Debris cover and the thinning of Kennicott Glacier, Alaska: in situ measurements, automated ice cliff delineation and distributed melt estimates

Abstract. Many glaciers are thinning rapidly beneath melt-reducing debris cover, including Kennicott Glacier in Alaska where glacier-wide maximum thinning also occurs under debris. This contradiction has been explained by melt hotspots, such as ice cliffs, scattered within the debris cover. However, melt hotspots alone cannot account for the rapid thinning at Kennicott Glacier. We consider the significance of ice cliffs, debris, and ice dynamics in addressing this outstanding problem. We collected abundant in situ measurements of debris thickness, sub-debris melt, and ice cliff backwasting, allowing for extrapolation across the debris-covered tongue (the study area and the lower 24.2 km2 of the 387 km2 glacier). A newly developed automatic ice cliff delineation method is the first to use only optical satellite imagery. The adaptive binary threshold method accurately estimates ice cliff coverage even where ice cliffs are small and debris color varies. Kennicott Glacier exhibits the highest fractional area of ice cliffs (11.7 %) documented to date. Ice cliffs contribute 26 % of total melt across the glacier tongue. Although the relative importance of ice cliffs to area-average melt is significant, the absolute area-averaged melt is dominated by debris. At Kennicott Glacier, glacier-wide melt rates are not maximized in the zone of maximum thinning. Declining ice discharge through time therefore explains the rapid thinning. There is more debris-covered ice in Alaska than in any other region on Earth. Through this study, Kennicott Glacier is the first glacier in Alaska, and the largest glacier globally, where melt across its debris-covered tongue has been rigorously quantified.

Download Full-text

Debris cover and the thinning of Kennicott Glacier, Alaska

10.5194/egusphere-egu21-15674 ◽

2021 ◽

Author(s):

Leif S. Anderson ◽

William H. Armstrong ◽

Robert S. Anderson ◽

Dirk Scherler

Keyword(s):

Satellite Imagery ◽

Hot Spots ◽

High Mountain ◽

Threshold Method ◽

Ice Dynamics ◽

Ice Flow ◽

Debris Cover ◽

A Chain ◽

The Relationship

Many glaciers in High Mountain Asia are experiencing the debris-cover anomaly. The Kennicott Glacier, a large Alaskan Glacier, is also thinning most rapidly under debris cover. This contradiction has been explained by melt hotspots, such as ice cliffs, streams, or ponds scattered within the debris cover or by declining ice flow in time. We collected abundant in situ measurements of debris thickness, sub-debris melt, and ice cliff backwasting, allowing for extrapolation across the debris-covered tongue. A newly developed automatic ice cliff delineation method is the first to use only optical satellite imagery. The adaptive binary threshold method accurately estimates ice cliff coverage even where ice cliffs are small and debris color varies. We also develop additional remotely-sensed datasets of ice dynamical variables, other melt hot spots, and glacier thinning.Kennicott Glacier exhibits the highest fractional area of ice cliffs (11.7 %) documented to date. Ice cliffs contribute 26 % of total melt across the glacier tongue. Although the relative importance of ice cliffs to area-average melt is significant, the absolute area-averaged melt is dominated by debris. At Kennicott Glacier, glacier-wide melt rates are not maximized in the zone of maximum thinning. Declining ice discharge through time therefore explains the rapid thinning. Through this study, Kennicott Glacier is the first glacier in Alaska, and the largest glacier globally, where melt across its debris-covered tongue has been rigorously quantified.We also carefully explore the relationship between debris, melt hotspots, ice dynamics, and thinning across the debris-covered tongue. In doing so we reveal a chain of linked processes that can explain the striking patterns expressed on the debris-covered tongue of Kennicott Glacier.

Download Full-text

Observation of natural and artificial features on the sea surface from SAR satellite imagery with in-situ measurements

2011 IEEE International Geoscience and Remote Sensing Symposium ◽

10.1109/igarss.2011.6048937 ◽

2011 ◽

Cited By ~ 1

Author(s):

Chris Maingot ◽

Biao Zhang ◽

Alexander Soloviev ◽

Mikhail Gilman ◽

Silvia Matt ◽

...

Keyword(s):

Satellite Imagery ◽

In Situ Measurements ◽

Sea Surface

Download Full-text

Debris cover and the thinning of Kennicott Glacier, Alaska, Part B: ice cliff delineation and distributed melt estimates

10.5194/tc-2019-177 ◽

2019 ◽

Author(s):

Leif S. Anderson ◽

William H. Armstrong ◽

Robert S. Anderson ◽

Pascal Buri

Keyword(s):

High Resolution ◽

Mass Balance ◽

In Situ Measurements ◽

Empirical Relationships ◽

Full Extent ◽

Part C ◽

Important Contributor ◽

High Resolution Imagery ◽

Debris Cover

Abstract. The mass balance of many valley glaciers is enhanced by the presence of ice cliffs within otherwise continuous debris cover. We assess the effect of debris and ice cliffs on the thinning of Kennicott Glacier in three companion papers. In Part A we report in situ measurements from the debris-covered tongue. Here, in Part B, we develop a method to delineate ice cliffs using high-resolution imagery and use empirical relationships from Part A to produce distributed mass balance estimates. In Part C we describe feedbacks that contribute to rapid thinning under thick debris. Ice cliffs cover 11.7 % of the debris-covered tongue, the most of any glacier studied to date, and they contribute 19 % of total melt. Ice cliffs contribute an increasing percentage of melt the thicker the debris cover. In the lowest 4 km of the glacier, where debris thicknesses are greater than 20 cm, ice cliffs contribute 40 % of total melt. Surface lake coverage doubled between 1957 and 2009, but lakes do not occur across the full extent of the zone of maximum glacier thinning. Despite abundant ice cliffs and expanding surface lakes, average melt rates are suppressed by debris, the pattern of which appears to reflect the debris thickness-melt relationship (or Østrem’s curve). This suggests that, in addition to melt hotspots, the decline in ice discharge from upglacier is an important contributor to the thinning of Kennicott glacier under thick debris.

Download Full-text

Debris cover and the thinning of Kennicott Glacier, Alaska, Part A:in situ mass balance measurements

10.5194/tc-2019-174 ◽

2019 ◽

Author(s):

Leif S. Anderson ◽

Robert S. Anderson ◽

Pascal Buri ◽

William H. Armstrong

Keyword(s):

Mass Balance ◽

High Mountain ◽

Glacier Surface ◽

Ice Dynamics ◽

Air Temperatures ◽

High Resolution Imagery ◽

Debris Cover ◽

The Alps

Abstract. The mass balance of many Alaskan glaciers is perturbed by debris cover. Yet the effect of debris on glacier response to climate change in Alaska has largely been overlooked. In three companion papers we assess the role of debris, ice dynamics, and surface processes in thinning Kennicott Glacier. In Part A, we report in situ measurements from the glacier surface. In Part B, we develop a method to delineate ice cliffs using high-resolution imagery and produce distributed mass balance estimates. In Part C we explore feedbacks that contribute to glacier thinning. Here in Part A, we describe data collected in the summer of 2011. We measured debris thickness (109 locations), sub-debris melt (74), and ice cliff backwasting (60) data from the debris-covered tongue. We also measured air-temperature (3 locations) and internal-debris temperature (10). The mean debris thermal conductivity was 1.06 W (m C)−1, increasing non-linearly with debris thickness. Mean debris thicknesses increase toward the terminus and margin where surface velocities are low. Despite the relatively high air temperatures above thick debris, the melt-insulating effect of debris dominates. Sub-debris melt rates ranged from 6.5 cm d−1 where debris is thin to 1.25 cm d−1 where debris is thick near the terminus. Ice cliff backwasting rates varied from 3 to 14 cm d−1 with a mean of 7.1 cm d−1 and tended to increase as elevation declined and debris thickness increased. Ice cliff backwasting rates are similar to those measured on debris-covered glaciers in High Mountain Asia and the Alps.

Download Full-text

Debris cover and the thinning of Kennicott Glacier, Alaska, Part C: feedbacks between melt, ice dynamics, and surface processes

10.5194/tc-2019-178 ◽

2019 ◽

Author(s):

Leif S. Anderson ◽

William H. Armstrong ◽

Robert S. Anderson ◽

Pascal Buri

Keyword(s):

Mass Balance ◽

Lower Limb ◽

Surface Velocity ◽

Glacier Surface ◽

Ice Dynamics ◽

Part C ◽

High Resolution Imagery ◽

Debris Cover ◽

Process Domains

Abstract. The mass balance of many valley glaciers is enhanced by the presence of melt hotspots within otherwise continuous debris cover. We assess the effect of debris, melt hotspots, and ice dynamics on the thinning of Kennicott Glacier in three companion papers. In Part A we report in situ measurements from the debris-covered tongue. In Part B, we develop a method to delineate ice cliffs using high-resolution imagery and produce distributed mass balance estimates. Here in Part C we describe feedbacks controlling rapid thinning under thick debris. Despite the extreme abundance of ice cliffs on Kennicott Glacier, average melt rates are strongly suppressed downglacier due to thick debris. The estimated melt pattern therefore appears to reflect Østrem’s curve (the debris thickness-melt relationship). As Kennicott Glacier has thinned over the last century Østrem’s curve has manifested itself in two process domains on the glacier surface. The portion of the glacier affected by the upper-limb of Østrem’s curve corresponds to high melt, melt gradients, and ice dynamics, as well as high ice cliff and stream occurrence. The portion of the glacier affected by the lower-limb of Østrem’s curve corresponds to low melt, melt gradients, and ice dynamics, as well as high ice cliff and stream occurrence. The upglacier end of the zone of maximum thinning on Kennicott Glacier occurs at the boundary between these process domains and the bend in Østrem’s curve. The expansion of debris upglacier appears to be linked to changes in the surface velocity pattern through time. In response to climate warming, debris itself may therefore control where rapid thinning occurs on debris-covered glaciers. Ice cliffs are most abundant at the upglacier end of the zone of maximum thinning.

Download Full-text

Validation of Physical Radiative Transfer Equation-Based Land Surface Temperature Using Landsat 8 Satellite Imagery and SURFRAD in-situ Measurements

Journal of Atmospheric and Solar-Terrestrial Physics ◽

10.1016/j.jastp.2019.105161 ◽

2019 ◽

Vol 196 ◽

pp. 105161 ◽

Cited By ~ 8

Author(s):

Aliihsan Sekertekin

Keyword(s):

Radiative Transfer ◽

Surface Temperature ◽

Satellite Imagery ◽

Land Surface Temperature ◽

Land Surface ◽

Transfer Equation ◽

Radiative Transfer Equation ◽

In Situ Measurements ◽

Landsat 8

Download Full-text

Atmospheric CO2 Record from In Situ Measurements at Amsterdam Island

Carbon Dioxide Information Analysis Center (CDIAC) Datasets ◽

10.3334/cdiac/atg.013 ◽

1996 ◽

Author(s):

A. Guadry, ◽

V. Kazan, ◽

P. Monfray,

Keyword(s):

In Situ Measurements ◽

Atmospheric Co2

Download Full-text

The Carbon:²³⁴Thorium ratios of sinking particles in the California Current Ecosystem 2: Examination of a thorium sorption, desorption, and particle transport model

10.26434/chemrxiv.7803698 ◽

2019 ◽

Author(s):

Michael Stukel ◽

Thomas Kelly

Keyword(s):

Organic Carbon ◽

Water Column ◽

Particulate Organic Carbon ◽

Transport Model ◽

California Current ◽

In Situ Measurements ◽

Power Law Distribution ◽

Sinking Particles ◽

Organic Carbon Concentration

Thorium-234 (234Th) is a powerful tracer of particle dynamics and the biological pump in the surface ocean; however, variability in carbon:thorium ratios of sinking particles adds substantial uncertainty to estimates of organic carbon export. We coupled a mechanistic thorium sorption and desorption model to a one-dimensional particle sinking model that uses realistic particle settling velocity spectra. The model generates estimates of 238U-234Th disequilibrium, particulate organic carbon concentration, and the C:234Th ratio of sinking particles, which are then compared to in situ measurements from quasi-Lagrangian studies conducted on six cruises in the California Current Ecosystem. Broad patterns observed in in situ measurements, including decreasing C:234Th ratios with depth and a strong correlation between sinking C:234Th and the ratio of vertically-integrated particulate organic carbon (POC) to vertically-integrated total water column 234Th, were accurately recovered by models assuming either a power law distribution of sinking speeds or a double log normal distribution of sinking speeds. Simulations suggested that the observed decrease in C:234Th with depth may be driven by preferential remineralization of carbon by particle-attached microbes. However, an alternate model structure featuring complete consumption and/or disaggregation of particles by mesozooplankton (e.g. no preferential remineralization of carbon) was also able to simulate decreasing C:234Th with depth (although the decrease was weaker), driven by 234Th adsorption onto slowly sinking particles. Model results also suggest that during bloom decays C:234Th ratios of sinking particles should be higher than expected (based on contemporaneous water column POC), because high settling velocities minimize carbon remineralization during sinking.

Download Full-text