Summertime formation of depth hoar in central Greenland
Alley, R. B., E. S. Saltzman, K. M. Cuffey, and J. J. Fitzpatrick
Geophysical Research Letters, Vol. 17, No. 12, p. 2393-2396, December 1990

Summertime solar heating of near-surface snow in central Greenland causes mass loss and grain growth. These depth hoar layers become seasonal markers which are observed in ice cores and snow pits. Mass redistribution associated with depth-hoar formation can change concentrations of immobile chemicals by as much as a factor of two in the depth hoar, altering atmospheric signals prior to archival in ice. For methanesulfonic acid (MSA) this effect is not significant because the summer maximum does not coincide with the density minimum, and the amplitude of the annual (MSA) signal is more than a factor of ten.

Boudinage: A source of stratigraphic disturbance in glacial ice in central Greenland
Cunningham, J. and E.D. Waddington
Journal of Glaciology, Vol. 36, No. 123, p. 269-272, 1990

A hydrodynamic model of interface stability in a stratified fluid is reviewed. The model predicts that irregularities on the boundaries of a stiff layer, embedded in a soft matrix, are unstable in pure shearing flow, when compression is normal to the layer. Perturbations on such a layer can grow to form symmetric pinch-and-swell structures called boudins. The model predicts initial perturbation growth rates on the boundaries of an interglacial period ice layer. We find that, beneath an ice divide, irregularities on the Sangamon layer boundaries will not kinematically decay, as the layer thins. Finite-element modelling is used to determine the strain history of Sangamon ice beneath the divide at Summit, Greenland. That history suggests boundary irregularities have grown, relative to layer thickness, at least 26 fold over the past 90,000 years. The result may be sever distortion or severing of the layer. Core holes penetrating the layer may recover anomalously thick or thin columns of ice resulting in erroneous environmental and climatic interpretations. Radio echo-sounding may be useful in searching for zones of boudinage, which should be avoided when coring. Initial perturbations might arise from mass-balance spatial variations or from transient flow fields.

Flow-law hypotheses for ice-sheet modeling
Alley, R. B.
Journal of Glaciology, Vol. 38, No. 129, p. 245-256, 1992

Ice-flow modeling requires a flow law relating strain rates to stresses in situ, but a flow law cannot be measured directly in ice sheets. Microscopic processes such as dislocation glide and boundary diffusion control both the flow law for ice and the development of physical properties such as grain size and c-axis fabric. These microscopic processes can be inferred from observations of the physical properties, and the flow law then can be estimated from the microscopic processes. A review of available literature shows that this approach can be imperfectly successful. Interior regions of large ice sheets probably have depth- varying flow-law "constants", with the stress exponent, n, for power-law creep less than 3 in upper regions and equal to 3 only in deep ice; n probably equals 3 through most of the thickness of ice shelves and ice streams.

Sensitivity of the ice-divide position in Greenland to climate change
Anandakrishnan, S., R. B. Alley, and E. D. Waddington
Geophysical Research Letters, Vol. 21, No. 6, p. 441-444, March 15, 1994

Model calculations of depth-age relations for deep ice cores in central Greenland are sensitive to stability of the ice-divide position. In addition, the folding of layers observed in the deep ice could be instigated by divide migration changing the velocity and particle path of ice flow. We use simple steady-state calculations to show that lateral divide migration of between 10 km and 50 km and elevation change of approximately 100 m is likely on glacial-interglacial time scales, enough to affect model dating. The ice- divide location appears to be most sensitive to the position of the ice-sheet margins. By contrast, the ice- divide elevation is most sensitive to the accumulation rate, the temperature profile, and the ice-stiffness profile.

Mapping c-axis fabrics to study physical processes in ice
Alley, R. B., A. J. Gow and D. A. Meese
Journal of Glaciology, Vol. 41, p. 197-203, 1995

Mapping the spatial distribution of c-axis orientations in ice thin sections is not much harder than preparing c-axis scatter-plots, but can reveal important additional information about processes responsible for the fabric and texture of the ice. Distributions of angles between c-axes of neighboring grains from the Byrd Station, West Antarctica ice core suggest that polygonization causes average grain size to stabilize below 400 m depth.

Shear-wave detection of asymmetric c-axis fabrics in the GISP2 ice core
Anandakrishnan, S., J.J. Fitzpatrick, R.B. Alley, A.J. Gow and D.A. Meese
Journal of Glaciology, Vol. 40, p. 491-496, 1995

C-axis fabrics of the GISP2 ice core from central Greenland have been measured rapidly and accurately in the field, using both compressional and shear waves generated by an inexpensive, commercially-available, "idiot-proof" device. Compressional-wave data were collected at 10 m intervals for the upper 2250 m of the ice sheet, and show progressive clustering of c-axes toward the vertical with increasing depth but no large steps at climatic boundaries in the core. The degree of clustering measured by ultrasound agrees closely with that measured using traditional optical techniques, but the ultrasound technique is easier and faster than optical methods. A slight asymmetry in the c-axis clustering is revealed by the shear-wave data and increases with increasing depth, indicating that deformation is not symmetric about the vertical at the site.

Dominant control of atmospheric circulation on snow accumulation in central Greenland
Kapsner, W.R., R.B. Alley, C.A. Shuman, S. Anandakrishnan and P.M. Grootes
Nature, Vol. 373, p. 52-54, 1995

Atmospheric circulation and not temperature is the primary control on snow accumulation in central Greenland over the last 17,000 years, based on correlation of accumulation to temperature calculated from the isotopic composition of a deep ice core. Within both warm (Holocene) and cold (Younger Dryas, glacial maximum) climate states, the sensitivity of accumulation to temperature is less than expected if accumulation is controlled primarily by the ability of warmer air to deliver more moisture. During transitions between warm and cold climate states, accumulation varies more than can be explained thermodynamically, probably because of storm-track shifts. In a greenhouse-warmed world, any circulation changes may be more important than the direct effects of temperature change in controlling accumulation in Greenland and its contribution to sea-level change.

Constraints on Holocene ice-thickness changes in central Greenland from the GISP2 ice-core data
Bolzan, J.F., E.D. Waddington, R.B. Alley, and D.A. Meese
Annals of Glaciology, Vol. 21, p. 33-39, 1995

The depth-age relation observed in the GISP2 ice core is the result of the integrated effects of ice-sheet changes over time, as well as the accumulation- rate history. Here we construct a forward model to compute ages at various depths in the core. In the model, the ages are functions of parameters that describe the thickness as a function of time. Using the maximum likelihood inverse method, these parameters are relatively adjusted until measured and computed ages agree satisfactorily. The results suggest that the thickness along the flowline connecting the GISP2 and GRIP drill sites has not changed significantly since the onset of the Holocene. We also derive bounds on the likely thickness changes. Because these bounds are independent of assumptions concerning the processes driving the ice-sheet evolution, they can provide useful constraints for other ice-sheet modeling efforts.

Fracture toughness of ice and firn determined from the modified ring test
Fischer, M.P., R.B. Alley and T. Engelder
Journal of Glaciology, In Press

The modified ring test is used to determine the fracture toughness of synthetic, granular, freshwater ice (average density 0.891 g cm -3) and firn (average density 0.605 g cm -3) from depths between 26-27.2 m in the E core of the Greenland Ice Sheet Project II. Average fracture toughness is 146.3 kPa m0.5 for the manufactured ice and 107.3 kPa m0.5 for the firn. Comparison between the ice and firn suggests that ice fracture toughness decreases with decreasing density (i.e., increasing porosity), suggesting lateral and vertical variations in the near-surface fracture resistance of glaciers and ice sheets may be related to firn densification. The modified ring test has many advantages over conventional, notch-based specimens in that complications which arise in notched specimens due to crack length, loading rate, notch acuity and specimen size effects are irrelevant for a modified ring specimen geometry.

Finite element analysis of the modified ring test for determining mode I fracture toughness
Fischer, M.P., D. Elsworth, R.B. Alley and T. Engelder
International Journal of Rock Mechanics, In Press

Plane strain fracture toughness (KIc) values are determined for the modified ring (MR) test through numerical simulation of crack growth to highlight the sensitivity of MR KIc values on applied displacement or force boundary conditions, slip conditions at the specimen-platen interface, and the Poisson ration (v) of the test material. Numerical calculation of fracture toughness in the MR test is traditionally conducted assuming a uniform force along the specimen loading surfaces and no slip between the specimen and the loading platens. Under these conditions KIc increases by 30-40% as v decreases from 0.4 to 0.1. When slip is allowed at the specimen-platen interface under a uniform force, KIc values are independent of v, and for any given v, are 5-25% less than those determined under a no-slip boundary condition. Under a uniform displacement of the specimen loading surfaces, KIc is essentially independent of v, regardless of specimen-platen interaction. Moreover, although KIc values determined under uniform displacement and no-slip boundary conditions are always higher than those determined under uniform displacement and slip-allowed boundary conditions, the average difference in KIc for these two methods is less than 5% for the two specimen geometries examined. This suggests that under uniform displacement conditions, KIc is essentially independent of specimen-platen interaction. Because KIc values determined from MR testing are strongly dependent on the modeling procedure, future reports of KIc determined from this test should be accompanied by detailed reports of the modeling procedure. Until further testing reveals the most accurate simulation technique, we advocate use of a uniform displacement formulation for KIc determination from MR testing because results from this method are insensitive to most modeling parameters. Numerical results from models conducted under uniform force, no-slip boundary conditions should be viewed as an upper bound to KIc.

Impurity influence on normal grain-growth in the GISP2 ice core
Alley, R.B. and G.A. Woods
J. Glaciology, In Review

Intercept analysis of approximately bi-yearly vertical thin sections from the upper part of the GISP2 ice core, central Greenland, shows that grain-size ranges increase with increasing age. Correlation of grain- growth rates to chemical and isotopic data indicates slower growth in ice with higher impurity concentrations, and especially slow growth in forest fire layers containing abundant ammonium; however, the impurity/grain-growth relations are quite noisy. Little correlation is found between growth rate and isotopic composition of ice.