Hazards in the Caribbean: The history of magma chambers, eruptions, landslides, streams, and fumeroles in Dominica (Keck Frontier Project)

What: This Frontier Project is an interdisciplinary project based on the volcanic island of Dominica in the Lesser Antilles. There will be two weeks of field work in Dominica, examining and sampling pyroclastic deposits, lava domes, landslide deposits, and hydrothermal waters and gasses. Our goal is to improve the understanding of the history and periodicity of these various geologic processes and how that informs the hazard potential on the island. The second half of the project will be sample preparation and initial analyses at Union College. Students will have access to rock processing equipment, an SEM with BSE and CL capabilities, an ICP-MS (laser ablation and fluid sample introduction), alkalinity titrators, ion chromatographs, and the stable isotope lab.

 Tentative Field/Laboratory Dates: June 11-June 25 (Dominica); June 26 – July 7 (Union College)

Where: Dominica, Lesser Antilles; Union College, Schenectady, NY

 Who: Holli Frey (Union College), Erouscilla Joseph (University of West Indies Seismic Research Center), Amanda Schmidt (Oberlin College), Laura Waters (Smithsonian Institute), and 16 students

Figure 1. Dominica is located in the central portion of the Lesser Antilles volcanic arc (map from Lindsay et al., 2005a). The geologic map of Dominica shows <1 Ma lava dome and ignimbrite deposits (modified from Smith and Roobol, 2013).

Figure 1. Dominica is located in the central portion of the Lesser Antilles volcanic arc (map from Lindsay et al., 2005a). The geologic map of Dominica shows

 Project Overview and Goals:  The Caribbean has been the site of significant historic volcanism, from the ongoing eruptions in Montserrat to the devesating eruptions of Mt. Pelee, Martinique (~ 32, 000 fatalities), and Soufriere, St. Vincent in 1902 (~1,500 fatalities). However, the island with the most volcanic hazard risk is Dominica, which experienced the largest explosive eruption in the Caribbean in the last 200 kyr (~58 km3) and features nine potentially active volcanic centers that are Pleistocene or younger in age (Carey and Sigurdsson, 1980; Lindsay et al., 2005a), and had phreatic eruptions in the Valley of Desolation in 1880 and 1997 (Fig. 1). The rugged landscape of Dominica has also been shaped by its tropical climate and landslides, caused by storms like Tropical Storm Erika in 2015. Today, Dominica is known as the Nature Island of the Caribbean and significant efforts have been made to increase tourism. There are numerous eco-tourism sites and the Waitukubuli National Trail, an island-wide 185 km hiking trail, was recently completed. The tourism sites feature active geothermal areas, with fumeroles and steam vents, as well as older lava flows and explosive deposits of pumice and ash. Recent shallow seismic swarms in northern and southern Dominica may be indicative of volcanic unrest.

The goals of this project are to understand how different processes have shaped the Dominican landscape and history in the past and how they may manifest in the future. We hope to elucidate the complicated explosive history of the island through geochronology and petrologic studies. We will build upon the work done by the UWI SRC by characterizing meteoric and hydrothermal waters, as well as volcanic gasses, from fumeroles to contribute to baseline monitoring of volcanic activity for use in detecting volcanic unrest. We will also study how the landscape has affected the settlement of people and the history of landslides.

Ignimbrite deposits on Dominica’s northwestern coastline.

Ignimbrite deposits on Dominica’s northwestern coastline.

Geologic setting:  Dominica is a 750 km2 island of rugged topography and pristine rainforest which features nine volcanic centers that are <2.6 Ma in age (Lindsay et al., 2005a). Dominica’s coastlines and interior valleys abound with thick (>20 m) ignimbrite deposits, composed of pumice clasts, rock fragments, and ash from solidified pyroclastic flows (Fig. 3). There have been no documented explosive eruptions of large magnitude in Dominica in the last 20 kyr. The most recent activity involving magma was a lava dome collapse. Arcaheological investigations in the late 1970s unearthed clay pots beneath an ash horizon near the village of Soufrierre. Charcoal within the ash was dated at 450 ± 90 years B.P. (Roobol et al., 1983). Dominica’s most recent volcanic activity was several explosions of steam and ash violently ejected from hydrothermal vents. Phreatic eruptions ocurred in the Valley of Desolation in 1880 and 1997, covering an area <1 km2 with a thin (~2 cm) layer of ash (Lindsay et al., 2005b). The Valley of Desolation is an active hydrothermal area and popular hiking destination of tourists (Fig. 3). At the far end of the valley lies Boiling Lake, a 75 m volcanic lake that is is typically very hot (80-90 ˚C) and acidic (pH of 3-5). In addition to the older volcanic deposits and active hydrothermal areas, Dominica’s volcanic present is recorded by earthquake swarms. The earthquakes are typically shallow (<5 km) and of fairly low magnitude (<4.0), often ocurring in rapid succession or swarms.

Potential student projects:  In order to better understand the magma plumbing system beneath Dominica, the potential for hazards, and how volcanism and landslides have shaped the landscape, we propose a variety of potential student projects that span several disciplines and research questions. The projects determined by the students may depend in part on the instrumentation, facilities, and faculty expertise at the home institution. Much of the analytical work can be done at Union for some projects, whereas others may require significant data collection during the fall term at national labs like U of AZ Laserchron facility or USGS-Stanford SHRIMP.

Grand Savanne Ignimbrite

Grand Savanne Ignimbrite

1. Petrology (2-6 students)

Although we have done some work on the chemistry, petrology, and ages of the ignimbrites, the plumbing system from which these deposits emanated and their potential relationship to the lava domes is still largely unresolved. Continued work will allow us to determine if the explosive and effusive deposits are from a single batholithic magma chamber and erupted over a short time period (Smith et al., 2013) or if the eruptions are derived from more discrete and isolated magma batches over an extended time scale. This has implications for hazards, with respect to larger catastrophic eruptions versus more frequent smaller eruptions. There have been very few detailed petrologic/geochemical studies of the lava domes, which are predominantly andesitic and have enclaves (Crampe et al., 2015), which are indicative of magma mixing.

  • Investigate the petrology of an individual lava dome using petrography, geochemistry, and mineral chemistry to constrain pre-eruptive magmatic conditions.
  • Comparative effusive deposits and explosive deposits that appear to emanate from the same vent and try to determine why the eruptive style changed (i.e. determine pre-eruptive water contents with plagioclase hygrometry or temperature with two oxide thermometry).
  • Characterize a particularly mineral phase (i.e. plagioclase, pyroxene, or apatite) and compare the textures, zoning, and compositions from different deposits throughout the island to address the potential interconnectedness of the magmatic plumbing system.
  • Determine volatile concentrations in melt inclusions to assess whether volatile saturation was an eruption trigger.

2. Detrital Geochronology (1-2 students)

Prior to our work, only two detrital zircons had been found in Dominica (Eocene age; Howe et al., 2015). However, we have found more than a dozen detrital zircons ranging in age from 50 Ma to 1.8 Ga. This is an exciting result, which needs a much more robust dataset. It has implications for the tectonic history of the Caribbean and the transport of detrital zircons from South America. Likely shed from mountainous terrains and shed into rive basins, the zircons were incorporated by ascending magmas that erupted explosively at the surface.

  • Determine the provenance and probability distribution of detrital zircons using U-Pb dating. Are the different zircon age populations distinctive in morphology, size, cathode-luminescence, or chemistry?
  • Analyze Hf isotopes and determine model ages for different zircon populations. What can we learn about crustal contributions versus derived mantle sources over time?

3. Geochronology < 350 ka (1-2 students)

We have done significant U-Th dating of zircon rims in most of the young ignimbrites which has yielded a mixture of crystallization ages (Brehm et al., 2015) and what we believe are eruption ages. The age populations are polymodal, suggesting a complex magmatic history.

  • Determine U-Th/(He) eruption ages from ignimbrites dated by U-Th to compare the dating techniques and refine the eruptive history of the explosive deposits
  • Determine U-Th ages of zircon cores or do depth profiling and compare to rim ages. This will give insight into the longevity of the magma chamber, crystal recycling, and potentially a shared history between deposits.

4. Physical Volcanology (1 student)

Throughout the island, there are numerous thick pyroclastic flow sequences and a few airfall deposits. These likely emanate from several vents on the island. Characterization thus far has been mostly limited to chemistry.

  • Compare the stratigraphy of different deposits and correlate units between deposits. This could involve sampling of different pumice clasts, with texture, porosity, vesicle size/shape examined on a SEM. One could also use a geochemical or mineralogical approach to compare units.

5. Aqueous Geochemistry (2-3 students)

There are >365 streams and >30 hydrothermal areas. Over the last three years, we have begun an island-wide field, chemical, and isotopic characterization of meteoric and hydrothermal waters (Metzger et al., 2015; DeFranco et al., 2016; Metzger et al., 2016). The prior stream coverage, particularly with respect to carbon isotopes (Rive et al., 2013), was limited, so our data is the basis for a comprehensive island-wide baseline for future chemical monitoring. With multiple years of data, we can start to ascertain normal variations, drought induced changes, the effects of Tropical Storm Erika (August 2015), and the possible sensitivity or reaction to seismic swarms.

  • What is the extent of the island where water has a hydrothermal signature? Is this best determined by field characterization (T and pH) or chemical tracers (Li, B, SO4) or stable isotopes (C, O, D)?
  • Have the hydrothermal areas changed since 2000-2006 (Joseph et al., 2011), during which time volcano seismicity increased in northern Dominica?
  • How do extremophile organisms (black and white filamentous ooze) affect the water chemistry and isotopic compositions?
  • Was the stream chemistry affected by Tropical Storm Erika in August 2015 and were the effects regional or island-wide? How long are the effects of such a catastrophic event observed, i.e. do the 2016 and/or 2017 datasets show greater elemental flux from weathering surfaces?
Valley of Desolation

Valley of Desolation

 6. Gas Chemistry (1-2 students)

Volcanic gases from passively degassing volcanoes have been widely studied over the past 20 years, as changes in their chemical composition could provide information to detect new intrusion of magma as a precursor of an eruption. There is a very limited amount of data on gas composition collected from hydrothermal systems in Dominica (Joseph et al., 2011). This project aims to obtain updated information on the gas composition of fumeroles and bubbling springs across the island, for comparison to previous samples.

  • Have the gas composition of hydrothermal areas changed since 2000-2006 (Joseph et al., 2011), during which time volcano seismicity increased in northern Dominica?
  • What does the gas data tell us about current reservoir temperatures, and subsurface magmatic activity?

7. Landscape Evolution (4-5 students)

In 2011, a landslide dam was breached, causing major flooding of the Layou River and the need for dredging, which continues to the present day. Prior to Tropical Storm Erika, a landslide assessment study of Dominica was done (Westin, 2015). In Schmidt’s prior work in China, students have used short-lived fallout radionuclides (137Cs, 210Pbex, and 7Be) to fingerprint relative depth of upstream erosion and sediment sources contributing to detrital river sediment.

  • What is the distribution of landslides caused by TS Erika? How did that compare to the assessment study? How can predictive models be improved?
  • Did landslide frequency increase after TS Erika during less extreme rainfall events?
  • How have landslides altered the landscape in Dominica over time?
  • Do watersheds with extensive upstream landsliding have different short-lived isotopic signatures than watersheds without landslides?
  • What are local floodplain sedimentation rates (determined using 137Cs) in areas downstream of watersheds with and without extensive active landslides?
  • Do watersheds with and without widespread landsliding have different clay composition in the downstream sediment? Is the sediment in detrital samples coming from surface or more recently exposed (and thus less weathered) material?

Logistics/Field Conditions:  Frey has been doing fieldwork in Dominica with students for the last 4 years and Joseph since 2000, and thus both are quite familiar with the concerns of doing fieldwork in the Caribbean. There are no venomous insects/snakes or aggressive fauna in Dominica. The main safety concern is mosquito-borne illnesses such as zika and chikungunya. Anecdotally, no one in our research groups has contracted these illnesses and mosquitoes are rarely sighted in our field areas or lodging site. Government sponsored fogging and education about mosquito breeding grounds and insect bite prevention has contributed to the reduction in cases. As a preventative measure, we regularly use DEET repellant, and wear long pants/shirts and/or permethrin treated clothing.

Effects of Tropical Storm Erika

Effects of Tropical Storm Erika

Students will fly into Melville Hall/Douglas Charles Airport in Dominica. While in Dominica, students will stay at the Springfield Guest House Archbold Tropical Research Center operated by the University of Clemson. The center has lab space, a classroom, and wi-fi. All meals are prepared on-site by the staff. In the field, students should be prepared for hot, humid weather, with temperatures 80-100˚F and potential for significant rain. The terrain is rugged and there may be a few days of long hikes depending upon the student project. However, most sampling sites are readily accessible by vehicle.

Students will fly from Dominica to Albany, NY. While at Union, students will reside in University housing on campus and be given a per diem to use either in the dining center (limited hours) or for groceries. During this time, students will be processing samples, running analyses, and/or becoming more familiar with the literature pertaining to their project.

Recommended Courses/Prerequisites:  The potential projects described above require familiarity with concepts introduced in courses focused on igneous petrology, volcanology, geochemistry, geomorphology and GIS/remote sensing. Previous coursework in these areas is encouraged for participation in this project. In your application essay, please indicate which project area is of most interest to you and rank your top 3 choices. Also discuss any analytical equipment or expertise of your home advisor, which may be beneficial to your project.

Student Expectations:  If you are selected to participate in this project, you are expected to be fully engaged in the field and laboratory, both for your own project and to assist others as needed. Following the summer experience, you will continue the project at your home institution under the guidance of a home advisor and the project directors. Each student is expected to produce a thesis or comparable work during the academic year. All students will be submitting an abstract and presenting at the 2017 Fall AGU meeting, held in New Orleans (mid December). At the meeting, we will have a workshop to discuss our data, and the directors will assist students in planning a meeting schedule and interfacing with potential graduate advisors.

Contact Info

Prof. Holli Frey (freyh@union.edu)

Dr. Erouscilla Joseph (pjoseph@uwiseismic.com)

Prof. Amanda Schmidt (Amanda.Schmidt@oberlin.edu)

Dr. Laura Waters (WatersL@si.edu)


Carey, S.N., Sigurdsson, H. (1980) The Roseau ash: deep-sea tephra deposits from a major eruption on

Dominica, Lesser Antilles arc. Journal of Volcanology and Geothermal Research 7: 67-86.

Dominica (Lesser Antilles). Journal of Volcanology and Geothermal Research 153: 200-220.

Howe, T.M., Lindsay, J.M., and Shane, P. (2105a) Eruption of young andesitic-dacitic magmatic systems

beneath Dominica, Lesser Antilles: Journal of Volcanology and Geothermal Research 297:9-88.

Howe, T.M., Schmitt, A.K., Lindsay, J.M., Shane, P., and Stockli, D.F (2015b) Time scales of intra-

oceanic arc magmatism from combined U-Th and (U-Th)/He zircon geochronology of Dominica,

Lesser Antilles: Geochemistry, Geophysics, Geosystems 16:347-365.

Howe, T.M., Lindsay, J.M., Shane, P., Schmitt, A.K., and Stockli, D.F. (2014) Re-evaluation of the

Roseau Tuff eruptive sequence and other Ignimbrites in Dominica, Lesser Antilles. Journal of Quaternary Science 29(6): 531-546.

Joseph, E.P., Fournier, N., Lindsay, J.M., Fischer, T.B. (2011) Gas and water geochemistry of geothermal

systems in Dominica, Lesser Antilles island Arc, Journal of Volcanology and Geothermal Research. 206:1-14.

Lindsay, J.M., Stasiuk, M.V., Shepherd, J.B. (2003) Geological history and potential hazards of the late-

Pleistocene to Recent Plat Pays volcanic complex, Dominica, Lesser Antilles, Bulletin of Volcanology 65: 201–220.

Lindsay, J. M., Trumball, R.B., Siebel, W. (2005a) Geochemistry and petrogenesis of Late Pleistocene to

recent volcanism in southern Dominica, Lesser Antilles. Journal of Volcanology and Geothermal Research 148: 253-294.

Lindsay, J.M., Smith, A.L., Roobol, M.J., and Stasiuk, M.V. (2005b) Dominica: Volcanic Hazards of the Lesser Antilles. Seismic Research Unit, The University of the West Indies, Trinidad and Tobago.

Rivé, K., Gaillardet, J., Agrinier, P. and Rad, S. (2013) Carbon Isotopes in the Rivers from the Lesser

Antilles: Origin of the Carbonic Acid Consumed by Weathering Reactions in the Lesser Antilles. Earth Surface Processes and Landforms 38: 1020-035.

Roobol MJ, Wright JV, Smith AL (1983) Calderas or gravity-slide structures in the Lesser Antilles island

arc? Journal of Volcanology and Geothermal Research 19: 121-134.

Sigurdsson, H. (1972) Partly-welded pyroclastic flow deposits in Dominica, Lesser Antilles, Bulletin of

Volcanology 36: 148-163.

Smith, A.L., Roobol, M.J., Mattioli, G.S., Fryxell, J.E., Daly, G.E., Fernandez, L.A. (2013) The Volcanic

Geology of the Mid-Arc Island of Dominica, Lesser Antilles—The Surface Expression of an Island-Arc Batholith. The Geological Society of America, Special Paper, 496.


Student Abstracts (*undergraduate)

Brehm, S.*, Frey, H.M., and Manon, M.R.F. (2015) Using trace element geochemistry of apatite

phenocrysts to fingerprint tephras in Dominica: Northeastern section Geological Society of America Abstracts with Programs, Vol. 47.

Brehm, S.*, Frey, H.M., and Manon, M.R.F. (2014) Hornblende in dacitic ignimbrites from central

Dominica: implications for magma plumbing systems: Northeastern section Geological Society of America Abstracts with Programs, Vol. 46, No. 2, p. 120.

Crampe, E.*, Frey, H.M., and Manon, M.R.F. (2015) Whole-rock geochemistry and petrography of the

Morne aux Diables volcanic center, Dominica: Northeastern section Geological Society of

America Abstracts with Programs, Vol. 47.

DeFranco, K.*, Frey, H.M. and Manon, M.R.F. (2016) Establishing island-wide water characterization in

a volcano-hydrothermal system in Dominica, Lesser Antilles. Northeastern section Geological Society of America Abstracts with Programs. Vol. 48, No. 2.

Flake, A.* and Frey, H.M. (2014) The Roseau and Layou ignimbrites of Central Dominica, Lesser

Antilles: Northeastern section Geological Society of America Abstracts with Programs, Vol. 46, No. 2, p. 120.

Frey, H.M., and Manon, M.R.F., Brehm, S.* (2015) Fingerprinting young ignimbrites in Dominica

(Lesser Antilles): Constraints from bulk REEs, apatite chemistry, and U-Th zircon chronology: Abstract V36-01 presented at 2015 Fall Meeting, AGU, San Francisco, Calif., 14-18 Dec

Frey, H.M., Manon, M.R.F., Brehm, S.*, and Flake, A.* (2013) Monotonous crystal-rich dacitic

ignimbrites from Dominica: Implications for magma recharge: Geological Society of America

Abstracts with Programs, Vol. 45.

Kittross, S.*, Frey, H.M., and Manon, M.R.F. (2016) LA ICP-MS U-Pb analysis of inherited zircons from

Dominica, Lesser Antilles. Northeastern section Geological Society of America Abstracts with Programs. Vol. 48, No. 2.

Main, L.* and Frey, H.M. (2014) The andesitic ignimbrites of northern Dominica, Caribbean:

Northeastern section Geological Society of America Abstracts with Programs, Vol. 46, No. 2, p. 120.

Metzger, T.*, Frey, H.M., Manon, M.R.F., and DeFranco, K.* (2016) Island-wide isotopic analysis of

meteoric and hydrothermal waters in Dominica, Lesser Antilles. Geological Society of America Abstracts with Programs. Vol. 48, No. 7.

Metzger, T.*, Frey, H.M., and Manon, M.R.F. (2015) Water geochemistry analysis of streams and

hydrothermal waters in Dominica, Lesser Antilles. Northeastern section Geological Society of America Abstracts with Programs, Vol. 47.