Research

Applications of cosmogenic nuclides

My research uses cosmogenic nuclides, rare nuclides produced through interactions with cosmic rays, on the Earth's surface. I primarily apply in situ 14C, 10Be and 26Al, which can all be extracted from the common mineral quartz.

In particular, in situ 14C is a powerful radionuclide since its short half-life (5.7 kyr) makes it sensitive to exposure over the last 25-30 kyr, as well as to periods of burial by ice.

Cosmogenic nuclide "dipsticks"

Rapid thinning of the Laurentide Ice Sheet

Field photos from New England Mountains

Compositionally dependent in situ 14C production rate calculator

Investigating the Holocene history of the Riukojietna ice cap, Sweden

Holocene ice surface lowering in western Dronning Maud Land, East Antarctica

Cosmogenic nuclide "dipsticks"

The concept is pretty simple - as an ice sheet thins, the top of the nunatak (mountain that sticks out of an ice sheet) is exposed first while the bottom of the nunatak is exposed last. Therefore, samples collected from the top of the nunatak should have the oldest exposure age while the bottom samples should have the youngest exposure age.

Nunatak in East Antarctica

Conceptual diagram of a dipstick

Boulder or bedrock samples along mountains/nunataks can then be measured for the amount of cosmogenic nuclides (in situ 14C, 10Be and 26Al) that have built up over time (amount of atoms per gram). If the production rate is well known (atoms per gram per year), then the exposure age can be calculated. The exposure ages reveal the past timing of ice surface lowering while the elevation transect reveals the rate of ice surface lowering.

Rapid thinning of the Laurentide Ice Sheet

During my M.S. work at Boston College, I collected boulder and bedrock samples along Mount Washington (NH), Acadia National Park (ME), and Mount Mansfield (VT). I dated samples primarily with 10Be and 26Al, though I used in situ 14C for a few samples on Mount Washington.

(left) Laurentide ice sheet deglaciation ages in New England. Isochrons are from North American varve chronology of deglaciation (from fig. 12 in Ridge et al., 2012). Black dots are moraines, and black triangles are mountain summits and slopes dated with 10Be. Heavy black line is the maximum extent of the Laurentide ice sheet (Dyke et al., 2003). ( Fig. 1 from Koester et al., 2020).

Conceptual diagram of ice surface lowering at Mount Washington

The cosmogenic nuclide data from the summit of Mount Washington indicate it was exposed early during the last deglaciation or at least for several thousand years before subsequent withdrawal of ice from the lowlands. Below the summit, the data showed rapid thinning of the ice surface between 15 and 13 ka, centered on 14.1 ± 0.9 ka. I made this conceptual diagram, though it wasn't included in the publication.

Field photos from New England Mountains

Below Mount Washington

This picture shows a good example of some bedrock we sampled below the summit of Mount Washington on the "lawn". The summit is in the background to the left.

Quartzite bedrock

Here I am taking the strike and dip of a surface before sampling.

On the auto road

Although we did hike up and down Mount Washington, we also took advantage of the auto road to access locations for sampling.

Mt. Pemetic

We sampled coastal mountains in Acadia National Park such as Mt. Pemetic. The granite boulders made our work a little easier.

Jordan Pond

We sampled near the shore of Jordan Pond for the base of our dipstick. (Read our paper to learn why I wouldn't sample there again.)

Glacial erratics galore

I got spoiled by this field area because its was easy to spot beautiful granite glacial erratics. Granite contains an abundance of quartz!

Compositionally dependent in situ 14C production rate calculator

Although methods to extract and interpret in situ 14C from quartz are well established, the ability to extract in situ 14C from quartz-poor rocks would open new avenues of research to a broader array of landscapes. As a precursor to sample extraction from non-quartz samples, I developed a compositionally dependent software framework in MATLAB to calculate the theoretical 14C production from non-quartz materials (Koester and Lifton, 2023).

I calculated the sea-level high-latitude (SLHL) compositionally dependent production rates for selected silicate minerals and rock types (see figures below).

The pie charts above show the amount of production from each element. Spallation production from oxygen (O) dominates throughout the compositional range at SLHL, making up >90% of overall in situ 14C production. The remaining production from other elements is difficult to see in the pie chart, and are shown as bar graphs to the right of each mineral. Note that all bar graphs are 5% and below, expect for Albite.

Predicted production rates for the investigated silicate minerals and rocks were generally less than quartz, except for albite due to significant production from Na. The colors refer to the element that corresponds to the highest proportion of production contribution after O and Si (Fig. 3 from Koester and Lifton, 2023).

Investigating the Holocene history of the Riukojietna ice cap, Sweden

Riukojietna ice cap is the last remaining ice cap in Sweden. Its mass balance has been significantly decreasing over the 20th century because it is relatively low-lying and sensitive to climate change. In 1989, its area was 4.6 km2 and extended in elevation between 1140 and 1456 m a.s.l.. However, the ice cap has shrunk. In 2015, its area covered 2.8 km2 which spanned less than 283 m between the ice divide and terminus.

Satellite imagery of the Riukojietna ice cap in 1960 (left), 2008 (middle), and 2015 (right).

During the Holocene Thermal Maximum (HTM; 8.2 – 4.2 ka), summer temperatures in Scandinavia were 1.5 – 2 ºC warmer than present, suggesting that glaciers may have disappeared or were reduced in size during this time. I combined 10Be–26Al–14C concentrations from bedrock collected directly adjacent to the Riukojietna Ice Cap, Sweden, to determine if the ice was smaller than present at any point since the LGM. I combined these results with sediment cores from proglacial lakes downstream of the ice cap that document Holocene glacier fluctuations. Results indicate that the ice cap was likely smaller than today during the Holocene; however, proglacial sediment records indicate that the ice cap likely survived the HTM. Coming to a publication near you soon!

Five bedrock samples were collected surrounding the Ruikojietna ice cap. Two samples were collected at high elevations, two from a bedrock knob that has recently been emerged in 2011, and one bedrock sample near Lake 1063.

Holocene ice surface lowering in western Dronning Maud Land, East Antarctica

Constraining past ice surface elevation changes of the East Antarctic Ice Sheet (EAIS) is essential to understand its response to climate change and contribution to global sea level change. Empirical evidence of how the Antarctic Ice Sheet (AIS) responded to past climate changes, such as cosmogenic nuclides, can therefore improve future numerical models of AIS behavior. MAGIC-DML (Mapping/Modeling/Measuring Antarctic Geomorphology Ice Change in Dronning Maud Laud) is an international collaboration that brings together data people (like me) with excellent ice sheet modelers to reconstruct the ice history in western Dronning Maud Land. Before the project began in 2017, little data existed in our study location.

The majority of the in situ 14C data are Holocene in age compared to the 10Be and 26Al.

My contribution to this project has been to analyze in situ 14C exposure ages from mountains protruding through the margin of the EAIS along the marine-terminating Jutulstraumen ice stream in western Dronning Maud Land (DML) to constrain ice sheet thinning since the Last Glacial Maximum (LGM).

In situ 14C more closely reflects the post-LGM deglacial signal in Antarctica than long-lived nuclides such as 10Be and 26Al because 14C is less likely to preserve inheritance (prior nuclides) under minimally-erosive ice (left).

We analyzed an ensemble of high-resolution experiments using the ice flow model Úa. Experiments were driven by a suite of climate models under a range of different model parameters, to produce Last Glacial Maximum (LGM) ice geometries and evaluate the outputs with in situ 14C data and ice core data. Our best-fit model shows significant (800 – 900 m) coastal thickening compared to present-day that has not been predicted by models to date in Dronning Maud Land. The in situ 14C exposure ages indicate Jutulstraumen ice stream experienced 850 m of coastal thinning and 300 m of inland thinning between 9.5 - 1 ka. Coming to a publication near you soon!

Page updated

Google Sites

Report abuse