Wiles, G.C., and Calkin, P.E., 1993. Neoglacial Fluctuations and Sedimentation of an Iceberg-Calving Glacier Resolved

NEOGLACIAL FLUCTUATIONS AND SEDIMENTATION OF AN ICEBERG-CALVING GLACIER RESOLVED WITH TREE RINGS (KENAI FJORDS NATIONAL PARK, ALASKA)

Gregory C. Wiles and Parker E. Calkin Department of Geology, State University of New York at Buffalo, 415 Fronczak Hall, Buffalo, NY 14260, U.S.A.

The McCarty Fjord iceberg-calving glacier and its land-terminating tributaries have fluctuated with some asynchroneity during the past two millennia. During advance, McCarty Glacier shed outwash along the fjord, and into ice-dammed tributaries inundating forests. A radioacarbon framework has revealed at least two major late Holocene glacial advances that occurred following a poorly documented expansion about 3600 BP. These two later advances are resolved within this fjord with tree-ring dating of in situ and transported tree trunks. The first event was an early medieval expansion of the McCarty trunk glacier beyond midfjord about 596 A.D.; this followed an interval of tree growth of at least 206 years. Tributary glaciers probably also advanced at this time. However, continuous tree-growth occurred in the distal (southern) tributary valleys during this advance, while northern tributaries were being dammed by the advancing trunk glacier. The down-valley extent of this expansion is thus constrained to a position within 12 km of the present contracted ice margin. The tree-ring refined chronology shows that a second McCarty ice expansion began in the 9th century, and reached midfjord by 900 A.D., and continued to advance through the Little Ice Age. In contrast, expansions of land-terminating glaciers here began after 1300 A.D. in concert with mountain glaciers worldwide. A log from a diamict cross-dated with a living tree-ring chronology shows that McCarty Glacier was advancing within a kilometer of its Little Ice Age maximum at 1790 A.D. Since about 1905 A.D., dramatic ice retreat has uncovered more than 20 km of McCarty Fjord.

INTRODUCTION chronology of McCarty Glacier and to compare this with that of local land-terminating glaciers. Objectives A record of late Holocene glacier fluctuations is Setting preserved in sediments at the foreflelds of tongues, Multiple glacial fluctuations and tectonic subsidence such as McCarty Glacier, which emanate from the of approximately 1 m per century have resulted in the Harding and the Grewingk-Yalik ice fields of the deeply incised McCarty trough stretching oceanward 60 southern Kenai Mountains (Fig. 1; Wiles and Calkin, km from the margin of McCarty outlet tongue. 1990). Preserved in these sediments are in situ and Mesozoic graywackes and slates that form the bedrock detrital trees representative of forests overrun by ice walls of the upper 20 km occupied by ice during during three intervals of late Holocene time. Our Holocene time, rise from sea level up to 1500 m investigations in this area have generally focused on altitude; they also produce a relatively uniform maxi- variations of land-terminating glaciers as proxies of mum depth of 270 m through this reach. Fjord walls climate change. These studies provide baseline data are interrupted by the valleys of the Dinglestadt outlet from which we can examine fluctuations of the fjord- tongue and by a set of contracted mountain glaciers in calving outlet tongues on the east flank of the moun- Delectable Bay (unofficial name) that were tributary to tains. McCarty Glacier during the Holocene (Fig. 2). Observations of the temperate fjord glaciers in The climate of the eastern mountain flank is mari- Alaska show that while in the calving condition, termini time. Moisture-laden Pacific air masses that move frequently undergo large-scale, slow advance and rapid westward from the Gulf of Alaska are intercepted by disintegration. These are often asynchronous with the icefield, producing an estimated 3-10 m year -1 of adjoining land-terminating tongues and with climatic precipitation (Rice, 1987) of which over 50-70% may fluctuations (Mann, 1986; Meier and Post, 1987). Such be snow (Clagett, 1988). The mean annual temperature fluctuations may be primarily related to water depth at from Seward (Fig. 1), located about 50 km northeast of the terminus, fjord configuration, and changes in McCarty Glacier, is 4.2°C (Ruffner and Bait, 1987). hypsometry with respect to equilibrium line altitudes Glacier tongues are fringed by a Sitka spruce and (Mercer, 1961; Post, 1975; Brown et al., 1982). mountain hemlock forest that have colonized the McCarty Fjord is remarkably uniform in both length deglaciated areas at lower elevations since 3000 BP and depth between the McCarty Glacier margin and (Ager, 1983). inside its Holocene terminal moraine (Fig. 2). In addition, confluent ice masses have been relatively FIELD DATA AND METHODS small. This geometry simplifies our objectives in the study reported here, which are to define the Holocene The dramatic retreat of the whole McCarty Glacier

35 36 o.c. Wiles and P. E. Calkin

system over the past century has uncovered extensive exposures of glacial sediments. These sediments in- clude stratigraphic sections of multiple units, paleosols and buried forests. Radiocarbon ages from organics in these sediments from the McCarty Fjord and a few from adjoining areas are summarized in Table 1. Ages have been converted to calibrated years using the Radiocarbon Age Calibration Program of Stuvier and Reimer (1986). Discussion below will refer to the mean calibrated age. Sections of subfossil logs obtained from a variety of stratigraphic settings have provided the opportunity to fine-tune the radiocarbon-based glacial history of the fjord using tree-ring cross dating. Ring-width time series were constructed at the Tree-Ring Laboratory of Lamont-Doherty Geological Observatory, Columbia University. Cross-dating and quality control for ring- width chronologies were performed using the computer program COFECHA (Holmes, 1983). Chronologies have been built using the ARSTAN computer program (Cook, 1985; Cook and Holmes, 1984). This program produces tree-ring chronologies FIG, 1. Index map of the southern Kenai Mountains. Numbers refer from ring-width measurement series. Detrending and to the following glaciers: (1) McCarty; (2) Dinglestadt (both the northwest and southeast flowing tongues); (3) Delectable glaciers; indexing the series removes most of the age-trend (4) Petrof; (5) Grcwingk. effects which are primarily caused by trees tending to

McCorty Glacier

.~,*,***~,*~.~ Location I Dinglestodt "GI;cie :3674 yr BP (UW-513) 3563 yr BP (BETA-39631) 554 A.D. (UW-514) • Location 5 0 1261 A.D. [] Delectable Boy (BETA-33800} Location 3 [ r~,.~,~ E]I Location 4 ...... ~ ) f'~p. = ,966 A.D. (BETA-35927) 596 ...... ~I/ ~ 1897 A.D. (BETA-39629) Q/ / 1968 A.D.(BETA-39630) / ~Locotion 2 / ('/.~,e,A.o.(uw-,,2) N James / Desire ~"

(3 ~__/~..~, L,/ % . 5km ' o,ocene _ moraine Rodlocorbon age loCotlons ,~r.~Or.j L cation of buried forests

FIG. 2. Index map of McCarty Fjord. Numbered locations refer to sites of radiocarbon ages and tree-ring ages. Neoglacial Fluctuations and Sedimentation 37 grow rings of decreasing width with increasing age. absolute age was limited by the calibrated radiocarbon Individual ring-width indices are calculated by dividing date. This chronology spans 508 years from 408 to 916 the actual measurements by values from the growth- A.D. and is composed of 15 time series of ring-width trend curve. Means of indices for each year are radii from 11 trees. computed using a biweight robust estimate. The The second strategy used in dating involved the processing also reduces the effects of any unusual construction of a living tree-ring chronology (Fig. 3b). individual tree growth variations and quantifies the This chronology, taken from the nearby Petrof Glacier common growth variations of all trees in a sample set. area (Fig. 1), is made up of 26 cores from 21 trees and Climate variations are usually the common factor spans 373 years, from 1616 to 1988 A.D. Cross-dating influencing growth ring variations for all trees from one tree-ring series from subfossil logs with this living location. The resulting chronologies were used in chronology allows us to assign calendar dates to the computer cross-dating of trees of uncertain age. logs preserved in the sediments. Tree-ring dating of logs preserved in sediments here was implemented using two strategies. In the first CHRONOLOGY OF THE McCARTY TRUNK strategy, logs of known radiocarbon age were cross- GLACIER dated with those of unknown age. These logs were derived primarily from forests composed of up to 20 in Early Advances situ and transported trees located in tributary valleys The earliest indication of McCarty Glacier advance along the east side of the fjord (Fig. 2). Most of these during the Holocene is inferred from glacial strati- subfossil logs are rotted; however, enough logs were graphy and two radiocarbon ages. Post (1980) sampled to construct a 'floating' tree-ring chronology obtained an age of 3674 BP (UW-513, Loc. 1, Fig, 2; (Fig. 3a) from trees in Delectable Bay and near Desire Table 1) from detrital wood lying on an erosion Lake (Fig. 2). This chronology was pinned down in surface, 6 km beyond the present east margin of absolute time by a calibrated 14C age of 897 A.D. McCarty Glacier and 10 m above high tide. We (BETA-39629) from the outer rings of one of the cross- obtained wood of similar age, 3563 BP (BETA- dated trees. That is, while the relative timing of events 39631) in this area near sea level from a diamict could be resolved in terms of years to decades, the overlain by tens of meters of sand and gravel.

Floating ring width chronology from McCorty Fjord

(a) ~ i

Z¢

IZ 897 A.D. (BETA-39629)

i | i i 'V, I 400 500 600 700 800 900 ~] Desire L~ke

Delectable I Range of subfossil trees Bay

i I i I I I I I I I I I 400 500 600 700 800 900 Year A.D.

Petrof chronology and cross-dated subfossil log from McCor~ Fiord IIII u

(b) £ "(3

g . Range of crou-doted subfoull tree

1600 1700 1800 1900 2000 Year A.D. FIG. 3. Tree-ring series: (a) Floatingtree-ring chronologyfrom subfossillogs of Delectable Bay and Desire Lake in McCarty Fjord. Also shown are the ranges of the 15 time series from the 11 subfossil trees; (b) Petrof living tree-ring chronologyand correlative time series from a subfossil log in McCarty Fjord used in dating. 38 G.C. Wiles and P. E. Calkin

Together these occurrences suggest McCarty Glacier for the floating chronology. Two other in situ tree may have been advancing near its present margin at stumps yielded ages of 968 and 968 A.D. (BETA- this time. 35927, BETA-39630), which are within one standard Drift containing both in situ tree remains and deviation of BETA-39629. In addition, trees near transported subfossil wood comprise evidence of a Desire Lake, to the south, were killed about 890 A.D. second advance (early medieval advance, Fig. 4). A (Fig. 3a). This nearly contemporaneous killing of trees, branch collected by Post (1980) from a bouldery till, 80 about 900 A.D., points to a third expansion of the m above sea level at Location 1 (Fig. 2), yielded an age McCarty trunk glacier. of 554 A.D. (UW-514). He also obtained an age of 561 Current direction indicators measured in sand and A.D. (UW-512) from a log 6 m above sea level at gravel that buried the forest in Delectable Bay show midfjord, Location 2 (Fig. 2). Furthermore, we ob- that the sediment here was derived from within this tained an age of 596 A.D. (BETA-33345, Table 1) from tributary valley. However, better sorted east-dipping the outer rings of an in situ tree that was apparently sequences of sediments that bear cross-dated tree overwhelmed by outwash along the margin at midfjord trunks in Desire Bay suggest deposition from the trunk (Loc. 3, Fig. 2). Ring counts from this stump indicate glacier into an ice-dammed lake. Therefore, we believe ice-free conditions existed for at least 200 years the McCarty Glacier was at an advanced position previous to burial. during the 9th century in order to dam Delectable and These ages are consistent with a well-established, Desire Lakes. Aggradation within these tributaries brief glacial advance dated at about 500 A.D., which is must have occurred due to a change in base level also displayed on the forefields of at least four outlet resulting from the filling of McCarty trough by ice. glaciers from the nearby icefields (Wiles and Calkin, McCarty Glacier probably continued to advance into 1990). This corresponds with the early medieval cold the 18th and 19th centuries. Cross-dating of trees that interval recorded in the Northern Hemisphere (Lamb, are preserved in diamicts with the Petrof tree-ring 1977). chronology (Fig. 3b) suggests that ice reached within a The terminal position of the McCarty Glacier during kilometer of its terminal moraine about 1790 A.D. this early medieval advance is not certain; however, (Loc. 6, Fig. 2). Grant and Higgins (1913) in 1909 limits of ice extent can be estimated based on the estimated that ice reached its terminal position here floating tree-ring chronology of Fig. 3a. Cross-dating of about 1860 A.D. (Fig. 4). The motions of the ice logs near Desire Lake, just beyond midfjord (Fig. 2), margin between 915 A.D. when it dammed Delectable attests to continuous forest growth from 408 to 890 A.D., coincident with this early advance. East-trending 2OOO sand and gravel fans exposed in Delectable Bay (Fig. 5) ~0 require a source from the adjacent advanced McCarty 1800 trunk glacier which must have dammed this tributary. 200 Little Ice Age This damming occurred prior to soil development, and odvonce spruce forest growth on these gravels, which began 1600 4O0 about 680 A.D. This age, obtained from the inner rings _ Grewingk Glacis,,. of the oldest tree cross-dated from Delectable Bay (Fig. 1400 1 3a), also provides a minimum for ice removal from the 600

adjacent fjord. Fan sedimentation into Delectable Bay C~ Delectoble tributory .~ 12oo glociers -- 800 (Fig. 5), continuous forest growth near Desire Lake, fl_ and forests buried in outwash with ages from the latter If) McCorty Glacier half of the 6th century (Fig. 2), all suggest early o iooo iooo ~- medie~val ice reached a maximum at Delectable Bay. Delectable Bay corresponds to a marked southward 800 -- 1200 widening (Fig. 2) of the fjord. /w / / ...w I I / / Later Advance 6O0 11 i -- 1400 The extent of retreat that occurred from midfjord by Grewir~k Glocier McCorty GIocier 680 A.D. is not documented; however, a readvance Eorly medievol 400 -- odvonce -- 1600 brought the McCarty Glacier back through the Delect- I ~ able Bay position (Fig. 2) within an interval as short as I I i 1 i I I two centuries to a terminus at James Lagoon (Figs 2, I 2 3 ~ 5 IO 15 Holocene Holocene Present McCorty 4). This terminus is now marked by a distinct end ice limit ice limit morgm moraine shoal and occurs where opposing tributaries Advonce/Retreot P allowed the glacier width to double, thus inhibiting Distance (km) further ice advance. FIG. 4. Late Holocene time--distance diagram comparing McCarty All cross-dated trees (Fig. 3a) sampled within De- Glacier fluctuations with those of Grewingk Glacier. The latter represents the general sequence of land-terminating glaciers in the lectable Bay were buried in outwash by 915 A.D., as southern Kenai Mountains. The maximum age for advance of the based on a radiocarbon age of 897 A.D. (BETA-39629) Delectable tributary glacier is indicated. Neoglacial Fluctuations and Sedimentation 39

TABLE 1. Radiocarbon ages from the southern Kenai Mountains

Calibrated interval Uncalibrated age Location on Laboratory no. BP BP Year A.D. Fig. 2

BGS-1278 1440± 70 1397 (1329) 1292 553 (621) 658 -- BETA-33345 1480± 70 1416 (1354) 1307 534 (596) 643 3 BETA-33800 690±70 685 (669) 643 1265 (1281) 1379 5 BETA-35927 1090±70 1064 (982) 938 886 (968) 1012 4 BETA-39629 1120 ± 60 1070 (1053) 96 880 (897) 986 4 BETA-39630 1090 ± 50 1059 (982) 950 891 (968) 1000 4 BETA-39631 3310± 60 3631 (3563) 3470 BC 1682 (1614) 1521 1 *UW-512 1500 ± 90 1516 (1389) 1307 434 (561) 643 2 *UW-513 3395 ± 75 3816 (3681, 3674, 3641) 3569 BC 1867 (1732, 1725, 1692) 1620 1 *UW-514 1510 ± 95 1524 (1396) 1310 426 (554) 640 1

*From Post (1980). Post (pers. commun., 1993) has noted that the locations of UW-512 and UW-514 were reversed in his original reference.

Bay, and 1790 A.D. when it reached near its terminus diamict facies overlie alluvial sediments preserving the 9.5 km to the south, are unknown. A continuous buried forest (Fig. 5). The lower diamict consists of 10 advance during this interval could have averaged about cm to more than a meter of pebbly compact silt and is 11 m year -1. Details of recent recession from its apparently a basal till. This lower till may record terminal position are considered in a separate section advance of ice from within Delectable Bay across below. lacustrine sediments prior to eventual confluence with McCarty Glacier. The superjacent diamict is a gravelly, INTERACTION BETWEEN McCARTY AND LAND- loose deposit interpreted as a supraglacial meltout till. TERMINATING GLACIERS The upper coarser till (Fig. 5), which occurs sporadic- ally and up to a meter in thickness, represents the The chronologies of tributary glaciers are con- medial and lateral moraine debris shown in aerial strained by glacial stratigraphy and radiocarbon ages. photographs of 1950 A.D. A maximum age for cirque glacier advance in Delect- Dinglestadt is the other significant land-terminating able valley (Loc. 5, Fig. 2) is inferred from a date of glacier entering McCarty Fjord (Figs 1, 2). The west- 1281 A.D. (BETA-33800) obtained from the upper flowing tongue of the Dinglestadt transection glacier on organic-rich beds of a lake sequence. These sediments, Kachemak Bay (Fig. 1) was advancing over forests which occur within a few hundred meters of the rapidly at its present margin by 621 A.D. (BGS-1278; Wiles backwasting margin of the easternmost glacier, are and Calkin, 1990). Therefore, it seems likely that capped by a diamict recording subsequent ice advance the eastern tongue of Dinglestadt Glacier entering down Delectable valley. This evidence is interpreted to McCarty Fjord may also have been advancing across suggest that ice buildup in tributary valleys and major its present marginal position and confluent with ice expansion lagged behind the McCarty advance the McCarty trunk glacier at this time. (-1300 A.D.) by more than 400 years. Evidence of later ice advance at Dinglestadt Glacier Near the mouth of Delectable Bay two superimposed consists of an undated deltaic sequence overlain by a till

Section in Delectable Bay (looking East )

~:'7~'::~ Lower diomict

g.:v',,'2"'t°'°° ~!':'." Sediment transport buried forest \ tkS~SBimnMl~-:--."' ~ - • direction 968 A.D. ( BETA - 35927 ) 897 A.D. Deltaic (BETA- 39629) 968 A.D, L(BETA-39630)

FIG. 5. Stratigraphic section at Delectable Bay (Loc. 4, Fig. 2). 40 G.C. Wiles and P. E. Calkin

showed the margin 1.5 km behind its 1909 position. Observations by Whitney (1932) in 1927 recorded an additional 1.6 km of recession. More recent analysis of air photographs by Post (1980) indicated that the glacier had receded approximately 5 km from 1927 to 1942. By 1950, McCarty Glacier had backed off an additional 11 km and in 1960, 1.6 km more, putting it within a few kilometers of the fjord head. Post (1980) reports a slight readvance from this 1960 position by 1977. Analysis of photographs taken in 1979 show the margin near its 1977 position. Bud Rice (pers. commun., 1990) has recorded an advance of approximately Delectable Boy i 1 km from 1988 to 1990. This late Holocene retreat from a 20-km reach of the ~re Lake •James /~ /I/ fjord (Fig. 6) has occurred in approximately 55 years. Lagoon L "%t./ Calving rates (Table 2) as a function of water depth can be calculated from the bathymetric data of Post (1980). [ Xl905 ~ DelightLoRe X xl 5km Iceberg-calving rates (Vc) of fjord glaciers are pro- / ~- - Holoceneterminal portional to the average water depth at the glacier aaron) margin with a constant of proportionality of 27 a -1 (Brown et al., 1982). Ice velocities can be calculated using the continuity equation Vc = V-X, where V is ice flow velocity and X is the rate of change of glacier length (which is positive in the direction of flow). Table 2 shows the calving rates and ice velocities of McCarty Glacier over the past 50 years (Fig. 6). Fjord depths (Post, 1980) are considered to be minimum values since FIG. 6. Map showing ice margin positions of McCarty Glacier since sedimentation followed deglaciation. Calving speeds 1905. and ice velocities, by continuity, must also be minima. Calving rates have exceeded ice velocities over the uncovered by margin retreat after 1950. This section 51-year interval resulting in this dramatic retreat. The records initial empounding of lake waters by McCarty retreat rates have increased as the glacier margin has Glacier in front of the advancing Dinglestadt tongue revealed deeper waters to the north and then, have during the most recent advance of the trunk glacier and decreased as the ice margin was again grounded near subsequent Dinglestadt (tributary) ice advance. There- the head of its fjord. The interval with the highest rate fore, this stratigraphy from the forefields of Delectable of retreat is that between 1942 and 1950 (Fig. 5) where and Dinglestadt Glaciers support near-synchronous McCarty Glacier was entering the deepest reach of its expansions of these land-terminating glaciers with the basin. McCarty iceberg-calving tongue during the early Contrasting with this rapid retreat of the McCarty medieval advance. However, two of these stratigraphic fjord glacier has been the slow wasting of the land- sections also support a later Little Ice Age advance of terminating tributary glaciers. U.S. Geological Survey the land-terminating glaciers that occurred at least 400 years after that of McCarty Glacier. (a)

/ V////AI ,,oe ICE-MARGINAL RECESSION IN McCARTY (b) FJORD ) I G'eciera, (c) Lituya Retreat of the glacier tongues from McCarty Fjord is Glacier well-documented by observations and aerial photo- (d) graphy (Fig. 6). During July and August of 1909, Grant and Higgins (1913) recorded the ice margin as still Southern grounded on the end moraine about 360 m behind its Mtns. terminal position. The oldest spruce tree cored on this I I I I I terminal moraine in James Lagoon, provided a mini- 0 500 I000 1500 2000 Year A.D mum date for retreat of 1905 A.D. This date is based Land-terminating glaciers ~ Iceberg-calving glaciers on a ring count with an added regional estimate of 15 years for the time of tree establishment (Wiles and FIG. 7. Comparison of late Holocene glacial history (last 2000 years) Calkin, 1990). from the southern Kenai Mountains with histories of other glaciers in southern Alaska. Sources: (a) Denton and Karl~n, 1977; (b) A series of photographs taken by Paul Smith of the Goodwin, 1988; (c) Goldthwait et al., 1963; Mann and Ugolini, 1985; U.S. Geological Survey in 1925 (Whitney, 1932) (d) Porter, 1989. Neoglacial Fluctuations and Sedimentation 41

TABLE 2. Retreat and calving rates of McCarty Glacier

Time period Retreat rate Ice velocity Calvingspeed Mean water depth years A.D. (m year -1) (m year -1) (m year -1) (m)

1909-1925 100 1800 1900 70 1925-1927 800 2400 3200 120 1927-1942 300 3800 4100 150 1942-1950 1400 3200 4600 170 1950--1960 200 2100 2300 85

1:40,000-scale mapping photos taken in 1950 show the terminating glaciers in the southern Kenai Mountains deglaciated fjord with tributary glacier margins from and of other fjord glaciers in Alaska. However, a later Dinglestadt and Delectable valleys still in their ex- advance of McCarty Glacier is recorded beginning by tended positions. 900 A.D. This preceded the Little Ice Age, and spanned a local nonglacial interval correlated with the COMPARISON OF THE McCARTY GLACIER medieval warm period (Optimum). FLUCTUATIONS WITH OTHER HOLOCENE Tree-ring dating constrained more precisely the GLACIAL RECORDS intervals of ice expansion and contraction than the radiocarbon chronology alone. For example, without Figure 7 compares expanded phases of iceberg- the aid of tree-ring analysis, the nearly contempor- calving fjord glaciers in maritime Alaska with those of aneous killing of at least 10 trees in tributary valleys land-terminating glacier tongues in the southern Kenai along the eastern margin of McCarty Fjord within a 50- Mountains and in the Wrangell-St Elias Mountains. A year period from 867 to 915 A.D. would not have been major characteristic of the McCarty Glacier chron- recognized from the few radiocarbon ages. ology, as well as that of the Lituya Bay ice tongue, is that the last ice margin expansion began considerably AC KNOWLEDGEMENTS earlier than the characteristic advances of the Little Ice Age (Grove, 1988). In the southern Kenai Mountains We are indebted to the staff of the Kenai Fjords National Park, as elsewhere in the Northern Hemisphere (Porter, particularly B. Rice and M. Tetreau for their considerable logistic support and encouragement. The help of Kachemak Bay State Park 1986), these advances began during the 13th century or and the Pratt Museum of Homer is gratefully acknowledged. We later. That these iceberg-calving glaciers may have thank G. Jacoby for review of the manuscript. B. Buckley of the chronologies that vary with those of land-terminating Lamont-Doherty Tree Ring Laboratory helped considerably in the tree-ring dating. A. Post provided early encouragement and advice. glaciers is consistent with many studies (e.g. Mann, This study was funded by a grant DPP8922696 from the National 1986), but differs somewhat from interpretations of Science Foundation to P.E. Calkin and partially supported by grants synchroneity made by Porter (1989) from his Alaskan from the Geological Society of America and the Mark Diamond Research Fund from the State University at Buffalo. studies. In the sense that episodes of advance encom- pass 300-1000 years compared with those of deglaciation in less than 200 years, the existing record from REFERENCES McCarty Glacier is compatible with the work of Porter Ager, T.A. (1983). Holocene vegetational history of Alaska. In: (1989) in Icy Bay and Mann and Ugolini (1985) in Wright, H.E. Jr (ed.), Late Quaternary Environments of the United Lituya Bay. A more detailed chronology may show that States, pp. 128--141. University of Minnesota Press, Minneapolis, the 1000-year interval encompasses many shorter ad- Minnesota. Brown, C.S., Meier, M.F. and Post, A. (1982). Calving speed of vances and retreats of the ice margin. Whether recent Alaska tidewater glaciers, with application to Columbia Glacier. advance of McCarty Glacier since 1960 may mark a U.S. Geological Survey, Professional Paper 1258-C, 13. major expansion is uncertain. Clagett, G.P. (1988). Alaska Snow Surveys. 1988 Annual Data Summary. United States Department of Agriculture, Soil Conser- Based on observations in Alaska over the last 200 vation Service, Anchorage, AK. years (Meier and Post, 1987) most modern tidewater Cook, E.R. (1985). A time series approach to tree-ring standardiza- glaciers in fjords have advance rates in the 20 to 40 m tion. Ph.D. dissertation, University of Arizona. Cook, E.R. and Holmes, R.L. (1984). Users Manual for Program year -~ range. Advance rates of McCarty Glacier of ARSTAN. Laboratory of Tree Ring Research, University of about 11 m year -1 measured over the last 1000 years Arizona, Tucson. are somewhat lower than this average. Denton, G.H. and Karl~n, W. (1977). Holocene glacial and tree-line variations in the White River Valley and Skolai Pass, Alaska and Yukon Territory. Quaternary Research, 7, 63-111. CONCLUSIONS Goldthwait, R.P., McKellar, I.C. and Cronk, C. (1963). Fluctuations of Crillon Glacier system, southeast Alaska. International Associa- tion of Scientific Hydrology Bulletin, 8, 62-74. Evidence from McCarty Fjord suggests advance of Goodwin, R.G. (1988). Holocene glaciolacustrine sedimentation in the trunk glacier about 3600 BP. An early medieval Muir Inlet and ice advance in Glacier Bay, Alaska, U.S.A. Arctic advance of the trunk glacier initiated about 550 A.D. is and Alpine Research, 20(1), 55-69. Grant, U.S. and Higgins, D.F. (1913). Coastal glaciers of Prince clearly documented. It appears to have spanned an William Sound and Kenai Peninsula. U.S. Geological Survey interval similar to that of expanded phases of land- Bulletin, 526. 42 G.C. Wiles and P. E. Calkin

Grove, J.M. (1988). The Little Ice Age. Methuen, London. Porter, S.C. (1986). Pattern and forcing of Northern Hemisphere Holmes, R.L. (1983). Computer-assisted quality control in tree-ring glacier variations during the last millennium. Quaternary Research, dating and measurement. Tree-Ring Bulletin, 43, 69-75. 26, 27-48. Kelley, J.S. (1985). Geologic setting of the Kenai Peninsula and Porter, S.C. (1989). Late Holocene fluctuations of the fjord glacier Cook Inlet Tertiary basin, south-central Alaska. In: Sisson, A. system in Icy Bay, Alaska, U.S.A. Arctic and Alpine Research, 21, (ed.), Guide to the Geology of the Kenai Peninsula, Alaska, pp. 3- 364-379. 19. Alaska Geological Society, Anchorage, AK. Post, A. (1975). Preliminary hydrography and historical terminal Lamb, H.H. (1982). Climate: past, present and future, II. Climate changes of Columbia Glacier, Alaska. U.S. Geological Survey, History and the Future, Methuen, London. Hydrologic Atlas 559. Mann, D.H. (1986). Reliability of a fjord glaciers fluctuations for Post, A. (1980). Preliminary bathymetry of McCarty Glacier, paleoclimatic reconstructions. Quaternary Research, 25, 10-24. Alaska. U.S. Geological Survey, Open File Report 80-424. Mann, D.H. and Ugolini, F.C. (1985). Holocene glacial history of Rice, B. (1987). Changes in the Harding Icefield, Kenai Peninsula, the Lituya District, southeast Alaska. Canadian Journal of Earth Alaska. Masters Thesis, University of Alaska. Sciences, 22, 913-928. Ruffner, J.A. and Bair, F.E. (1987). Weather of U.S. Cities. Gale Meier, M.F. and Post, A. (1987). Fast tidewater glaciers. Journal of Research Group, Detroit, Michigan. Geophysical Research, 92(B9), 9051-9058. Stuvier, M. and Reimer, P.J. (1986). A computer program for Mercer, J.H. (1961). The response of fjord glaciers to changes in the radiocarbon age calibration. Radiocarbon, 28, 1022-1030. firn limit. Journal of Glaciology, 3, 850--858. Whitney, P.C. (1935). The recent retreat of McCarty Glacier, Plafker, G. (1969). Tectonics of the March 27, 1964 Alaska Alaska. Geographical Review, 2, 389-391. earthquake. United States Geological Survey, Professional Paper Wiles, G.C. and Calkin, P.E. (1990). Neoglaciation in the southern 543-I. Kenai Mountains, Alaska. Annals of Glaciology, 14, 319-322.