EOHDMOHsmHm HZ: EmoHum gumamq cwm: AmHs\mm.MHv AmHexmo.sv AmHs om.oc Eumcmq Hmuoe maxxma.m Nex\mm.m N Ex mm.N mocmsvmnm Emmuum huHmcmo mmmaHmnn mono :Hmmm oGflmMQ OUMGflMHU MOOHU Hflflgmmllovm "HHQH. 175 mm.m oo.mH mN oo.H Ho.H oo.H Hw.H H m mm.H oo.v mw.o oo.H mo.N 5N.v H N vv.N mN.m hN.o «v.0 o>.m mH.m HN H oHumm He ex Ha ex OHumm “mnesz “mono zumcwq summon cams numcoq Hmuoe GOHHmonzmHm . Emwhum mstmo.mH max\5m.m Hocmsvmum Emmnum NHE\mv.o NEM\vo.v muHmcmc mmmchun NHE vv.H NEH mw.m comm :Hmmm .mxmmho cowcmu mam HHHEBMm cmeDwn Homno omqucs mo :Hmmn mmmchHQII.m< mnmda 176 mm.¢H mv.MN Nv on.H vn.N on.H vh.N H m Nn.v oo.m mm.o mm.o mm.N mm.v m N mo.H MH.v mm.o mm.o oo.0H 0H.mH mm H oHumm He ex HE ax OHDmm Hmnesz nmcno zumcwq sumcmq cwm: aumcmq HmHOB COHumoHsHHm Emmuum NHE\Hm.mH Nex\mH.m Noamswmnm ammuum ~H5\vo.v max\mm.~ mpHmcmo mmmchHa NHs 4H.m max mH.m noun :Hmmm .chmn mmmcHMHU xmmuu comcmuul.m4 mqmma APPENDIX B CIRQUE AND CLIFF GLACIER ORIENTATION DATA 177 .178 mHmHOMHm Awo.mmv Ham.mvv Amo.ch AwH.omc Amn.mmc Awe.mmc 1w¢.HNV 1mm.o~c HNHHo can onwuHo m>Huom ism.mmc ism.m~1 AwH.Hvo Awm.mev Amo.avc 1mm.HHV Awm.HHV Amv.mmc mwswnHo 0H 4 H ma m N N m cmumHomHmmc nmwao .oH iwv.mmc Awm.mmv Awo.aqc Amo.omc Awm.Hv. imn.mflc Awe.mac Awo.mmv mmsvuflo omumHomHmmo h v m m m N N m ANHHmmc HOV hHucmomm .QH Awe.mmc 13m.mmc 1mm.ovc Awm.emv Awm.mmv Awe.HNv Awa.m V Awm.mmc mumHomHm , Hm mm mm ow. om EH m om m>Huom mchHmucoo mmsvHHo .mH 1w¢.mmv “wo.mmc Awe.ovc Awo.noc Awm.omv Ama.mmv 13m.m v 1mm.omc mmsuuHo cmdoam>mcuaam3 4m om Hm Hm mm am H mm mm some mama :oHumucmHHo .Hm 3 m m z 32 25 mm mz "Amoum OomH Mo mEHmH CHV cOHumucmHHo "Amoum Com mo mEHmu :HV COHumncwHHo .mumo COHumucoHHo uoHomHm MMHHU paw mstHUII.m wacommd APPENDIX C DESCRIPTIONS OF TRENCHED SECTIONS IN NORRIS GLACIER OUTWASH 179 DESCRIPTIONS OF TRENCHED SECTIONS IN NORRIS GLACIER OUTWASH TABLE Cl.--Norris Glacier 67 sec 1. Depth (cm) Description 0 Surface (moss, matted spruce needles, organic litter), underlain by 2.5 to 5.0 cm dark brown organic material. 3.8 to 6.9 Gray silt, 1.3 to 1.8 cm thickness. 8.9 to 17.0 Brown-stained silt. 12.7 to 25.4 Top of grayish zone, transitional with the overlying brown zone. 56.4 Top of deeply stained zone. Sand and rounded pebbles. 91.0 Base of trenched section ended in stained gravel. TABLE C2.--Norris Glacier 67 sec 2. Depth (cm) Description 0 Surface (moss, matted spruce needles, organic litter), underlain by 5.0 cm dark brown organic litter. 5.0 to 7.0 Gray-brown sand and organic matter. 7.0 to 12.0 Light brown stained sand. 12.0 to 19.6 Sand, slightly brown-stained. 50.0 to 96.0 Medium sand; coarser sand toward base. Slightly stained. 180 181 TABLE C2.--Continued. Depth (cm) Description 96.0 Transition to fine sand. 130.0 to 152.0 Very fine sand and interbedded gray silt. Cross-bedding present. 152.0 to 156.0 Silt, horizontally bedded. 156.0 to 160.0 Finely interbedded sand and gray silt. 160.0 to 175.0 Medium to coarse sand and gray silt. Abundant biotite is present in several thin layers. 175.0 to 180.0 Very coarse sand and small granules. 180.0 to 191.0 Coarse sand. 191.0 to 220.0 Interbedded and cross-bedded fine sands and thin dark horizons containing biotite. 226.0 to 236.0 Medium sand. 236.0 to 241.0 Gray silt. Sharp contact with over- and under-lying strata. Faint cross-bedding indicates current flow toward southeast. 241.0 to 267.0 Finely interbedded find sand and silt. 267.0 to 297.0 Interbedded fine sand and silt, darker than overlying stratum. 297.0 to 320.0 Gray silt. Micaceous. 320.0 to 344.0 Fine gray silt, darker than overlying stratum. 344.0 Top of organic horizon about 7.6 cm in thickness. 352.0 Top of orange-brown stained gravel stratum. Staining is largely restricted to pebble surfaces. 366.0 Base of trenched section, ended in stained gravel. APPENDIX D DESCRIPTIONS OF TRENCHED SECTIONS ALONG DAVIDSON GLACIER PROGLACIAL STREAM 182 DESCRIPTIONS OF TRENCHED SECTIONS ALONG DAVIDSON GLACIER PROGLACIAL STREAM TABLE Dl.--Section 1, located on south side of Davidson Glacier proglacial stream ca 10 meters below proglacial lake outlet. Depth (cm) Description 0 to 20.0 Recent alluvium. Pebbles and sand deposited by the proglacial lake outlet stream. 20.0 to 180.0 Unweathered gray till. 180.0 Locally, a thin layer of gray silt occurs at the base of the unweathered gray till. 180.0 to 250.0 Weathered till, slightly stained orange- brown. Staining is largely restricted to surfaces of cobbles and boulders. A few decomposed rocks are present. This till contains a higher proportion of boulders than does the overlying till. Most of the large particles are rounded or sub- rounded, but angular fragments are present. Base not exposed. 250.0 to 305.0 Covered by slumped material. 305.0 Level of proglacial stream surface. 183 184 TABLE D2.--Section 2, located 4.1 meters downstream (NE) of Section 1. Dep th (cm) Description 0 to 40.0 40.0 to 60.0 60.0 to 140.0 140.0 to 145.0 145.0 to 154.0 154.0 to 290.0 290.0 Recent alluvium. Pebbles and sand deposited by the proglacial lake outlet stream. Sand and pebbles. Gray till, containing more sand than the underlying till. 2.5 to 7.5 cm gray silt, underlain by organic material (forest floor debris). Weathered till, stained orange-brown. Staining persists through uppermost 61 cm. Granodiorite pebbles only superficially stained; micaceous schist fragments have generally disintegrated. Base not exposed. Covered by slumped material. Level of proglacial stream surface. 185 TABLE D3.--Section 3, located 4.3 meters downstream (NE) of Section 2. Depth (cm) Description 0 to 22.0 22.0 to 95.0 95.0 to 118.0 118.0 to 122.0 122.0 to 126.0 126.0 to 190.0 190.0 to 250.0 250.0 Recent alluvium. Pebbles and sand deposited by the proglacial lake outlet stream. Unweathered gray till. Upper part trun- cated by recent stream erosion. Sand and gravel lenses are present. Cross-bedded granules and pebbles. Cross-bedded sand, 2.5 to 20.0 cm in thickness. Organic material, 2.5 to 5.0 cm in thickness. Weathered till, stained orange-brown. Base not exposed. Covered by slumped material. Level of proglacial stream surface. 186 TABLE D4.--Section 4, located 3.0 meters downstream (NE) of Section 3. Depth (cm) Description 0 to 12.0 Recent alluvium. Pebbles and sand deposited by the proglacial lake outlet stream. 12.0 to 115.0 Unweathered gray till. Contains sand and gravel lenses. Logs and stumps present. 115.0 to 118.0 Cross-bedded silty sand and granules. 118.0 to 122.0 Gray silty clay. 122.0 to 186.0 Weathered till, slightly stained orange- brown. Base not exposed. 186.0 to 246.0 Covered by slumped material. 246.0 Level of proglacial stream surface. TABLE D5.--Section 5, 187 I located 4.9 meters downstream (NE) of Section 4. Depth (cm) Description 0 to 8.0 8.0 to 110.0 110.0 172.0 182.0 193.0 200.0 206.0 212.0 216.0 263.0 307.0 to to to to to to to to to 172.0 182.0 193.0 200.0 206.0 212.0 216.0 263.0 307.0 Recent alluvium. Pebbles and sand deposited by the proglacial lake outlet stream. Unweathered gray till. Many large boulders present. Sand and pebbles. Several blocks of till are present. Sand and granules. Clay and silt with organic fragments. Cross-bedded sand and granules. Sand. Silt. Peaty organic material. Till; pebbles and cobbles are abundant. Base not exposed. Covered by slumped material. Level of proglacial stream surface. TABLE D6.--Section 6, 188 located 80 meters downstream (NE) of Section 5. Depth (cm) Description 0 to 12.0 12.0 to 118,0 118.0 150.0 155.0 160.0 200.0 218.0 264.0 272.0 277.0 278.0 280.0 282.0 338.0 353.0 357.0 360.0 434.0 450.0 to to to to to to to to to to to to to to to to to to 150.0 155.0 160.0 200.0 218.0 264.0 272.0 277.0 278.0 280.0 282.0 338.0 353.0 357.0 360.0 434.0 450.0 457.0 Recent alluvium. Pebbles and sand deposited by the proglacial lake outlet stream. Unweathered gray till. Lenses of brown silty sand are present. Brown silt and sand. 5.0 to 7.5 cm bluish clay. Organic material. Silt. Sandy gravel, slightly stained orange- brown. Silt and sand. Silt. Sand. Coarse sand and granules. Silt and clay. Plant fragments. Peaty organic material. Silt and clay, with a few small sand lenses. Sand. Silt. Peaty organic material. Gravelly till. Fine silty sand. Deeply weathered till(?) with abundant pebbles. Identity uncertain because exposure lies below present stream level. 189 TABLE D7.--Section 7, located 40 meters downstream (NE) of Section 6. Depth (cm) Description 0 to 10.0 10.0 to 230.0 230.0 266.0 300.0 306.0 312.0 318.0 326.0 395.0 488.0 to to to to to to to to 266.0 300.0 306.0 312.0 318.0 326.0 395.0 488.0 Recent alluvium. Pebbles and sand deposited by the proglacial lake outlet stream. Unweathered gray till. cated by recent stream erosion. logs. Upper part trun- Contains Silt and clay. Sand, granules, pebbles. Clay, stained orange-brown where overlying organic zone. Organic material (former forest floor litter). Silt and clay. Clay. Weathered till, stained orange-brown. Sand lenses present. Base not exposed. Covered by slumped material. Level of proglacial stream surface. 190 TABLE D8.--Section 8, located 30 meters downstream (NE) of Section 7. Depth (cm) Description 0 to 12.0 Recent alluvium. Pebbles and sand deposited by the proglacial lake outlet stream. 12.0 to 217.0 'Unweathered gray till. Upper part trun- cated by recent stream erosion. 217.0 to 290.0 Silt. 290.0 to 294.0 Compacted organic material. 294.0 to 306.0 Brown and gray clay. Roots are present. 306.0 to 445.0 Weathered till, stained orange-brown. Base not exposed. 445.0 to 520.0 Covered by slumped material. 520.0 Level of proglacial stream surface. APPENDIX E LIST OF VALLEY GLACIERS IN THE VICINITY OF THE JUNEAU ICEFIELD AND DENDRO- CHRONOLOGICALLY DATED LATE NEOGLACIAL ADVANCES 191 192 APPENDIX E LIST OF VALLEY GLACIERS IN THE VICINITY OF THE JUNEAU ICEFIELD AND DENDRO-CHRONOLOGICALLY DATED LATE NEOGLACIAL ADVANCES Glacier name Bucher Glacier Davidson Glacier Eagle Glacier Gilkey Glacier Herbert Glacier Lemon Creek 61. Mendenhall Gl. Norris Glacier Taku Glacier Twin Glaciers Location: (U.S.G.S. 1:63,360 (sheets) Juneau D-2 (v. fig. 10) Skagway A-2 (v. fig. 12) Juneau C-3 (v. fig: 8) Juneau D-2 (v. fig. 10) Juneau C-3 (v. fig. 8) Juneau B—2 (v. fig. 5) Juneau B-2 (v. fig. 5) Juneau B-l (v. fig. 4) Juneau B-l (v. fig. 4) Taku River C-6 (v. fig. 16) Maximum Minimum age advance beyond maximum date advance late 1500's 1750 - 55 i 425 1410:100 ? 12101100 ? ca. 1785 ca. 1783 ca. 1765 600 ca. 1750 10,300 1767 — 69 630 1910 - 16 550+ 773+ 1755 - 65 ? 1750 - 55 570+ 1775 - 77 Sources Beschel and Egan (1965) Egan Egan Egan Lawrence (1950) Huesser and Marcus (1964) Lawrence (1950) Huesser and Marcus (1964) Lawrence (1950) Lawrence (1950) Muntz (1955) Egan Lawrence (1950) Lawrence (1950) APPENDIX F WRIGHT GLACIER RECONNAISANCE STUDY 193 APPENDIX F WRIGHT GLACIER RECONNAISANCE STUDY Location and Description Wright Glacier is located south of Taku River and about 25 km. (16 mi.) southeast of Juneau (v. Figure 3). Most of its drainage area is shown on the 1:63,360 Taku River B-S, B-6, C-5, and C-6 quadrangles (Figures l3, 14, 15, and 16, respectively). Vertical aerial photography of the terminal zone was obtained in 1948 by the U. S. Geo- logical Survey, and in 1962 by the U. S. Forest Service (v. Figure 21). Little information regarding Wright Glacier is available in the literature. F. A. Kerr (1948, p. 16), in the course of a bedrock mapping project carried out nearby in 1932, reported that Wright Glacier showed ". . . evidence of rapid retreat in the last 20 years." D. B. Lawrence (1950) did not conduct field work at the Wright terminus, but noted on the basis of aerial photographs that ". . . its position of maximum advance was only 0.7 mile beyond the 1948 ice front, and the posi- tion of the apparent shore line would suggest that the terminal moraine was formed under water, as it was at the Twins." (The "apparent shore line" to which Lawrence 194 195 refers is that produced by the damming of upper Taku Valley by Taku Glacier in the mid-18th century.) General Assessment of Wright Glacier's Terminal Fluctuations The writer's 1966 field observations at the Wright Glacier terminal zone were made without an increment borer long enough to date the large trees on the older de- glaciated surfaces, and without a set of the 1962 aerial photos. Access to much of the deglaciated terrain was not possible at the time of the field study because of two unfordable streams (Hidden River and the proglacial lake outlet); a boat or inflatable raft is a necessity for effective work in this area. In View of these limit- ing factors, the field time was spent in a general reconnaissance from which specific recommendations of future objectives might be made. Evidence of Past Fluctuations Although end moraines are present on the outwash apron between the proglacial lake and Taku River, this area was repeatedly submerged about two centuries ago by lake waters when Taku River was dammed by Taku Glacier during its maximum recent advance. This area could not be safely reached by the 1966 field party. However, high trimlines are well preserved at places on the Wright 196 Glacier valley walls, and reveal a long history of former advances. The largest trees on the highest distinct trim- line could not be dated with the short increment borer available, but on the basis of the cores obtained, the Wright Glacier maximum recent advance is considered to have occurred no later than 1780. The actual date may have been many decades earlier, but cannot be confidently estimated at present. At an unknown date since the 18th (?) century advance and subsequent recession, a significant readvance of Wright Glacier took place. This readvance is marked by a conspicuous lateral moraine a few tens of meters below the highest main lateral moraine. The oldest trees found on the lower moraine were ca. 75 years of age (by ring counts to the growth centers of Sitka Spruce trees); substantially older trees may be present elsewhere on this moraine, however, and thus the 75-year age is minimal. At the time of each of these two main advance stages, Wright Glacier formed a dam across the lower end of a large valley to the southwest of the terminus (Hidden Creek valley). At the site of this former ice dam, erosion by Hidden Creek has since cut a lS-meter vertical section in which two distinct tills are exposed. Lack of time prevented a detailed examination 197 of this significant exposure. The lower of the two tills was characterized by a gray color, a large proportion of angular boulders and cobbles (including many rock types which do not outcrop locally), and little evidence of associated fluvial action. The upper till varied from tan to light gray in color, contained much interbedded silt and sand, with few cobble--or larger-sized frag- ments. This till was overlain by a variable thickness of silt which may have been deposited by a glacier- marginal lake. End Moraine Morphologic Zones The end moraine zone was not examined in the field; the following discussion is based on subsequent examination of the 1962 vertical aerial photographs. On the basis of vegetational differences visible on the 1962 aerial photos, four main end moraine morphologic zones may be delineated. The outermost of these zones is about 200 to 300 meters in width, has an exceptionally smooth distal curvature in plan view, appears to lack visible end moraine ridges, and is covered by vegetation of apparently the same stage of development as that on the adjacent Taku River valley floor. The second end moraine zone is ca. 200 meters in width, and includes at least six closely Spaced morainic 198 ridges, all of which curve in a broad arc without large indentations. A coniferous forest cover has developed on this morainic zone. A few water-filled depressions occur in this zone. A third morainic zone is ca. 150 to 250 meters in width, and consists of two principal morainic ridges which display a tendency toward local irregularities in crestal trend, marking an increasingly uneven former ice edge. The vegetation cover on this zone is markedly less mature than that of the previous zone. The fourth and innermost zone is 100 to 300 meters in width, and ends at the proglacial lake shore. This zone differs significantly from the others in that it has been molded by moving ice into a series of alter- nating ridges and valleys. The ridges generally increase in height and width in the direction of glacier flow, while the intervening valleys become narrower and shallower, leading to a conspicuous "saw-tooth" proglacial lake shoreline. The vegetation cover on this youngest zone does not appear to be more than a few decades older than that on the adjacent belt. Deglaciation of this sector probably occurred during the decade of the 1940's, or perhaps slightly earlier. Comparison of the 1948 and 1962 aerial photos reveals that this retreat was particularly great from the 1940's to the early 1960's. During this short 199 period, more than 2.5 square kilometers (1.0 sq. mi.) of the terminus disappeared and was replaced by a large proglacial lake. In 1966, most of the ice front still ended as a receding ice cliff in the proglacial lake, but part of the southern ice front terminated on land. Although serious limitations in direct field evidence make the following interpretations strictly provisional, the subdued appearance of the outermost end moraine zone shown on the 1962 aerial photos, together with the apparent similarity of the forest cover on this belt and on the adjacent outwash surface, suggests that this zone was submerged by lake waters dammed by Taku Glacier's mid-18th century advance. The admittedly minimal dendro-chronological data for Wright Glacier trimlines do not conflict with this possibility. The later, but lesser, main advance of Wright Glacier probably culminated at a time when the Taku Glacier-dammed lake in Taku Valley no longer existed. The character and timing of these events would be readily resolved by additional field work on the outer- most end moraines and proglacial zone. An additional item recommended for future field examination concerns the possibility of at least one higher (and probably very much older) ice level than that attained during the 18th-century expansion. Indications of this possible high trimline are best 200 observed on the valley wall south of the proglacial lake, and consist of faint linear patterns considerably higher than, but generally parallel to, the obvious 18th-century trimline. .— Figure 4 JUNEAU (B—lI OUADRANGLE ALASKAHGREATER JUNEAU BOROUGH 1:63 560 SERIES (TOPOGRAPHIC) Figure I I ALASKA I TOPOGRAPHIC SERIES I ‘fi I I I I I ‘3 6% [WIN FEFT 1340090130 ‘ V T‘ 5 °( ' I‘ I I l‘ < x j: K ' K "j l: 1‘ I u .. \‘ _‘ I; ‘ I" 'r"\ :3 I? ‘ IN)“ f ‘ ‘ “I ' ,,,2/ L. I I ’ : I I I " j / i ‘ I I I 3 \J % I ! ‘\\\ 1 I I ' I 'I \ I l J ! \f/ I I I m , {I / I I I II I j I I ’ ' “I I j \ I . H I .49 Z ‘.‘~ I 4 I ‘r. I g . N ,_ " C r ‘ ‘ \