Dating the Moraines and Recession of Athabasca and Dome Glaciers, Alberta, Canada

Transcription

1 Arctic and Alpine Research ISSN: (Print) (Online) Journal homepage: Dating the Moraines and Recession of Athabasca and Dome Glaciers, Alberta, Canada B. H. Luckman To cite this article: B. H. Luckman (1988) Dating the Moraines and Recession of Athabasca and Dome Glaciers, Alberta, Canada, Arctic and Alpine Research, 20:1, To link to this article: Copyright 1988, Regents of the University of Colorado Published online: 07 May Submit your article to this journal Article views: 162 Citing articles: 3 View citing articles Full Terms & Conditions of access and use can be found at

2 Arctic and Alpine Research, Vol. 20, No.1, 1988, pp DATING THE MORAINES AND RECESSION OF ATHABASCA AND DOME GLACIERS, ALBERTA, CANADA B. H. LUCKMAN Department of Geography, University of J#?stern Ontario London, Ontario N6A 5C2, Canada ABSTRACT Little Ice Age moraines at Athabasca and Dome glaciers, Alberta, Canada, were dated using dendrogeomorphic techniques and old photographs. Ice-scarred or tilted trees provide precise dates for the Little Ice Age maximum advance of 1843/44 (Athabasca) and 1846 (Dome). Moraine fragments from an earlier advance of the Athabasca Glacier were formed in Ecesis periods on these moraines are between 40 and 60 yr. Most trees colonizing these moraines date from periods of warmer summers in the present century. Dated ice-front positions and landform development between 1906 and 1981 are presented and discussed. Maximum recession rates for both glaciers occurred between ca and 1960 but have decreased markedly in the last two decades. INTRODUCTION The Athabasca Glacier is the most frequently visited and accessible glacier in Canada. It has been the locale for many glaciological studies (Ommanney, 1976; Kucera and Henoch, 1978) but, apart from Heusser's pioneer work (Heusser, 1956) and a number of specialized studies (e.g., Welch, 1970; Mills, 1977; Gilbert and Shaw, 1981), no detailed study has been carried out on the chronology of landform development in the glacier forefield. This paper will present an integrated review ofavailable material and new dendrogeomorphic data on the timing of glacial events in the forefields of the Athabasca and adjacent Dome glaciers during the Little Ice Age. It will also discuss the origin and developmental sequence of selected landforms in these areas. Dome and Athabasca are outlet valley glaciers that flow northeastwards from the Columbia Icefield (Figure 1). Although the glaciers have similar lengths, there are significant differences in their dynamics and surface debris cover. The Athabasca Glacier is fed by a staircase ofthree icefalls and the glacier tongue is between 250 and 320 m thick (Paterson and Savage, 1963). Only the eastern third of the glacier terminus has significant debris cover. Dome Glacier is formed by the coelescence of three smaller glaciers that spill over bedrock cliffs, 700 to 1000 m high, into the Dome valley. The medial moraines between these ice-tongues merge downvalley and the surface of the lower glacier tongue is completely debris-covered with incised supraglacial drainage. Contrasts between the glacier snouts emphasize some of these differences. The snout of Dome Glacier is smooth and completely debris covered: the primary controls of its morphology are the type and rate of ablation processes. Most of the Athabasca snout is debris free, has significant areas of erevassing, and has ice velocities of about 15 m yr" (Kucera, 1981). During the Little Ice Age these two glaciers coalesced and spread across the valley of the Sunwapta River, blocking drainage from the tributary valley ofwilcox Creek (Figure 2). DATING THE MORAINES OF ATHABASCA AND DOME GLACIERS Athabasca and Dome glaciers are typical of many glaciers in the Canadian Rockies in that their maximum 40 / ARCTIC AND ALPINE RESEARCH Holocene extent occurred during the Little Ice Age or Cavell Advance (Luckman and Osborn, 1979; Osborn 1988, Regents of the University of Colorado

3 117"15'W CSi a t i o n a I P a r~ Human Modified Surfaces melres kilometr es FIGURE I. The Columbia Icefield and environs. FIGURE 2. Place names and human modified areas at the Athabasca Glacier forefield. and Luckman, in press). Deposits of the earlier Crowfoot Advance have been identified immediately beyond the Little Ice Age moraines but have not been precisely dated. The general outline of the Holocene history is known from previous studies that are briefly summarized below. PREVIOUS STUDIES Heusser (1956) visited the Athabasca Glacier in 1953 with Field (Field and Heusser, 1954) and carried out a program of tree-ring dating of the moraines as part of a regional study of glacier fluctuations in the Canadian Rockies (12 glaciers were studied). He identified two periods of glacier advance: ca (based on a tilted tree, hereinafter referred to as the "Heusser log") and in the mid-19th century. The date for the earliest recession from the second advance was established as 1841 (see below), and successive recessional moraines were dated at 1900, 1908, 1925, and 1935 based on tree cores and old photographs, many of which are published in Kite and Reid (1977). Annual or biennial surveys of the Athabasca ice front were carried out by the Water Survey of Canada (WS.c.) from 1945 to 1980 and are summarized in WS.c. (1982). However, these surveys, like the photogrammetric maps produced between 1959 and 1979 (e.g., Reid and Charbonneau, 1979), only map the boundary of the "clean" ice and the debris-covered eastern margin is excluded. Some of the earlier ice limits are shown on a map in Mercer (1958, drawn by Field). TECHNIQUES USED IN THE PRESENT STUDY Reconstructing the recent history of the Athabasca forefield presents a number of unique problems due to the "popularity" of this area. There has been considerable modification of the landscape due to tourist-related activities (Figure 2) that have removed or significantly altered several landforms. In addition, published maps showing former ice-front positions (e.g., Welch, 1967; Kucera, 1972; Parks Canada, 1981) have errors or are inconsistent with other sources. Therefore, in this study, maps of the forefield or ice-front positions were prepared from primary data sources, i.e., dated oblique photographs, aerial photographs, and the original WS.c. maps. The landform map of the Athabasca Glacier forefield (Figure 3) is based on detailed photo-interpretation with extensive field checking. In addition, significant landforms that had been removed by human activity (moraines, channels, and even a small pond) have been included, on the basis of aerial photography taken in 1959, to produce a relatively complete picture of the suite of landforms in this area. Field investigations included an extensive tree-coring program on the moraines in 1980/81. Tree cover on the moraines is sparse but over 100 trees were cored. This site is exceptionally rich in dendrogeomorphic evidence for ice-push or scarring of trees along the trim line and several snags and a living tree were sectioned from the outer moraine areas. B. H. LUCKMAN / 41

4 THE OUTER MORAINES OF THE ATHABASCA GLACIER The morphological expression of the outer limit of the Athabasca Glacier is extremely variable ranging from the classic, towering, lateral moraines to vegetational trimlines without associated moraines. Sections in the east lateral moraine suggest that, despite its morphological simplicity, the feature is a complex, multi-age feature that '>:~~:::::::::::::::::::".. <.»...:::::-,.,<>...:::... o ;.. '====;====== metres approx,. Map ij'ased ona (4August 1.981) withsome detail reconstructed from A c.,::::::. (1 September 1959) <,.; o Till, Ground Moraine I Fluted Till fz:~iie~ Ablation Till L Lacustrine Veneer FluviailFluvioglacial Deposit S Snowbank ~ Bedrock ~ Undercut Slope ~ Patchy Till Veneer over Bedrock r Moraine Ridge Former Channel " 1/ if :: Ii /j Dirty Ice Athabasca Glacier / FIGURE 3. Landforms/surficial deposits map of the Athabasca Glacier forefield. (Letters refer to localities discussed in the text.) Reproduced with permission of Parks Canada. 42 / ARCTIC AND ALPINE RESEARCH

5 may incorporate deposits of several Holocene glacial events (see, for example, Osborn, 1986; Ryder and Thomson, 1986). Several authors (e.g., Jennings, 1951; Kucera, 1981) have suggested that these high lateral moraines are ice cored. Examination of historic photographs and present sections suggests that this is not the case. The earliest photography (e.g., Cautley and Wheeler, 1924; frontispiece) indicates that the glacier ice lying against the inner margins of both lateral moraines had a thick debris cover. Differential ablation preserved some of this debriscovered ice against the proximal slope of the lateral moraine and sections of this ice have been exposed by slumping from time to time. DATING THE LATERAL MORAINES The conspicuous lateral moraines usually have a single moraine crest and are treeless. However, in the western margin of the forefield, near the former confluence with Dome, several distinct ridges diverge from the crest as the moraine is draped over a bedrock knoll. The higher, inner ridges are almost treeless but Heusser (1956) reported the oldest tree from the low, outer moraine had a pith date of This tree was too rotten to yield a good core in 1981, but cores from other trees confirm. that the moraine dates from the 18th century (Table I). During the 19th century advance, drainage from the glacier cut through the older moraine and flowed between the two laterals (A, Figure 3). Immediately north of this channel, Heusser sectioned a tree that was tilted by the glacier in 1714 (Figure 4). Examination of the cores from an adjacent tree, 5 m "down-ice" revealed an internal scar between the 1713 and 1714 rings at the base of the tree on the side facing the glacier. The corroborative evidence from these trees provides a precise date for the emplacement of the outer lateral moraine at this site. The oldest tree growing on the inner moraines indicates they date from the nineteenth century advance (Table I). DATING FEATURES IN THE TERMINAL AREA At its Little Ice Age maximum position the Athabasca Glacier advanced against the lower slopes of Mount Wilcox, blocking drainage from Sunwapta Pass. Drainage flowed westwards along the ice-margin to the Sunwapta flats from a point immediately west of the highest point of the glacier margin (B, Figure 3). Small, arcuate moraine ridges, I to 3 m high, delimit the eastern margin of this snout but, westwards, the trimline follows the top of a 2 to 5 m high bluff that marks the northern bank of a former ice-marginal stream. In places segments of terminal moraine occur along the southern flank of this channel and a small, isolated terrace fragment occurs at the downstream end (C, Figure 3). Farther west all traces of the terminal position have been completely removed by proglacial drainage from the Athabasca Glacier north of the Sunwapta River. I Well-developed, treeless moraine fragments mark the downvalley extent of the Athabasca Glacier in the zone where the two glaciers coalesced (Figure 3). The dating control for the terminal moraines comes '.JI'TH J611 ;;.~-;{'~, TREE TILTED ~ '?~. -V IN 1714 Y I n ATHABA o GLACJE~ ~ o z; 0> 7'.,. 7- ' (' -.l-... 1:- FIGURE 4. Cross section of the "Heusser log. " This spruce tree was tilted by the early 18th century advance of the Athabasca Glacier. Note the marked reduction in ring widths between 1820 and 1900 corresponding to the 19th century advance of the glacier to within a few meters of the tree. The sample was cut by C. J. Heus ser in from the lower slopes of Mount Wilcox: the few scattered trees on the terminal landforms south of the Icefields Parkway all date from the present century. A mid-18th century date from the only tree (A81131, Table 1) growing on the most westerly moraine fragment (0, Figure 3) indicates this moraine is a remnant from the 18th century advance. Cores from trees on other moraine segments flanking this channel indicate these moraines date from the 19th century. Heusser (1956) inferred that ice withdrew from these moraines ca based on the oldest tree he cored on the cutbank of the former icemarginal channel. He dated the 19th-century advance to 1841 based on "the oldest tree on the outwash north (actually west) of the chalet" (Heusser, pers. comm., 1981). A slightly older date was determined for this tree in 1981 (A8149) but other trees on the terrace are much younger (Table 1). One of the trees on the slope backing the terrace dates from 1785 (A8153, Table 1)and the only other 18th-century tree on that slope (A81132) lies immediately upslope of the 18th-century moraine fragment. The dates from these two trees may therefore be used to infer an 18th-century age for the terrace remnant. As the dating control and morphological relationships for this feature are poorly defined and more precise dating is 'The present course of the Sunwapta River on the flats is artificial and dates from the construction of the Icefield Parkway in Formerly the river flowed against the bluff on the north side of the Dome-Athabasca outwash. B. H. LUCKMAN / 43

6 available elsewhere (see below) this evidence is not used to date the 19th-century advance. Heusser (1956) determined younger recessional dates from trees on moraine fragments north of the old road (V, Table I). The 1981 results are reported in four groups. Most of these moraine areas have scattered tree cover which does not appear to provide closely limiting dates (lb, II, III, and IV, Table 1). The fourth sample area (Ia) is a grove (ca. 10 m long) of alpine fir rooted at or near the base of the distal face of a small terminal moraine, only ca. 1 to 1.5 m high, about 20 m north of the old highway. These trees were significantly older than trees on other, equivalent, moraine fragments, and several had deformed lower trunks that bowed away from the moraine. This configuration suggested that the trees may have been deformed during moraine emplacement. The oldest of these trees was sectioned in 1984 and revealed that the main trunk had developed as a "leader" from an earlier stem that had been knocked down, twisted and broken during emplacement of the moraine (Figure 5). The original stem survived but lost apical dominance and was completely enclosed in callous growth from the new trunk. The oldest pith date obtained from the new leader was 1850, about 1.5 em above the point it branched from the old stem. As the leader grew vertically at ca. 1.6 em yr' between 1850 and 1860, it must have begun development in the late 1840s. Examination of the basal section of the tree indicated a pith date ca for the original stem. The early rings of this tree are very narrow (ca. 95 rings in 10 mm) but following the 1843 or 1844 ring year there is an abrupt (over 50070) decrease in ring width and the next 10 or II annual rings occupy only 0.65 mm (some of these rings are only two or three cells wide). After 1855 the rings widen and ring TABLE 1 Tree-ring dates for landforms in the terminal zone ofthe Athabasca and Dome forefields" Athabasca Glacier Terminal Moraine Segments I: Moraine north of old highway (east of B, Figure 3) (a) Abies grove IS5Sp (AS I77) ISS4p (ASI7S) (b) Other IS93p (ASI66) IS61p (ASI76) < ISS7 (ASI79) IS7Sp (AS 1SO) ISS9p (ASlS2) 1907p (ASI63) 1907p (ASI72) II: Moraine south of old highway (at cesspool) IS94p (ASI116) 1916p (ASI120) 1924p (ASI114) III: Moraine segments along cutbank (north of E, Figure 3) < 177S (ASI131) 1915p (ASI126) 1921p (ASI122) IV: Moraine adjacent to present highway (west of E, Figure 3) 1911p (ASI45) < 1914 (ASI41) 1920p (ASI4S) V: Heusser's oldest trees (equivalent to areas I and II, above) IS71 IS73 ISS3 ~ ISSI (ASlSl) 1925p (AS1124) _._---~..._---- *IS53p (ASI49) 1934p (ASI54) Outwash (south of C, Figure 3) *IS5S (Heusser) 1902p (ASI5S) < 1915 (ASI55) Undercut Bluff Bordering Former Ice-marginal Stream (east of 0, Figure 3) *173S (Heusser) *173Sp (ASI132) -e 17S5 (ASI53) IS23p (ASI127) <IS45 (AS53) IS61p (ASI12S) ISSSp (ASI50) IS91p (ASI123) 1761 (Heusser) < ISS5 (AOSI14) ISS7 (Heusser) 1920p (OSI20) Outermost West Lateral Moraine < IS16 (AOSI15) IS5Sp (AOSI13) < IS60 (AOSI17) Other Laterals Dome Glacier Terminal Moraine IS9Sp (OSI7) 1905p (OSI5). 1_92_S-,-pS_0_Sl_1). _ Eastern Lateral 1923p (OSI21) < 1937 (OSI22) < 1937 (OSI23) aan trees were cored as close to the root crown as possible and the dates given are earliest ring dates. 17 yr have been added to Heusser's published dates to allow for his 17-yr ecesis estimates (Heusser, 1956; pers. comm., 19SI). ( )= sample number; p = pith date; < = within 5 yr of pith; ~ = within 10 yr of pith; *=same tree. 44 / ARCTIC AND ALPINE RESEARCH

7 Pith ~resapprox. L~:::J TIll, Ground Moraine.- Fluted Till [IE; Ablation TIll Lacustrine Veneer r-1 Fluviallfluvioglacial,_.-~ Deposit S Snowbank ~Bedrock r' Undercut Slope ~" Patchy TillVeneer b= over Bedrock J' Moraine Ridge Lower Branches FIGURE 5. Plan sketch of tree A84-1 from the west, outer terminal moraine, Athabasca Glacier. widths of 1 to 2 mm are common in the late 1800s. This ring-width record is interpreted as indicating a prolonged period of growth suppression of the young tree, probably as a result of competition from adjacent saplings. (Similar conditions prevail in the present forest, ca. 1 km to the west, just beyond the trimline [see Luckman et ai., 1984].) Major injury to the tree by the glacier in 1843 or 1844 resulted in the sequence of very narrow rings for a period of about 10 yr, after which the tree gradually recovered. Over the last century tree growth has been quite vigorous due to the sheltered microhabitat of the tree and removal of competitors by the glacier. Rotted stem or branch sections from trees that were killed occur at the base of the moraine slope in at least two localities (Figure 5). On the basis of topographic setting and trunk morphologies it seems likely that many FIGURE 6. Landforms/surficial deposits map of the Dome Glacier forefield and dated ice-front positions (for sources see Table 4). Reproduced with permission of Parks Canada. of the oldest trees in this grove are survivors from other damaged trees, i.e., they did not develop from seedlings on a freshly exposed moraine surface. OUTER MORAINES OF DOME GLACIER Extensive areas of the Dome Glacier forefield are occupied by Dryas-covered outwash flats originating from the former proglacial drainage of Dome or the western flank of Athabasca (Figure 6). Large sections of the terminal moraine have been destroyed and the remnants have only a sparse tree cover. All significant trees on or just inside the moraines west of Dome Creek were cored in The oldest tree was a spruce, 2 em in basal diameter, growing at the foot of the proximal slope of the outer moraine. Five other trees from the moraines (D815 was on outwash) provided pith dates between 1928 and 1948 (Table 1). Heusser's 1870 estimate for the maximum advance of Dome was based on a tree with a pith date of 1887 that grew on the moraine, probably east of the creek. The four largest trees on the distal slope of the east lateral moraine yielded pith dates of between 1920 and This ice limit can be precisely dated from a unique assemblage of snags in a small area on the western lateral at the point where it joins the outwash (Figures 7, 8, 9). At this locality the glacier was expanding laterally and advancing obliquely across a steep bedrock slope, draping a veneer of till (often less than 1 m) over the surface. living trees adjacent to this trimline are at least 200 to 300 B. H. LUCKMAN / 45

8 yr old. Three standing snags of similar age (200 to 269 yrs, Table 2) occurred just within the Little Ice Age limits. Two of these trees were rooted on a 20 0 bedrock slope less than 3 m horizontally and 1 m vertically below the till limit marking the trimline at this point. Although these trees were killed or damaged by the glacier they were too large (or close to the glacial limit) to be removed by the ice. By 1981 tree 0814 had fallen but the basal shape of the snag clearly matches one of the two adjacent rooted stumps (0814 and 0815, Table 2). The 0814 snag has a large bark scar which formerly faced the glacier, has maximum dimensions of 1 m vertically and 0.25 m horizontally, and begins 0.7 m above the root crown. Using tree-ring densitometric data this tree has been cross-dated with the Heusser log and lived between 1689 and The outermost ring on the scar face is However, several rings have been lost by surface weathering of the scar and a scarring date of 1846 was determined by ring counts where the scar tissue had been protected by callous growth. This date is confirmed by a reaction wood series initiated in 1847 on the opposite side of the tree indicating that it was tilted away from the glacier at that time. There is no evidence to indicate how long the ice was in contact with the tree. The section shown in Figure 9 was cut from the uppermost section of the scar and shows very little evidence of other damage, even though the lower part of the tree must have been surrounded by ice. Tree 0813 does not show external damage but, growing on a slope, it does have a reaction wood series on one radius and therefore is unsuitable for cross-dating studies. Tree 0812 was situated on a small gravel bar or island in a former proglacial channel (Figure 8). It appears to be located about 5 to 10 m within the Little Ice Age maxi- FIGURE 7. Terminal moraine and west lateral of Dome Glacier, July The arcuate line of boulders marks the terminal position on the outwash, and the growth sites of 0812 and 0814 are shown at A and B, respect ively. The log being excavated, foreground, is 0841, killed in FIGURE 8. Trimline of the west lateral moraine of the Dome Glacier, July 1981, showing the position of samples 0812, 0813, and The position of the cros s section for 0814 is indicated by the arrow. 46 / ARCTIC AND ALPINE RESEARCH

9 --~- -~- ~ _.._- -_.._---_.-----~. _----~ ~ -~----- mum as determined by projecting the alignment of adjacent, preserved moraine fragments. This tree was crossdated with the Heusser log and several older living trees at the site using ring-width data and has a pith date of As many as 40 rings have been removed from the snag surface facing the glacier which is exposed to prevailing regional and topographic winds. The outer ring on the north side of the tree is ca This tree has no external damage or significant growth irregularities; it may have died naturally or been slowly killed by the prograding outwash shortly after Several tree-fragments protrude from the lateral mo- TABLE 2 Snags sampled on the western margin of Dome Glacier --_.- Basal Sample diam. Height Ring 10 (ern) (m) count Comments ". "~ S 7 > Lying on outwash, 20 m outside limit, rootstock attached. 0812S 21 > Standing snag, lived from , topped at 6 m; in outwash ca. 5 to 10 m within projected LIA limit. 0813S II Standing snag, ca. 2 m from 0814S, grew on slope 4 m inside LIA limit. 0814S ca. 28 > Picea, fallen snag, lived from Rootstock 2.5 m inside LIA limit on tillveneered slope. Scar facing ice 0.7 to 1.7 m above base. 0815S 20 Rotten, rooted snag, I m upslope of 0814S (1.5 m inside limit). 0841S 29 > Picea, lived from , root present lying on bedrock, base covered with < 1.0 m of till. 2.5 m from LIA limit, 50 to 70 m up-ice of 0814S. 0842S 6 76 Topmost 1.5 m of Larix trunk, sticking out of moraine crest: does not cross-date. 0843S ca. 30? >200 Partial section 4 to 5 cm across from outer base of stem? Rings tight and contorted., I ri r r t: I t. ~ (- -.r----~, / ~... OUTER MOST-A- RING 1902 I 1 r 1 / I PITH ,-...p~ SCAR 1846 o C'-p., ~o~. 'P..r. ~ (''''1' ;J'""' _/ I I ~~l ~~/ l. ; 1 J I /, I FIGURE 9. Annotated cross section of tree 0814, Picea engelmannii, scarred by Dome Glacier in B. H. LUCKMAN / 47

10 raine and till 50 to 100 m upslope of D814 (Table 2). The largest of these (D841, Figure 7) was killed in 1838(crossdate based on ring-width and densitometric data). As the tree was not in growth position it could have been transported by the glacier prior to burial. Some snags lying on the outwash of Dome Glacier (e.g., D811) were probably washed out of the till as they contain complacent ring records of equivalent length to the oldest trees presently growing in the forefield. The 1846 date provides precise chronological control for the Little Ice Age maximum advance at Dome Glacier. There is no morphological evidence for an 18th century event at this glacier and the ring-width series in D814 and D812 show no clear evidence of a period of significant growth suppression in the early 18th century similar to that seen in ring-width series from trees in the area between the Dome and Athabasca glaciers (Johnson, 1981 ). ECESIS ESTIMATES The precise dates for moraine formation at Athabasca and Dome glaciers enable direct determination of the period between moraine formation and the earliest dates obtained from trees growing on the moraine surface. These estimates of ca. 40 to 60 yr (Table 3) are much greater than the 17-yrecesis period used by Heusser (1956) or the 10 to 20 yr commonly reported for the Canadian Rockies (Luckman, 1986a, Table 1). However, these differences might be anticipated from the small number of trees on these moraines vis-a-vis the sites studied by those authors. Kearney (1982) has demonstrated that seedling establishment at treeline on the slopes of Mount Wilcox, overlooking these moraines, has been limited to periods of warmer summers in the 1940s and 1960s. The general range of dates obtained from trees on the Athabasca Dome moraines and the clustering of many tree-ring dates in the 1930s and 1940s(Table 1 shows only the oldest trees sampled) suggest that there is a strong climatic control on the colonization of moraines at these sites. Although the moraines are about 100 to 150 m below treeline, severe local microclimatic effects (e.g., katabatic drainage from the Icefield) make them marginal sites for colonization under present conditions. Effects such as this must be taken into account when estimating the age of moraines with a sparse tree cover because periods of colonization may be related to periods of warmer summers rather than the period of elapsed time since moraine stabilization. DATING OF LANDFORMS WITHIN THE LITTLE ICE AGE LIMIT PRE-LITTLE ICE AGE FEATURES Within the Little Ice Age limit there are some landforms that appear, on morphological grounds, to predate the Little Ice Age deposits. Two rounded ridges, 10 to 20 m high, 50 to 100 m wide, and 200 to 300 m in length, flank the area of exposed bedrock between the Icefields Parkway and the terminal position of the Athabasca Glacier. These ridges trend northwest-southeast and the smaller recessional/terminal moraines are clearly superimposed over the pre-existing ridge surface. The alignment and position of these major ridges suggests they were moraines (surface exposures indicate they are composed of till) smoothed by the overriding glacier. As these two ridges are approximately orthogonal to the retreating ice front they were significant controls of ice-marginal drainage systems during recent deglaciation. MORAINES AND MELTWATER CHANNEL SYSTEMS The Athabasca Glacier blocked drainage from Sunwapta Pass and Mount Athabasca and during glacier recession a number of channel systems were cut at or under the ice-margin before the present drainage was established. Several major changes in the proglacial stream outlets from the snout also occurred. It is not pos- TABLE 3 Ecesis estimates from the outer moraines ofathabasca and Dome glaciers Locality Date of moraine Pith date of oldest tree Ecesis estimate Terminal, Athabasca Terminal, Dome Lateral, Athabasca 2nd lateral, Athabasca Oldest terminal, Athabasca Eastern lateral, Dome 1843 or 44 (A841S) 1846 (D814S) 1714 (Heusser log) 1843 or 1844 a 1714 b 1846 c 1893 (A8166) 1887 (Heusser) 1761 (Heusser) <1885 (AD8114) 1778 (A81131) 1920 (D8120) ca. 50 ca. 41 ca. 47 ca. 41 ca. 64 ca. 74 aassumes second lateral = to terminal. bassumes downvalley terminal = to Heusser log moraine. cassume = to terminal. 48 / ARCTIC AND ALPINE RESEARCH

11 sible to date these features precisely by tree rings or lichenometry (see Osborn and Taylor, 1975; Luckman, 1977). Fortunately, good documentary sources are available, particularly between ca and 1924 and after 1945, from which a detailed chronology of recession can be established (see Figures 6 and 10, Table 4). The most conspicuous terminal moraine sequences of the Athabasca forefield were created by readvances between 1843 and Except for the innermost moraine, dated ca from the earliest photograph (see Figure 10), only relative ages can be ascribed to these features. However, on the basis of better-dated moraine systems elsewhere in the Rockies (Luckman and Osborn, 1979), they probably date from the period. For most ofthis time drainage flowing westwards along the ice front disappeared into the ice south of the long pond in the forefield (see Schaffer, 1908, and frontispiece). The absence of other channel remnants suggests that this submarginal drainage occupied the present course of Wilcox Creek. This was certainly the case after 1917(frontispiece) but earlier the lower outlet was the well-defined channel at the western margin of the exposed bedrock area (E, Figure 3). This subglacial stream course may account for the absence of till from the large bedrock area immediately upslope. The subglacial channel walls could intercept the debris-rich basal ice which would ablate into and be evacuated by the stream while relatively clean ice continued to flow over the stream towards the terminus upslope. Moraine accumulations occur at the 1917, ca. 1923and 1938 ice-front positions (Figure 10). Between ca and 1930/31 the ice-front position is clearly delimited by a series of well-preserved annual push-moraines, rarely more than 1 m high, that extend continuously for over almost 100 m across the gentle slope south of Wilcox Creek (Figure 3). These moraines are spaced at 10 to 30 m intervals and well-preserved flutes extend onto the proximal moraine slope. The formation of similar moraines at the present ice front between ca and 1980 has been documented by Kucera (1972, 1981) and Welch (1967). A thin veneer of lacustrine deposits and a channel sequence cutting across the slope west of these moraines indicates small temporary ice-marginal ponds in this area between ca and By 1938 this local ponding had ceased and the stream occupied a broad, braided channel flowing across the base of the slope to join a proglacial stream, issuing from beneath the ice along the line of the present Sunwapta River (Figures 3 and 10). Between 1935 and 1955 a sequence of small channels were cut by streams flowing westwards along the ice-margin into Sunwapta Lake. The course of the present icemarginal stream was established in The western third of the former Athabasca Glacier snout was debris covered and the suite of landforms in this area is quite distinctive. A fairly uniform blanket of 0.2 to 0.5 m of dark, angular ablation till covers the area within the 1919 ice-front position although fluting of the underlying basal till is visible in a few places. Recession of this ice was by a combination of slow downwasting and accelerated recession of exposed ice-cliffs generated by meltwater erosion (Figure 10 and Luckman, 1986c). The western margin of the glacier was the source of a number of pro glacial streams and head ward expansion of this drainage between 1919 and ca. 1959created a large gravel spread (Figure 3), flanked to the east and south by an active ice cliff. After the initiation of Sunwapta TABLE 4 Primary sources for Figures 6 and 10 Date Photographer Identification Comments" 1906 or 1907 Schaffer ACR NG20-92 b, #4 Lot 1 Athabasca and Dome from Wilcox (see Luckman, 1986b) 1907 or 1908 Schaffer ACR NGIO-92 Schaffer (1908, 1911) 1917 Boundary Commission Panoramas and closeup of Athabasca (See Cautley and Wheeler, 1924: 5) 1919 Boundary Commission Panoramas of Athabasca and Dome Sept Field Athabasca Glacier Field (1949; pers. comm., 1986) 1923 Thorrington View Athabasca Toe (Loaned by Field, 1986) 1924 Harmon ACR NA E. flank Athabasca 10/7/1938 NAPU A First aerial photography ca Harmon ACR NA E. flank Athabasca ca (very dark) 19/9/1948 NAPL Al Air photos 1954 Map, Plate 7 in Davies et al. (1970) 1959 NAPL A Scale 1: NAPL A NAPL A NAPL A "Ful] details, discussion, and other sources are given in Luckman (1986b, 1986c). "Whyte Museum of the Canadian Rockies, catalogue number. CNAPL = National Air Photo Library, Ottawa. B. H. LUCKMAN / 49

12 Lake a second, parallel, ice cliff developed along the western shore of the lake and, during the 1950s, up to 20 to 30 m of clean ice were exposed along the lake. Accelerated ablation at this cliff line rapidly reduced the extent of the buried ice mass and, between 1962 and 1964, drainage from the western flank of the glacier broke through the diminished ice mass and flowed directly into the lake. By 1968 most of this drainage had transferred to the present ice-marginal stream and there was no evidence of extensive buried glacier ice north of the original breakthrough (marked by a small flow in 1968, see Figure 10). Morphological evidence suggests that, as the Athabasca Glacier receded, its front assumed a bilobate form, and the main proglacial drainage issued from the contact between the debris-covered and clean ice. From at least 1919 to 1938 the main proglacial stream discharge via the sinuous abandoned channel ("west" channel) that marks the ablation till/fluted till boundary in the forefield. Two earlier outlets are visible to the west (X and Y, Figure 3) at the intersection of moraines from these former ice lobes. The initiation of Sunwapta Lake cannot be docu- FIGURE 10. Dated ice-front positions of Athabasca Glacier (for sources see Table 4). Photograph reproduced courtesy National Air Photo Library, Ottawa, and Parks Canada. 50 / ARCTIC AND ALPINE RESEARCH

13 mented directly. Both "west" channel and the present river course had proglacial sources in 1938 with most of the flow in "west" channel. Examination of available photographic (Table 4, particularly NA ) and morphologic evidence suggests that the main flow switched to the lower, central channel in 1939 or 1940 (Luckman, 1986c). The first WS.c. survey in 1945 (Davies et al., 1966) indicated that the lake was about 120 m wide at that time and extrapolation of the 1945 to 1949 recession rates (WS.c., 1982) would indicate that the north shore of the lake became ice free in 1940 or Transfer of all the lake drainage to this lowest outlet took place between 1940 and The lake grew rapidly to reach its maximum area ca and has subsequently decreased in size due to delta formation and, by 1986, the southern basin was almost completely filled (Figure 10; Mathews, 1964; Gilbert and Shaw, 1981). The landscape of the Dome Glacier forefield is less complex than that of the Athabasca. Glacier recession has been principally by downwasting of the debriscovered snout with accelerated ablation at ice-cliffs generated by proglacial drainage trimming the snout. Arcuate moraine segments can be identified at the 1906 and 1919 ice-front positions but older moraine and till surfaces have been extensively modified by proglacial outwash. The present stream course was adopted between 1938 and 1948 when the braided channel shifted about 100 m east of its former course. The bedrock rise to the present snout is thinly mantled with ablation till and contains few discrete landforms. Since about 1960 the snout has been fronted by a small outwash plain that is undergoing active sedimentation and kettling. Proglacial drainage emerges from a number of siphons in this outwash up to 100 m beyond the ice-cliff marking the apparent glacier snout and indicates that the outwash still contains considerable buried ice. Over the last 10 to 15 yr two main proglacial source areas have been evident on opposite sides of the valley. Considerable switching of flow between these outlets (Table 5) indicates that the glacier-outwash hydrological system is quite complex. RATES OF ICE-FRONT RECESSION Quantitative estimates of the rate of ice-front recession for both glaciers are given in Table 6. Estimates for the Athabasca Glacier only refer to the eastern half of the glacier and there are obvious differences in the accuracy and measurement techniques on which these estimates are based. Nevertheless, they are probably the most complete yet published for glaciers in the Canadian Rockies and provide a comprehensive picture of changes over the last 70 yr. These data appear similar to, and therefore may be considered representative of, other glaciers in this area (see McPherson and Gardner, 1969; Gardner, 1972; WS.C., 1982). Although the Athabasca Glacier changed from a partially calving snout to a totally land-based snout in the early 1960s, the decrease in the rate of frontal recession in the 1960s and 1970sis also seen at Dome Glacier, Saskatchewan (WS.C., 1982) and other glaciers. Distribution of Flow" TABLE 5 Changes in flow dominance on the upper outwash of Dome Glacier Western Eastern channel channel Date (Ufo) (%) Source 1/9/ A b 3118/ A b 22/9/ A b 29/6/68? Most Visit 7/7/ A /8/77 99 Dry? CN3724 b 20/7/ Visit 7/8/79 95 A b 19/6/ (Photo from west lateral) 4/8/ A b 27/7/ Visit 24/9/ Visit 17/7/ Visit Comment East channel starts from ice at present lip of outwash. East siphon in ice. Considerable flow on west flank, large ice cave at head. Water bubbling up from under east ice margin. East site close to present valley side source, west has ice-cliff but low, clean flow. West ice-cliff well developed. Huge east siphon, some siphon ponds on west side. Pond at east margin, turbid water from west siphon in gravels. West clean, east very turbid, well developed "ponds." Ditto, more discharge in west, major channel shift on east fan. Main turbid discharge from siphons on west, ca. 20 m beyond icecliff (see Luckman, 1986c). Bulk of discharge from beneath the center of the ice margin. "These are qualitative estimates based on the width of the channel on the photographs used. Drainage from the western channel began ca baerial photograph. B. H. LUCKMAN / 51

14 Some smaller glaciers from adjacent areas have undergone recent frontal advance (Gardner and Jones, 1985; Luckman et al., 1987) but there is no unequivocal evidence for advance at the Athabasca Glacier during this period. The spacing of the annual moraines at the snout shows a marked reduction from about 18 m yr" between the moraines (Welch, 1967; Figure 9) to less than 1.5 m yr' for the moraines and the 1977 and 1978 moraines are indistinguishable (Luckman, 1986b). However, these moraines only occur along a small section of the snout and the W.S.C. data, averaged over a wider area, show minima between 1970 and (The maps in Reid and Charbonneau [1979, 1981], derived photogrammetrically, indicate frontal advances of ca. 50 m in two sections of the ice-front after This conclusion is not supported by the detailed photographic records of Kucera [1972, 1981] or myself.) Since 1981 the thin central lobe of the glacier toe has been receding more rapidly up a steep bedrock slope. The general pattern of recession at Dome Glacier is similar to that at the Athabasca but is complicated by the changing shape of the snout, the absence of easily detectable markers in front of the glacier and the presence of buried ice. No field measurements have been made. CONCLUSIONS Despite the sparse tree cover and the absence of Rhizocarpon sp. lichens on the moraines, the forefields of Dome and Athabasca glaciers have yielded the most precise dating of Little Ice Age limits in the Canadian Rockies and a comprehensive picture of recession during the present century. Unfortunately, few other sites in the Canadian Rockies have such abundant (and varied) dendrogeomorphic and documentary evidence available. The outermost terminal moraines at both glaciers date from the 1840s, somewhat earlier than Heusser's (1956) estimates. However, at two small sites in the Athabasca forefield, moraine fragments from a locally more extensive eighteenth century advance (maximum 1714) are preserved. Local topographic controls inhibiting ice flow may also account for the preservation of three standing conifer snags within the lateral margin of the 19th century TABLE 6 Ice-front recession of Dome and Athabasca glaciers ca From To Distance" (m) Athabasca Glacier ca /8 1906/ ca. 200 ca ca ca (land) ca (lake) ca Rate m yr " From To b c 17.0 c Distance" (m) Rate m yr" Dome Glacier (estimated from Figure 6)d " 2.5-7" 100 e c e e aestimates for the central part of the tongue east of Sunwapta River based on Figures 3 and 10. "Values for 1945 onwards were measured by the Water Survey of Canada and refer only to the eastern (clean ice) part of the snout (see W.S.C., 1982). 'These figures seem anomalous, given the spacing of the annual moraines (see Luckman, 1986b, Figure 5). Possibly the snout was snow-covered during the 1962 survey. dthese are approximate given the detailed changes in configuration of the snout. "These distances are along the present stream channel. 52 / ARCTIC AND ALPINE RESEARCH

15 advance of Dome Glacier. Ecesis estimates for trees on these moraines are between 40 and 60 yr. The major periods of tree establishment on these sites are more probably linked to periods of warmer summers in the present century than any standard "ecesis period" since moraine stabilization because of the marginal climatic nature of these sites. Some trees at the glacier margin may survive glacier damage and produce new stems. After 100to 150 yr such trees may appear to be the first colonizers of the moraine surface and tree-ring dating could produce anomalously "old" ages for the moraine if standard ecesis estimates are used. Careful evaluation of site conditions and tree characteristics are necessary to avoid such errors. Prominent moraine systems were constructed by the Athabasca Glacier in the late 19th century but cannot be precisely dated. No spatially extensive moraine systems were formed in the present century although, locally, annual moraines were formed. Drainage from the eastern flank of the Athabasca Glacier occupied a variety of marginal and submarginal positions prior to deglaciation of the present course of Wilcox Creek ca Although ice-marginal channels and temporary lakes occurred along the upslope (eastern) margin, these streams drained under the northern margin of the glacier to reappear as major proglacial streams at the downvalley snout. The morphological record of recession at Dome Glacier is much simpler than that of the Athabasca; most of the glacial landforms have been replaced by extensive proglacial outwash gravels. During the 20th century, maximum ice-front recession rates (25 to 30 m yr") occurred in the 1940s and 1950s at both glaciers. In the last 20 yr rates of recession have decreased considerably, but no periods of net annual advance have been recorded at either glacier. Detailed reconstruction of recessional patterns at the Athabasca and other glaciers in the Canadian Rockies in the last 40 yr is possible because of the excellent documentary sources provided by several government agencies. Many of these programs have now been terminated. The lack of good periodic updates of aerial photography or continued measurement will severely curtail research in the future. ACKNOWLEDGMENTS This work has been supported by grants from the Natural Sciences and Engineering Research Council of Canada and contracts from Parks Canada, which are gratefully acknowledged. I thank the following: Fred Dalley, Gordon Frazer, Jim Hamilton, and David and Helen Luckman for assistance in the field; Dr. C. J. Heusser for providing his specimen for Figure 4; Dr. W. O. Field for provision of old photographs; Mr. E. Cavell, Whyte Museum of the Canadian Rockies, for assistance with their photographic archives; L. Jozsa, Forintek Canada Corporation, for densitometric work on several tree samples; National Air Photo Library and the Geodetic Survey of Canada for permission to reproduce Figure 10 and the frontispiece, respectively; Parks Canada for permission to reproduce Figures 2,3, and 6; G. O. Shields, Cartographic Section, U.WO. and Parks Canada Interpretive personnel, particularly Sue Wolff, Dave Biedermann, Kevin van Tighem, and Jim Todgsham for their assistance and support. REFERENCES CITED Cautley, R. W. and Wheeler, A. 0., 1924: Report ofthe Commission Appointed to Delimit the Boundary between the Provinces oj Alberta and British Columbia: Part II 1917 to 192/, From Kicking Horse Pass to Yellowhead Pass. Office of the Surveyor General, Ottawa. 157 pp. Davies, K. F., Froelich, C. R., Heinze, L., and Kerber, R., 1966: Surveys of glaciers on the eastern slope of the Rocky Mountains in Banff and Jasper National Parks. Water Resources Branch (Canada), Report, 2160-(66), Calgary. 30 pp. Davies, K. F., Warner, L. A., Anderson, J. E., and Dahl, B. M., 1970: Survey of glaciers on the eastern slopes of the Rocky Mountains in Banff and Jasper National Parks-1970 Report. Water Survey of Canada Division, Inland Waters Branch, Department of Energy, Mines and Resources, Calgary. 29 pp. Field, W.O., Jr., 1949: Glacier observations in the Canadian Rockies, Canadian Alpine Journal, 32: , 1986: Personal communication. Great Barrington, Massachusetts. Field, W. 0., Jr. and Heusser, C. J., 1954: Glacier and botanical studies in the Canadian Rockies, Canadian Alpine Journal, 37: Gardner, J. S., 1972: Recent glacier activity and some associated landforms in the Canadian Rocky Mountains. In Slaymaker, H. O. and McPherson, H. J. (eds.), Mountain Geomorphology. Vancouver: Tantalus Press, Gardner, J. S. and Jones, N. K., 1985: The Neoglacial advance of Boundary Glacier, Banff National Park. Canadian Journal oj Earth Sciences, 22: Gilbert, R. W. and Shaw, J., 1981: Sedimentation in proglacial Sunwapta Lake, Alberta. Canadian Journal oj Earth Sciences, 18: Heusser, C. J., 1956: Postglacial environments in the Canadian Rocky Mountains. Ecological Monographs, 36: , 1981: Personal communication. Department of Biology, New York University, Tuxedo, New York Jennings, A. H., 1951: Glacial geomorphology of the Sunwapta Pass area, Jasper National Park, Alberta, Canada. M.Sc. thesis, University of Iowa. 99 pp. Johnson, M., 1981: A dendroclimatological study at the Columbia Icefields, Jasper National Park. B.A. thesis, Department of Geography, University of Western Ontario. 108 pp. Kearney, M. S., 1982: Recent seedling establishment at timberline in Jasper National Park, Alberta. Canadian Journal oj Botany, 60: Kite, G. W. and Reid, L. A., 1977: Volumetric change of the Athabasca Glacier over the last 100 years. Journal ofhydrology, 32: Kucera, R. E., 1972: Probing the Athabasca Glacier. Vancouver: Evergreen Press. 32 pp. ---, 1981: Exploring the Columbia Icejield. Canmore, Al- B. H. LUCKMAN / 53

16 berta: High Country Press. 64 pp. Kucera, R. E. and Henoch, W. E. S., 1978: Glacier and landform features in the Columbia Icefield area, Banff and Jasper National Parks, Alberta, Canada. A study carried out for Parks Canada by Glaciology Division, Inland Waters Directorate, Environment Canada, March 1978, 131 pp. Luckman, B. H., 1977: Lichenometric dating of Holocene moraines at Mount Edith Cavell, Jasper, Alberta. Canadian Journal ofearth Sciences, 14: , 1986a: Reconstruction of Little Ice Age events in the Canadian Rockies. Geographie physique et Quaternaire, 40: , 1986b: Historical ice-front positions of the Athabasca Glacier, Report to Parks Canada, Contract KJP 03850, Feb pp. ---, 1986c: Landform development in the fore fields of Athabasca and Dome Glaciers. Report to Parks Canada, Contract KJP-02960, Dec. 1986; 190 pp. Luckman, B. H. and Osborn, G. D., 1979: Holocene glacier fluctuations in the middle Canadian Rocky Mountains. Quaternary Research, 11: Luckman, B. H., Jozsa, L. A., and Murphy, P. J., 1984: Living seven-hundred-year-old Picea engelmannii and Pinus albicaulis in the Canadian Rockies. Arctic and Alpine Research, 16: Luckman, B. H., Harding, K. A., and Hamilton, J. P., 1987: Recent glacier advances in the Premier Range, British Columbia. Canadian Journal of Earth Sciences, 28: Mathews, W. H., 1964: Sediment transport from Athabasca Glacier, Alberta. International Association for Scientific Hydrology, Publication, 65: McPherson, H. J. and Gardner, J., 1969: The development of glacial landforms in the vicinity of the Saskatchewan Glacier. Canadian Alpine Journal, 52: Mercer, J. H., 1958: Glaciers of the Canadian Rocky Mountains. In Field, W. 0., Jr. (ed.), Study ofmountain Glaciation in the Northern Hemisphere, Chapter 2, Part B, Western Canada and Arctic Canada. New York: Department of Exploration and Field Research, American Geographical Society. 23 pp. Mills, H. H., 1977: Differentiation of glacier environments by sediment characteristics: Athabasca Glacier, Alberta, Canada. Journal of Sedimentary Petrology, 47: Ommanney, C. S. L., 1976: Information relating to glaciers of the Columbia Icefield area, Banffand Jasper National Parks, Alberta. Glaciology Division Internal Report, April pp. Osborn, G. D., 1986: Lateral moraine stratigraphy and the late Neoglacial history of the Bugaboo Glacier, British Columbia. Quaternary Research, 26: Osborn, G. D. and Luckman, B. H. (in press): Holocene glacier fluctuations in the Canadian Cordillera (Alberta and British Columbia). Quaternary Science Reviews. Osborn, G. D. and Taylor, J., 1975: Lichenometry on calcareous substrates in the Canadian Rockies. Quaternary Research, 5: Parks Canada, 1981: Columbia Icefield 1:50,000 map (I.W.D. 1011) produced by Parks Canada and National Hydrological Research Institute, printed by Energy, Mines and Resources, Ottawa, Paterson, N. S. B. and Savage, J. c., 1963: Geometry and movement of the Athabasca Glacier. Journal ofgeophysical Research, 68: Reid, 1. A. and Charbonneau, J. O. G., 1979: Glacier surveys in Alberta Report Series No. 65, Inland Waters Directorate, Ottawa. 17 pp. ---, 1981: Glacier surveys in Alberta Report Series No. 69, Inland Waters Directorate, Ottawa. 19 pp. Ryder, J. M. and Thomson, B., 1986: Neoglaciation in the Southern Coast Mountains of British Columbia: Chronology for events prior to the late Neoglacial maximum. Canadian Journal ofearth Sciences, 23: Schaffer, M. T. S., 1908: Untrodden ways. Canadian Alpine Journal, 1: , 1911: Old Indian Trails of the Canadian Rockies. (Reproduced in A Hunter ofpeace, E. J. Hart (ed.). Banff: Whyte Foundation, pp.) Water Survey of Canada, 1982: Survey ofthe Athabasca and Saskatchewan Glaciers. Water Survey of Canada, Calgary District Office. 37 pp. Welch, D. M., 1967: Slope evolution on recessional moraines. M.Sc. thesis, Department of Geography, University of Alberta. 100 pp. ---, 1970: Substitution of space for time in a study of slope development. Journal of Geology, 78: Ms submitted April / ARCTIC AND ALPINE RESEARCH