The Imminent Calving Retreat of Taku Glacier – Eos – Science News by AGU – Campbell et al.

Long an anomaly among glaciers, advancing while most others shrank, Taku Glacier is starting to succumb to climate change, offering an unprecedented look at the onset of tidewater glacier retreat.

By , J. M. Amundson, S. O’Neel, R. J. Motyka, L. Sass, M. Truffer, J. M. Zechmann, and S. Campbell 18 February 2021

Along the rugged Southeast Alaska coast, 30 kilometers northeast of the state capital Juneau, a tidewater glacier has largely defied global trends by steadily advancing for most of the past century while most glaciers on Earth retreated. This 55-kilometer-long and nearly 1,500-meter-thick tidewater glacier, named Taku Glacier, or T’aaḵú Ḵwáan Sít’i in the language of the Indigenous Tlingit people, has been the focus of continuous scientific study for more than 70 years. Some records even extend back to the mid-18th century. With this long observation record and the glacier’s year-round accessibility and proximity to Juneau and adjacent research facilities, Taku provides an unparalleled locale to study tidewater glaciers and their response to Earth’s rapidly changing climate.

Uncertain Impacts from Tidewater Glaciers

Earth’s approximately 3,000 tidewater glaciers—those that interact directly with the ocean—represent 42% and 57% of the planet’s glacierized area and volume, respectively, excluding the Greenland and Antarctic Ice Sheets (Figure 1). Tidewater glaciers differ from their land-terminating counterparts, where climate directly drives glacier change in that processes at or near the glacier-ocean boundary, such as sediment deposition, iceberg calving, and submarine melting, can lead to self-induced instabilities [e.g., Enderlin et al., 2018; Sutherland et al., 2019]. These additional processes can drive slow centennial-scale advances and rapid decadal-scale retreats known as tidewater glacier cycles, even in the absence of climate variability [Post et al., 2011; Brinkerhoff et al., 2017].

Figures illustrating regional percentages of tidewater, lake calving, and land-terminating glacier volumes as well as global areas and volumes
Fig. 1. (a) The relative percentages of tidewater, lake calving, and land-terminating glacier volumes by region. The size of each circle reflects the total ice volume of each region. Percentages of tidewater, lake calving, and land-terminating glacier (b) areas and (c) volumes globally. Click image for larger version.

Tidewater glaciers contribute significantly to local ecosystems through their high sediment production and freshwater discharge. The nutrient-rich freshwater directly discharged into the nearshore environment sustains highly productive marine ecosystems and fisheries [Meire et al., 2017], highlighting their important and dynamic role at local scales.

Climate warming can initiate tidewater glacier retreat; indeed, most tidewater glaciers are currently retracted or retreating. Compared with land-terminating glaciers, tidewater glaciers contribute disproportionately to sea level rise during their retreat because of nonlinear feedbacks between ice flow, calving, submarine melting, and sediment deposition—that is, relatively small changes at a glacier’s terminus can produce large changes in the glacier’s total volume. Although these additional processes can cause tidewater glaciers to retreat rapidly, they can also allow tidewater glaciers to stabilize and even advance under warming climate conditions [Brinkerhoff et al., 2017].

Overall, limited understanding of the processes driving tidewater glacier cycles leads to large uncertainties in current projections of how they will influence sea level rise and ecosystem change. These uncertainties in turn affect issues ranging from fisheries management to forecasting coastal community inundation from sea level rise, thus challenging human adaptation to climate change [Catania et al., 2020].

A Lack of Tidewater Glacier Observational Records

Land-terminating glaciers have been intensively and increasingly studied around the world during the 20th and early 21st centuries [e.g., Zemp et al., 2019], but our understanding of tidewater glaciers stems from only a few glaciers. Arguably the best documented tidewater glacier retreat continues today at Alaska’s Columbia Glacier. Highlighting the catastrophic nature of some tidewater glacier retreats since it began in about 1980, this glacier has thinned by more than 500 meters in its lower reaches, retreated more than 20 kilometers, and shed more than half its volume [Enderlin et al., 2018]. During the past 2 decades, Columbia Glacier’s average annual mass loss represents 5% of sea level rise contributions from the Alaska region [Larsen et al., 2015], outpacing all other mountain glacier regions on Earth [Zemp et al., 2019].

More than 50 publications on Columbia Glacier illustrate how ice flow and calving can drive tidewater glacier retreats, outweighing the effects of variations in surface mass balance and climate. Research elsewhere has highlighted the ocean as another prominent driver of tidewater glacier retreat, particularly through its influence on submarine melting [e.g., Sutherland et al., 2019]. Sedimentary processes at or near the glacier’s terminus may also influence ice flow, calving rates, and submarine melting [Eidam et al., 2020].

As a tidewater glacier advances through its fjord, it excavates sediments along the ocean floor and redeposits them in front of the glacier, producing a moraine along the glacier’s terminus [Motyka et al., 2006]. This moraine helps to stabilize the glacier, introducing resistance that prevents fast ice flow while reducing calving by supporting the ice face and limiting submarine melting by restricting direct glacier-ocean contact. However, a retreat from this sediment moraine can destabilize a glacier by allowing ocean water to contact and melt the glacier once again. The then unsupported ice face can subsequently experience increased calving and ice flow. These processes can initiate a runaway retreat, with the glacier rapidly backpedaling through its overdeepened basin—where the bed is below sea level—that was excavated upstream of the moraine during the glacier’s advance.However, tidewater glaciers with high sediment production can produce a new moraine relatively quickly [Eidam et al., 2020], allowing for stabilization or even readvance without a change in climate. Hence, sediment production likely controls the periodicity of tidewater glacier cycles and susceptibility to catastrophic retreat [Brinkerhoff et al., 2017]. Yet a lack of tidewater glacier observational records, specifically of sediment processes, confounds modeling of tidewater glacier cycles and of these glaciers’ impacts on 21st-century sea level rise and ecosystem change.

The Taku Anomaly

Figure showing locations of the Taku Glacier terminus in 1948 and 2018
Fig. 2. (a) The location of Taku Glacier east of Juneau in Southeast Alaska is shown in this satellite image acquired by Sentinel-2 on 1 September 2019. Late summer images show Taku Glacier’s terminus in (b) 1948 and (c) 2018. In Figure 2c, the early stages of retreat are visible as a thin “bathtub ring” of recently deglaciated rock surrounding the terminus. The red dots indicate the location and viewing orientation of the scenes in Figure 3a (left dot) and Figure 3b (right dot).

In Southeast Alaska, 725-square-kilometer Taku Glacier (Figure 2) recently began retreating from its sediment moraine [McNeil et al., 2020]. The glacier winds 55 kilometers through the Juneau Icefield, with an overdeepened bed reaching depths of 600 meters below sea level and remaining below sea level for about 40 kilometers upstream from its present-day terminus [Nolan et al., 1995]. The glacier meets the ocean at its termination in Taku Inlet, where its activity has been observed for more than 2 centuries.

During a previous retreat between 1750 and 1793, Taku Glacier retreated 8.5 kilometers of the 40-kilometer-long overdeepened section of its bed. Simple conceptual models of tidewater glacier cycles suggest that a tidewater glacier will retreat the length of its overdeepened bed before stabilizing and readvancing [e.g., Post et al., 2011]. In the latter half of the 18th century, however, Taku Glacier retreated less than 40% of the distance of its overdeepened bed before readvancing [Nolan et al., 1995], never reaching the catastrophic retreat rates observed recently at Columbia Glacier [Post and Motyka, 1995]. And by 1890, the glacier stabilized and was advancing again while other glaciers in the region retreated.

From 1890 until 2015, a time when most glaciers worldwide retreated, Taku Glacier advanced. The glacier’s sustained advance, especially under a warming climate, has captivated researchers for more than 70 years. The resulting records of mass balance, ice flow, ice thickness, and sediment processes make Taku Glacier’s advance perhaps the best documented of any tidewater glacier in the world.Taku Glacier calved icebergs into Taku Inlet through the early 20th century, but by 1950 calving had largely ceased as its sediment moraine rose above the water surface, protecting it from the ocean’s influence. As the glacier continued to advance during the next 7 decades, it pushed this moraine into Taku Inlet. Today the moraine extends 12 kilometers downstream from the glacier’s terminus, largely filling Taku Inlet. Despite its reach, the moraine rises only a few meters above sea level at its highest point, and high tides and storm surges infiltrate Taku Inlet to within a few tens of meters of the present-day terminus.

Taku Glacier’s advance significantly slowed in the late 1980s and has now transitioned to retreat and mass loss [McNeil et al., 2020]. Regional temperatures rose approximately 2ºC in the past 70 years, increasingly influencing the glacier’s mass. The glacier gained mass and advanced between 1946 and 1989, but the mass change approached equilibrium, and the advance slowed by 72% after 1989. Since 2013, the glacier has lost mass, thinning at an average rate of 1.3 meters per year, and it has retreated from its moraine (Figure 3), with record annual mass loss in 2019. The past few decades of observations imply that climate warming has caught up with Taku Glacier, likely initiating the retreat phase of the tidewater glacier cycle.

Photos of scenery, including the moraine, near Taku Glacier
Fig. 3. Taku Glacier’s moraine, shown here on 30 June 2019, was advancing until at least 2015, as documented with time-lapse photography. The glacier terminus undergoes a seasonal advance in winter and retreat in summer on the order of 20 meters. The seasonally advanced terminus position in 2019 was farther back than the seasonally retreated terminus position in 2015. (a) The western portion of the terminus was previously advancing and prograding the moraine into the Norris River. The glacier is now separated from the moraine by a moat. (b) Portions of the moraine that were actively deforming in summer 2015 are now covered with Nootka lupine and equisetum (horsetail), two colonizing plants commonly found in Southeast Alaska. Credit: Jason Amundson

What Can Taku Teach Us About Tidewater Glaciers?

The accessibility of and the existing research infrastructure at the historically well studied Taku Glacier provide a unique and timely opportunity to observe a tidewater glacier’s transition from advance to retreat and to better understand how this retreat affects fjord ecosystems as well as local economies, such as the tourism and salmon fishing industries in and around Taku Inlet.

Priorities of continuing research at Taku include better characterizing (1) sedimentation rates and moraine stability; (2) the evolution and influence of submarine melting; (3) terminus destabilization, initiation of calving, and evolution of glacier flow and mass balance; (4) changes in basal properties within the overdeepened bed; (5) the evolution of fjord chemistry and ecology; and (6) impacts on local tourism and fishing economies.

Specifically, studies of Taku Glacier will elucidate the role of sediment during retreat, which is underrepresented in models of tidewater glaciers and of their impacts on sea level rise and ecosystem change. After retreat begins, glaciers with submarine moraines, like Columbia Glacier, might experience a quick onset of increased ice flow, calving, and submarine melting, allowing for rapid mass loss due to the immediate oceanic influence. However, subaerial moraines like the one at Taku Glacier may greatly reduce or entirely prevent contact with the ocean. Lacking direct ocean interaction, submarine melting and calving would be reduced, potentially preventing a catastrophic retreat like that observed at Columbia Glacier. As Taku Glacier retreats and the moraine stops receiving sediment, though, adjacent rivers and tides may still erode the moraine, progressively increasing glacier-ocean interactions [Zechmann et al., 2020].Either scenario—moraine preservation or degradation—will likely result in disproportionate glacier mass loss compared with loss at terrestrial glaciers because of ice flow, calving, and submarine melting. However, the timescales of these loss scenarios differ from decades (degradation) to centuries (preservation).

As research at Columbia Glacier enhanced our understanding of mechanisms driving catastrophic tidewater glacier retreat, detailed analyses of Taku Glacier—with conditions and a configuration that differ from Columbia’s—will help scientists better grasp the spectrum of tidewater glacier behaviors. Given the uncertain longevity of tidewater glaciers, this understanding is crucial to further elucidate the susceptibility of these glaciers globally to retreat and to predict their effects on 21st-century sea level rise and ecosystem change.


Immense gratitude must be shown to Matt Nolan for sharing aerial imagery of Taku Glacier. We additionally thank Caitlyn Florentine and Adrian Bender for their thoughtful and constructive comments, which greatly enhanced our article. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. government.


Brinkerhoff, D., et al. (2017), Sediment transport drives tidewater glacier periodicity, Nat. Commun., 8, 90,

Catania, G. A., et al. (2020), Future evolution of Greenland’s marine-terminating outlet glaciers, J. Geophys. Res. Earth Surf., 125(2), e2018JF004873,

Eidam, E., et al. (2020), Morainal bank evolution and impact on terminus dynamics during a tidewater glacier stillstand, J. Geophys. Res. Earth Surf., 125(11), e2019JF005359,

Enderlin, E., et al. (2018), Evolving environmental and geometric controls on Columbia Glacier’s continued retreat, J. Geophys. Res. Earth Surf., 123(7), 1,528–1,545,

Larsen, C. F., et al. (2015), Surface melt dominates Alaska glacier mass balance, Geophys. Res. Lett., 42(14), 5,902–5,908,

McNeil, C., et al. (2020), Explaining mass balance and retreat dichotomies at Taku and Lemon Creek Glaciers, Alaska, J. Glaciol., 66(258), 530–542,

Meire, L., et al. (2017), Marine‐terminating glaciers sustain high productivity in Greenland fjords, Global Change Biol., 23(12), 5,344–5,357,

Motyka, R. J., et al. (2006), Rapid erosion of soft sediments by tidewater glacier advance: Taku Glacier, Alaska, USA, Geophys. Res. Lett., 33(24), L24504,

Nolan, M., et al. (1995), Ice-thickness measurements of Taku Glacier, Alaska, USA, and their relevance to its recent behavior, J. Glaciol., 41(139), 541–553,

Post, A., and R. J. Motyka (1995), Taku and Le Conte glaciers, Alaska: Calving-speed control of late-Holocene asynchronous advances and retreats, Phys. Geogr., 16(1), 59–82,

Post, A., et al. (2011), A complex relationship between calving glaciers and climate, Eos, 92(37), 305–306,

Sutherland, D. A., et al. (2019), Direct observations of submarine melt and subsurface geometry at a tidewater glacier, Science, 365(6451), 369–374,

Zechmann, J., et al. (2020), Sediment redistribution beneath the terminus of an advancing glacier, Taku Glacier (T’aakú Kwáan Sít’i), Alaska, J. Glaciol.,

Zemp, M., et al. (2019), Global glacier mass changes and their contributions to sea-level rise from 1961 to 2016, Nature 568, 382–386,

Author Information

Christopher McNeil (, U.S. Geological Survey, Anchorage, Alaska; J. M. Amundson, University of Alaska Southeast, Juneau; S. O’Neel, Cold Regions Research and Engineering Laboratory, Hanover, N.H.; R. J. Motyka, University of Alaska Fairbanks; L. Sass, U.S. Geological Survey, Anchorage, Alaska; M. Truffer and J. M. Zechmann, University of Alaska Fairbanks; and S. Campbell, University of Maine, Orono

Citation: McNeil, C., J. M. Amundson, S. O’Neel, R. J. Motyka, L. Sass, M. Truffer, J. M. Zechmann, and S. Campbell (2021), The imminent calving retreat of Taku Glacier, Eos, 102, Published on 18 February 2021.
Text © 2021. The authors. CC BY 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.