Student Projects

Student Projects 2022

Pairs of two students will work on a project.

In the past at least two summer school projects have evolved into scientific publications:
Altena et al., 2019
https://www.the-cryosphere.net/13/795/2019/
Brinkerhoff et al., 2016
https://www.cambridge.org/core/journals/annals-of-glaciology/article/inv...

PROJECT 1: Modeling the development of “glacier tables”

STUDENTS: Kyle Blum, Thomas Frank
ADVISOR: Eric Petersen
DESCRIPTION: Glacier tables develop from locally reduced melt underneath boulders, leading to a raised stem of ice on which the boulder is perched. In this project, threshold conditions for the development of glacier tables will be investigated from modeling of energy balance and heat conduction. Sensitivity analysis will be employed to show the relative effects of boulder size, geometry, and thermal conductivity. Further exploration is dependent on student interest and can include (1) a comparison of steady state vs. diurnal forcing, (2) investigating the melt and expected lifetime of the glacier stem, as well as boulder migration as it topples, (3) investigating glacier burrows, wherein smaller boulders melt cryoconite holes.
SOFTWARE REQUIREMENTS: Python, matlab, or any programming language in which the student is proficient, running locally on your machine. Advisor can provide help with python and limited help with matlab.
REQUIRED STUDENT BACKGROUND: Basic proficiency in programming language.

PROJECT 2: Surface energy balance and sensitivity of melt to climate change

STUDENTS: Morag Fotheringham, Anirudha Vijay Mahagaonka
ADVISOR: Regine Hock
DESCRIPTION: The project involves calculating the energy balance at the surface of a glacier using data over one melt season from an automatic weather station on a glacier. You will code the energy balance model from scratch (equations will be provided), or you can use a template code in Matlab. The next steps will evolve with student interests. For example, the project may compare different parameterizations for various components of the energy balance, or explore the sensitivity of energy components and glacier melt to melt to changes in meteorological input variables, or compare energy balance calculations to a simple degree-day melt model.
SOFTWARE REQUIREMENTS: Any programming language, e.g. Matlab. Project can in principle also be done using spreadsheets (e.g. Excel)
REQUIRED STUDENT BACKGROUND: Basic knowledge of any of the software requirements above.

PROJECT 3: A simple flowline model

STUDENTS: Myron Malisse Lumuus, Theresa Diener
ADVISOR: Regine Hock
DESCRIPTION: We will investigate the behavior of an idealized simple one-dimensional glacier model which simulates changes in glacier topography with time based on prescribed surface mass-balance gradients and equilibrium line altitudes. You will code the model in Matlab (based on a template), and then explore the behavior of the glacier in response to varying boundary conditions. You may also investigate the sensitivity of the glacier evolution in time to changes in glacier bed, ice softness etc.
SOFTWARE REQUIREMENTS: Matlab
REQUIRED STUDENT BACKGROUND: The project does not require prior ice flow modeling experience, but some basic knowledge of calculus/differential equations and Matlab is needed (or sufficient programing skills in another programing language in which case you would need to be able to code the model without help).

PROJECT 4: Kennicott Glacier geodetic mass balance: 1957-1978

STUDENTS: Chimira Andres, Albin Wells
ADVISOR: Sarah Child
DESCRIPTION: This project estimates the geodetic mass balance of Kennicott Glacier from 1957 to 1978 using 1950s USAF aerial photographs and structure-from-motion photogrammetry. To date, full areal coverage of calculated mass balance for Kennicott Glacier does not exist over this timeline. Introducing this timestep in Kennicott Glacier’s mass balance series will provide new insight into the temporal scale of the glacier’s degradation during the 20th century. Students will generate a 1957 DEM of the glacier from manually extracted ground control via present-day high resolution satellite imagery and DEMs. The resulting DEM’s hypsometry will be used to estimate glacier volume change and converted geodetic mass balance.
SOFTWARE REQUIREMENTS: For SfM processing: MicMac and CloudCompare (the install files will be provided if students are unable to do so ahead of time); any GIS software; Python, Matlab, or Excel for analysis
REQUIRED STUDENT BACKGROUND: Basic skills using any GIS software. A familiarity with optical imagery and SfM processing is beneficial, but not essential.

PROJECT 5: Kennicott Glacier geodetic mass balance: 1978-present

STUDENTS: Kamil Kachniarz, Brandon Tober
ADVISOR: Sarah Child
DESCRIPTION: For this project, ~40-year geodetic mass balance chances are estimated from observations of Kennicott Glacier covering 1978 to the present. This is accomplished by using a historical DEM to estimate ~40-year hypsometry–from present-day high-resolution satellite derived elevations–that is converted to mass balance using a density factor. The historical DEM is produced from high-altitude aerial photography collected by the NASA Johnson Space Center with structure-from-motion photogrammetric software. XYZ positions of exposed bedrock are determined from the present-day elevations for ground control used in the bulk bundle application. A DEM of Kennicott Glacier from 2000 will also be provided for mass balance comparisons of the later 20th century with the early 21st.
SOFTWARE REQUIREMENTS: For SfM processing: MicMac and CloudCompare (the install files will be provided if students are unable to do so ahead of time); any GIS software; Python, Matlab, or Excel for analysis
REQUIRED STUDENT BACKGROUND: Basic skills using any GIS software. A familiarity with optical imagery and SfM processing is beneficial, but not essential.

PROJECT 6: Satellite-based flow history of the lower Kennicott

STUDENTS: Emily Glen, Hannah Verboncoeur
ADVISOR: Mark Fahnestock
DESCRIPTION: Mapping the flow history of the lower Kennicott Glacier from the full satellite record. This project will use feature-tracking velocity fields from the ITS_LIVE archive (https://its-live.jpl.nasa.gov) to document the last several decades of ice flow changes in the lower Kennicott and tributary Root Glaciers, and will try to connect these changes to the evolving appearance of the ice in multi-decade satellite time series. Questions to be addressed: What are the primary changes in ice flow over the record, when did they occur, and what does this record suggest for the future? Students will work with ice flow data and satellite imagery starting with existing Python-based tools and QGIS.
SOFTWARE REQUIREMENTS: QGIS, Python environment with a number of libraries - contact instructor (mfahnestock@alaska.edu) for instructions on a conda-based installation that needs to be done prior to departure because of internet limitations in McCarthy.
REQUIRED STUDENT BACKGROUND: Some initial experience with Python and/or other data analysis (matlab, etc) programing helpful.

PROJECT 7: Seasonal flow variations in coastal and inland glaciers

STUDENTS: Karla Boxall, David Polashenski
ADVISOR: Mark Fahnestock
Understanding patterns of seasonal ice flow variation in coastal and inland glaciers using satellite-derived ice flow records. Archives like ITS_LIVE (https://its-live.jpl.nasa.gov) provide previously unavailable records of seasonal and longer-term variation, but capture ice displacement sampled non-uniformly in time. Students will work with time series of ice flow at points on several glaciers, with the goal of documenting the shape of the seasonal component of flow variation, how tightly constrained the transitions in flow speed through the year are, and what challenges are presented by these large, but potentially noisy, datasets. If time allows, spatial patterns of seasonal flow will be mapped. Students will work with ice flow data starting with existing Python-based tools.
SOFTWARE REQUIREMENTS: QGIS, Python environment with a number of libraries - contact instructor (mfahnestock@alaska.edu) for instructions on a conda-based installation that needs to be done prior to departure because of internet limitations in McCarthy.
REQUIRED STUDENT BACKGROUND: Some initial experience with Python and/or other data analysis (matlab, etc) programing helpful.

PROJECT 8: Feedbacks between orographic precipitation, subglacial erosion, and glacier flow

STUDENTS: Victor Devaux-Chupin, Joel Wilner
ADVISOR: Andy Aschwanden
DESCRIPTION: We will investigate how the feedbacks between orographic precipitation, which occurs when moist air is forced upwards over rising terrain, can lead to asymmetric ice flow if the ice overtops the mountain. The Antarctic Peninsula is an extreme case of these feedbacks because the peninsula extends north-south across the prevailing winds and precipitation changes from as high as 10 m/yr on the windward side of the mountain range to 10 cm/yr on the lee side, over less than 50 km. Despite the low snowfall on the leeward side, the northern Antarctic Peninsula has large trunk glaciers on the lee side. This asymmetry is driven by feedbacks between glacier dynamics and orographic precipitation. On long (10s of millennia) time scales, glacier erosion sculpts landscapes yet developing erosion theories is challenging because of the complex nature of the erosion processes. It is commonly assumed that the erosion rate is proportional to some power of the basal sliding velocity. In this project we will use the combination of a model of orographic precipitation and an ice flow model that includes erosion on an idealized geometry to better understand these feedbacks. We will also use the flowline model by Brinkerhoff et al (2017, Nat. Comm.) as the basis for the project.
SOFTWARE REQUIREMENTS: Familiarity with Python. Students should download the model code from Brinkerhoff et al. (Suppl. Software) and make sure the necessary libraries are installed on their laptops. The code of the coupled model is currently being update and will be posted here: https://github.com/aaschwanden/flowline-glacier-model
REQUIRED STUDENT BACKGROUND: Proficiency in Python. Experience with finite element methods is a plus, but not required.

PROJECT 9: Estimating the parameters of a temperature index model using Markov-chain Monte Carlo methods

STUDENTS: Coline Bouchayer, Eric Gagliano
ADVISOR: Andy Aschwanden
DESCRIPTION: Are you interested in exploring probabilistic programming methods and parameter estimation using the python package “PyMC3”? So does the instructor. First we will familiarize ourselves with the concept of probabilistic programming by recasting linear regression in a probabilistic framework. Instead of finding the slope and intercept that best fit the data, probabilistic methods seek to find the posterior parameter distributions that best fit the data. Then we will replace the linear model with a slightly more complicated model frequently used in glaciology: a temperature index model. Temperature-index models remain widely used to calculate surface melt because of their simplicity, yet empirical parameters such as the melt factors for snow and ice are not well constrained. In this project we will estimate the parameter distributions that best explain the surface melt calculated by a more sophisticated energy balance model. A reanalysis for 1980-2020 Greenland air temperature, precipitation, surface melt and mass balance are provided by the regional climate model HIRHAM5.
SOFTWARE REQUIREMENTS: Python with pymc3 and xarray installed.
REQUIRED STUDENT BACKGROUND: Interest in or basic knowledge of statistics. Some experience with Python will be helpful.

PROJECT 10: Recovering surface mass balance from kinematic observations

STUDENTS: Emily Glazer, Veronica Tollenaar
ADVISOR: Ed Bueler
DESCRIPTION: Remote observations of glaciers using visible and radar imaging allow increasingly-routine determination of ice surface elevation and ice surface velocity. On the other hand, surface mass balance (SMB) rates can be hard to observe remotely, for instance in ablation areas and/or firn areas with ice lenses. The surface kinematical equation (SKE) is worth understanding in this regard, in addition to its application to modeling glacier evolution. We will use the SKE to invert time-dependent synthetic observations to extract SMB, a feat attempted before (e.g. Gudmundsson & Bauder, 1999). A newly-discovered form of the SKE makes the topic worth a second look. Simplified 1D geometry and the linear nature of the problem make it achievable in a week-long project.
SOFTWARE REQUIREMENTS: Matlab or Octave or Python/numpy, running locally on your machine.
REQUIRED STUDENT BACKGROUND: Exposure to differential equations and linear algebra.

PROJECT 11: Stokes modeling of glacier velocity: the effects of surface bumps

STUDENTS: Philipp Arndt, Davor Dundovic
ADVISOR: Ed Bueler
DESCRIPTION: How does the velocity field of a glacier change if you put a pile of ice somewhere on its surface, or if you dig a pit somewhere? The change in the velocity field reveals the boundary Green's functions of the Stokes problem for glaciers. The goal of this project is to build new understanding of these Green's functions. How do they depend on glacier thickness or sliding? How do they depend on nearness to margins or bed topography? How can they be exploited to build faster numerical models? Students will use a modern finite-element framework for solving the Glen-Stokes equations, in a flowline plane, for student-chosen glacier geometries, as we explore.
SOFTWARE REQUIREMENTS: Python/numpy, running locally on your machine. Ideally, functional Firedrake, Gmsh, and Paraview installations. (But I'll bring an extra laptop, pre-loaded.)
REQUIRED STUDENT BACKGROUND: Exposure to linear algebra, numerical methods, partial differential equations, and (ideally) the finite element method.

PROJECT 12: The influence of climate on the tidewater glacier cycle

STUDENTS: Hui Gao, Ann-Sofie Zinck
ADVISOR: Martin Truffer
DESCRIPTION: The tidewater glacier cycle is a dynamic cycle consisting of relatively slow advance followed by rapid retreat. A flowline model by Brinkerhoff et al (2017, Nat. Comm.) shows how the interaction between ice and water flow with subglacial sediment dynamics can lead to such cyclicity even under constant climate. In this project students will investigate how the tidewater glacier cycle is affected by climate change. Another extension of the Brinkerhoff model will explore the influence of seasonality on the cycle.
SOFTWARE REQUIREMENTS: Familiarity with Python. Students should download the model code from Brinkerhoff et al. (Suppl. Software) and make sure the necessary libraries are installed on their laptops.
REQUIRED STUDENT BACKGROUND: Proficiency in Python. Experience with finite element methods is a plus, but not required.

PROJECT 13: Formation of gives

STUDENTS: Naomi Ochwat, Yoram Terleth
ADVISOR: Martin Truffer
DESCRIPTION: Ogives form a wave-like pattern of topography at the bottom of ice falls. The formation appears to have seasonality with one new wave formed each year. We will use a simple kinematic model to investigate the formation of ogives and the sensitivity of their formation to various environmental and geometrical factors.
SOFTWARE REQUIREMENTS: Matlab, Python or similar software
REQUIRED STUDENT BACKGROUND: Some familiarity with Matlab, Python or similar software.

PROJECT 14: Modeling subglacial hydrology of a mountain glacier

STUDENTS: Anna-Mireilla Hayden, Johanna Klahold
ADVISOR: Aleah Sommers
DESCRIPTION: The subglacial hydrological system evolves dynamically, draining meltwater beneath the ice, and influencing ice sliding velocity via effective pressure. In this project, we will explore seasonal subglacial drainage using the SHAKTI model (Sommers et al., 2018, GMD). Depending on interest, the project will focus on 1D simulations to explore the subglacial response to different meltwater forcings in an idealized setting and/or 2D simulations applied to a real glacier (Kennicott Glacier or otherwise, to be selected before the course). Experiments will be designed, run, and analyzed by the students to simulate the subglacial evolution forced by a variety of spatial and temporal inputs of meltwater to the bed, which may include moulins, distributed inputs, and/or lake drainage events.
SOFTWARE REQUIREMENTS: Matlab (required), ISSM binaries (optional, required for 2D simulations; Ice-sheet and Sea-level System Model, available at https://issm.jpl.nasa.gov/download/)
REQUIRED STUDENT BACKGROUND: Some basic knowledge of fluid mechanics, partial differential equations, and experience with Matlab will be helpful. You will not need to write the model from scratch, but will modify existing code. To become familiar with SHAKTI in the ISSM framework for 2D simulations, please download the ISSM binaries ahead of time and go through the SHAKTI tutorial (https://issm.jpl.nasa.gov/documentation/tutorials/shakti/).