Alan Whittington - Research Interests

I regard the combination of field- and lab-based studies is an essential component of research in the geosciences. Doing both enhances my understanding of geological processes and increases my enjoyment of both kinds of work.
My research has spanned crustal melting and granitic plutonism in orogenic belts; shallow basaltic sheet emplacement in sedimentary basins; measurements of the viscosity, other physical properties and thermodynamics of silicate liquids; metamorphic petrology in high-grade metamorphic rocks. Research field areas to date include the western Himalaya, Antarctica, the Rocky Mountains, Brazil and Guatemala.
My lab research is focused on measuring the physical properties of silicate melts, particularly viscosity.

Recent and ongoing projects include:

• Rheology of dacite-rhyodacite block and ash "stealth" flows: Santiaguito Dome, Guatemala
• Transport properties of silicate glasses, liquids and magmas
• Crustal contamination of basaltic magma / Emplacement of magma at Spanish Peaks, Colorado
• Intrusive Architecture and Flow Directiions in the Ferrar Large Igneous Province, Southern Victoria Land

Past research includes studies of continental collision and crustal melting in two quite different areas:

• Nanga Parbat, Pakistan - exhumation, metamorphism and melting in the western Himalayan syntaxis
• The Araçuaí Belt, Brazil - metamorphism and magmatism in a "confined" Neoproterozoic orogenic belt

I also pay attention to the geology of wherever I happen to be, and have some nice pictures as a result:

• What I did on my holidays

For more details, read on...

The Lab

Group members, September 2005. From L to R: Dr. Harald Behrens (visiting from University of Hannover); Geoffroy Avard (PhD student); Urmidola Raye (PhD student), Bastian Joachim (back) and Andre Stechern (front), both visiting students from University of Hannover; Fred Davis (undergraduate), Bridget Hellwig (MS student), Jackie Getson (MS student), Jen Cooper (MS student), and Dr. Alan Whittington (Assistant Professor). Not pictured: Mike O'Malley (MS student), Brooks Hafey (undergraduate).

Recent and ongoing projects

Ph.D. student: Geoffroy Avard (in prog)
M.S. students: Jackie Getson (in prog), Bridget Hellwig (in prog), Mike O'Malley (in prog)

We are investigating the rheological and thermodynamic properties (particularly viscosity and heat capacity) of hydrous dacitic to rhyodacitic magmas from an active subduction zone. Results will include measurements of the actual temperature and water contents of silicic block-lavas at Santiaguito volcano in Guatemala, and their viscosity and heat capacity as a function of temperature, water content and bulk-rock chemistry. The results will be applied directly to refine existing models for the emplacement of slow-moving block lava flows, and will also be used to develop a predictive model of the viscosity of hydrous (rhyo-)dacitic liquids based on configurational entropy theory, providing important data required for improving current physical and thermal models of magmatic and volcanic processes.

Santiaguito erupting on 26th March, 2005. Looking north from the well leveed 2001 flow. Note pyroclastic flow to E and rock avalanches to SW of Caliente.

Key scientific objectives and questions to be answered by this project include:
(i) Determine the viscosity and heat capacity of rhyodacitic to dacitic magmas as a function of temperature, bulk composition and X(H2O). Determine how they differ from those of andesitic magmas, which do not form block lava flows, whether there is a “critical melt structure” at which these changes occur and whether the effect of water is a simple function of anhydrous composition.
(ii) Integrate the results of (i) with calorimetric and viscosimetric data for synthetic liquids previously collected by Whittington and others, to develop a predictive model for the viscosity of hydrous dacitic and related liquids based on configurational entropy theory. Test the hypothesis that the effect of trace components on liquid viscosity is significant because of a greater contribution from the configurational entropy of mixing, compared with that observed in simple systems of only a few components.
(iii) Integrate the measured viscosities and heat capacities into the FLOWGO thermo-rheological model of cooling lava flows, developed by Harris and colleagues. Determine the influence of initial temperature and water content on the subsequent behavior and evolution of silicic flows.

Log viscosity as a function of H2O content for phonolites (Whittington et al., 2001;Chem Geol)
We aim to fill a critical gap in our knowledge of viscosity for dacites and rhyodacites.

This project will several broader impacts beyond the research aims stated above.
(a) Volcanic hazard mitigation. Block flows have the potential to remain mobile for months to years, and to extend great distances. The greatest hazards posed by them are block and ash flows derived from flow front collapse, which have occurred both at Santiaguito and Unzen (Japan) in recent times. The work will be carried out in close collaboration with volcano observers in Guatemala (INSIVUMEH). Our aim is to provide them with an improved predictive model of block flow emplacement, contributing to improved hazard monitoring and prediction capabilities. The computer model will be written in a generic PC format, and made freely available to others.
(b) Graduate student training. The project will support graduate students based at MU, who will be trained in viscometry and calorimetry techniques, and will conduct thermodynamic modeling of hydrous liquid viscosity. Students will be involved in modifying and testing the FLOWGO thermo-rheological model for block lava flows, in collaboration with Harris and colleagues.
(c) Enhancing research infrastructure for magma viscosity measurement in the USA. The MU experimental petrology laboratory has both concentric cylinder (high-T, low viscosity), and parallel-plate (low-T, high viscosity) viscometers. This combination of techniques allows viscosity to be determined over a temperature range exceeding that encountered in geological settings. This is important because it allows interpolation of viscosity at magmatic temperatures, rather than extrapolation of this highly non-Arrhenian property.

This will be a major step toward the long-term goal of developing a predictive thermodynamic model for the physical properties of hydrous silicate liquids.

(ii) Measurement of Transport Properties of Silicate Melts with applications to Crustal Anatexis (NSF-EAR, 01/05 - 12/07)

This project is in collaboration with co-PI Anne Hofmeister (Washington University, St. Louis).

Ph.D. student: Urmidola Raye (in prog)

Heat and mass transfer within the lithosphere are governed by the transport properties, thermal conductivity and viscosity. Available data on these physical properties are inadequate to model evolving lithospheric thermal structure and magmatic processes during orogeny with confidence. Models of lithospheric thermal structure typically assume constant thermal conductivity, although this property is known to depend strongly on temperature and rock composition. The aims of this proposal are two-fold. First, we will provide new data on the viscosity and thermal conductivity of natural and synthetic silicate melts and chemically equivalent glasses, minerals, and rocks. We will focus on high-temperature measurements that are currently scarce. Second, we will apply our results to numerical modeling of crustal anatexis in orogenic belts.
The new thermal conductivity data will significantly advance previous measurements because of: (i) much greater accuracy and precision of the laser flash technique that will be employed compared to contact methods, and (ii) far greater temperature range (limited only by sample melting). Furthermore, few studies have measured the viscosity of naturally occurring (compositionally complex) hydrous granitic or rhyolitic liquids, particularly over the large viscosity range required for reliable interpolation of viscosity under magmatic conditions.
Based on preliminary data, we have developed several hypotheses to be tested, including:
(i) Granitic liquids will have lower thermal conductivities than their protoliths at temperatures pertinent to crustal anatexis (~650-800 ¾C).
(ii) Silicic liquids and glasses will show marked variations in thermal conductivity and viscosity with normative quartz component.
(iii) Water will decrease the viscosity of simple liquids more strongly than that of complex natural melts. Existing viscosity models for silicic liquids will be tested.

Temperature of the 10^11 Pa.s isokom (a proxy for viscosity) as a function of H2O content (Whittington et al., 2005;Trans Roy Soc Edin)

We will apply to the new data to leucogranite generation by crustal anatexis, a common signature of continental collisional orogenic belts. Thermal orogenic models have struggled to explain the attainment of the high temperatures required for anatexis in the upper parts of thickened crust. Preliminary data suggest that decreasing thermal conductivity during melting may have a positive feedback relationship with increasing melt fraction during crustal anatexis. More realistic thermal models for the crust need to incorporate both temperature-dependent thermal conductivity, and the consequences of melt extraction (whose timescale is dependent on viscosity). We will conduct numerical simulations to address the following questions:
(i) How is the thermal structure of the crust affected by a T-dependent conductivity structure, and what are the relative effects of lattice and radiative components of conductivity?
(ii) How does development of a low-conductivity partial melt layer affect thermal structure, and what is the influence of the thickness of the partially molten zone?
(iii) What is the effect of melt fraction and melt extraction rate on heat transfer in the crust?
Broader impacts arising from this project will include the training of graduate students in thermal conductivity measurements by laser-flash analysis, parallel-plate and concentric-cylinder viscometry, and electron microbeam analysis. This study will impact our understanding of large-scale lithospheric processes, including heat-flow and magma transport, and volcanic hazards, in which the transport properties of magmas play a key role.

Cinder cones at dawn on the flank of Mauna Kea

The emplacement of basaltic (mantle-derived) magma into the Earth’s crust is a primary geological process, manifested in volcanic activity at the Earth’s surface, and in the emplacement of dikes and sills into shallow sedimentary rocks. This process transfers both heat and material from the Earth’s mantle to the crust, increasing the volume of the Earth’s continents over time, and substantially affecting crustal rocks which are intruded by basaltic magma (basalt is typically emplaced or erupted at temperatures in excess of 1100¾C). Understanding how magma is formed and emplaced into the Earth’s crust is essential to the interpretation of volcanic regions, whether active or extinct. By sampling along shallowly intruded dikes and sills, one can trace the magma’s chemical evolution, through time and space. When intrusive igneous rock bodies are analyzed over a large area, changes in grain size (relating to cooling rate), and in chemical composition (relating to crystallization and the assimilation of surrounding host rocks), can frequently be observed. These changes can occur along the length of a single dike, and as regional changes throughout a system of dikes from the same eruptive period. Changes in magma chemistry influence the eruptive nature of the magma, e.g. silica-rich magmas such as rhyolites are more viscous and hence much more likely to erupt violently than silica-poor magmas such as basalts. Changes in magma volatile content (water, CO2, etc.) are also very important in controlling the explosiveness of eruptions. This is of particular interest for shallowly emplaced magma, since most magmas only become saturated with water at depths of less than about 3 km in the Earth’s crust.

Analysis of the orientations of planar intrusions such as dikes and sills, which are large magma-filled cracks, can provide valuable information on the orientation of the overall regional stresses at the time of emplacement, and on the stresses imposed on the region by the intrusion of magma. Before these relationships can fully be addressed, the source of the magma must first be determined. In the case of a stock-like intrusion (a blob, frequently sketched as a vertical cylinder but many geometries are known), the source of the magma is assumed to be below the stock, the buoyant magma having flowed or percolated upwards from below. In the case of dike intrusions, flow of magma may occur laterally and is not necessarily directly up from a source below it. The source of magma in a genetically linked set of dikes may be inferred from the directions of magma flow measured at different localities. In the case of a horizontal sheet intrusion (sill), magma flow must be horizontal, usually away from a feeder dike that links in to the base of the sill.

An excellent example of magma intrusion as stocks, dikes and sills occurs in the vicinity of Spanish Peaks, Colorado. The peaks themselves are stock-like intrusions, with a radial dike swarm radiating from the central stocks. The radial dikes have been well studied, but in the same area there is also a set of sub-parallel dikes trending ENE, that have a different composition and connect to a series of sills. Based on cross-cutting relations and limited radiometric dating, the sub-parallel dike set appears to be older than the radial system.

The purpose of this study is (1) to determine whether the orientation of the sub-parallel dike set was controlled by regional or local stresses; (2) to determine flow directions in the sub-parallel dikes and related sills to locate their source; (3) to examine changes in magma chemistry and volatile content as a function of distance traveled from the source, and (4) to study interactions between magmas and the sedimentary rocks which they intrude.

Paula Lee, M.S. thesis (2005). "Spatial, temporal, and petrogenetic relationships of basaltic and lamprophyric dikes and sills of the Raton Basin, southern Colorado and northern New Mexico"

(iiib) Crustal contamination of basaltic magma: a field and experimental study (Univ. Missouri Research Board, 09/03 - 08/05)

Measured dissolution rates, combined with microscopic textural observations and chemical analyses of both experimental products and samples from the field, will be used to calculate minimum rates of magma emplacement in the southwest US and the Trans-Antarctic Mountains (see Ferrar LIP project, below). The results will be compared with rates derived from previous theoretical studies, which estimate that basaltic magma can be injected through the crust at a few meters per second. M.S. student: Jennifer Cooper (defense 11/29/05) Undergraduate student: Fred Davis (senior thesis, Fall 2005)   Jen Cooper contemplating a lateral step in a vertical dike: Raton Basin, Colorado (2003)

This project is in collaboration with PI's Tom Fleming (Southern Connecticut State), Steve Marshak (Illinois) and Anne Grunow (Ohio State)

tholeiitic dolerite sills (black) intruding Beacon Sandstone; Beacon Heights, South Victoria Land, Antarctica (2003)

We are studying the emplacement of the Ferrar dolerite sills in south Victoria Land, Antarctica. Field work over Christmas 2002-2003 indicates varying flow directions within single dikes, indicated by the morphology of sandstone contact surfaces. Petrographic and AMS analysis are ongoing, and we hope to construct a history of flow direction from initial fracture propagation, through first magma emplacement (preserve as a chilled margin), to the main flow stage of these large (up to >200 m thick) sills. This will allow construction of a regional scale flow model, which it is hoped can explain how the Ferrar dolerites were apparently emplaced very rapidly (all have indistinguishable ages and there is no evidence of significant crustal contamination) over >1600 km, apparently as sills (dikes are relatively rare and much smaller than the sills).

camping in Windy Gully, South Victoria Land, Antarctica (2003)

An extension of this work is to measure the rates of quartz dissolution on basaltic liquids as a function of temperature, for which we will use real Ferrar dolerite and real Beacon sandstone to enable petrographic comparison of textures produced in the lab and in nature, as well as dissolution rate data that may enable the timescale of emplacement to constrained by the degree of quartz contamination found in the basalt (also see previous project)

Undergraduate student: Fred Davis (senior thesis, Fall 2005)

Previous research

During my Ph.D. research in the western Himalayas I became very interested in the complex relationships between different orogenic processes, especially those which contribute to the exhumation of metamorphic belts. Although classically thought of as the retrograde part of the orogenic P-T-t path, the exhumation stage can produce several melting episodes, depending crucially on the spatial and temporal patterns of heat and fluid fluxes. Dehydration melting reactions, which have positive slopes in P-T space, are particularly favored by rapid exhumation.

The Nanga Parbat Massif is perhaps the best place in the world to study metamorphism and crustal melting during rapid exhumation. The massif is a bivergent wedge, constantly accreting new material to its base and shedding old material through glacial erosion and landslides. This activity results in rapid exhumation (about 3 to 4 mm/y, see next section) and the associated advection of heat towards the surface drives a vigorous hydrothermal system. I found evidence of three separate melting reactions occurring within the last few million years, including (i) dehydration melting of muscovite (to produce leucogranites), (ii) biotite (leaving spinel-cordierite restites,and (iii) water-saturated melting in shear zones where fluid flux is concentrated. A P-T-t path for rocks currently exposed at the surface was constructed based on the different melting reactions, which matched the predicted P-T-t path from thermal modeling. The combination of forward and inverse approaches (thermal model on the one hand; geochemistry and P-T work on the other) was very successful here.

In the western Himalaya, dehydration melting was achieved in the absence of extensional structures, with rapid unroofing accommodated by exceptionally vigorous erosion which has led to extreme topographic relief (about 7 km vertical relief in about 20 km horizontal distance). I modeled the exhumation history of this region, with geochronological constraints from published data and my own work. The two major findings were that (i) rapid exhumation leads to elevated near-surface geotherms, so that exhumation rates may be overestimated where cooling rates are rapid, and (ii) accelerating cooling rates do not necessarily imply accelerating exhumation rates. My study placed important quantitative constraints on a well-studied extreme example of rapid exhumation and denudation.

Models of collisional orogenesis indicate that accretion of heat producing elements is required for melting to occur; redistribution of these elements via melting may make melting in subsequent cycles substantially less likely, allowing the signatures of polymetamorphism to remain relatively intact. For example, using isotopic model ages we have identified the protolith of the western Himalaya, and a Proterozoic metamorphic event, demonstrating a complex polymetamorphic history for this region and shedding new light on the deep structure of the orogen.

(vi) Metamorphism and melting in a "confined" Neoproterozoic orogen: The Araçuaí belt, Brazil

This project is in collaboration with PI Steve Marshak (Illinois), Jim Connelly (Texas), Fernando Alkmim (UFOP, Ouro Preto) and Antônio-Carlos Pedrosa-Soares (UFMG, Belo Horizonte)

The Neoproterozoic Araçuaí belt of eastern Brazil records the closure of a Red Sea-type basin between the São Francisco and Congo cratons of West Gondwana, and is part of the Brasiliano (Pan-African) orogenic system. In its southern part this basin contained oceanic crust, so that closure involved subduction and arc magmatism prior to continental collision. In its northern part, the basin was entirely ensialic. After the cessation of subduction, a period of magmatic quiescence and possible plateau development ended with the collapse of the orogen. Extensional structures are observed across the preserved Brazilian portion of the orogen (the eastern segment of the orogen is now in West Africa). Six distinct syn- to post- kinematic granite suites were emplaced within the orogen, although the temporal relation of the different granite suites to contractional and post-collision extensional tectonism has remained enigmatic.

mafic enclaves in a "G5" post-orogenic granite, Araçuaí belt, Brazil (2000)

Preliminary U-Pb ages for key granite suites are related to the structural history of the orogen, placing important constraints on the timing and duration of orogenic collapse. The age of the northernmost pluton of the subduction-related (G1) granite suite is consistent with published ages for this suite of 625 to 575 Ma. The G2 suite consists of an "anatectic sea" of variably foliated S-type migmatitic leucosomes, leucogranitic sheets and plutons, commonly associated with locally extensional mesoscopic shear zones. This migmatite-leucogranite association covers >75% of the outcrop over an area of ~ 20,000 km2. A preliminary U-Pb zircon age of ca. 536 Ma from a well-foliated G2 pluton is significantly younger than published ages from other G2 bodies (590 to 570 Ma). Bodies sampled from the other suites are: an I-type granite showing flow foliation but little or no tectonic fabric (G3I suite), an undeformed S-type leucogranite (G4 suite), and an undeformed I-type granite containing dioritic enclaves (G5 suite). Preliminary U-Pb zircon ages from all three are within error and slightly younger than the age of G2, consistent with published ages from other G4 and G5 bodies.

The new younger age of at least some G2 granites alleviates a major problem in the tectonic evolution of the Araçuaí belt, which was the association of these granites with extensional shear zones apparently at ca. 590 to 570 Ma (a time when subduction-related magmatism, and presumably crustal thickening, were still active). If widespread generation of S-type migmatites, sheets and plutons in fact occurred at ca. 530 to 540 Ma, part of this suite can instead be related to orogenic collapse (the older S-type bodies may be related to syncollisional magmatism). Since the body we sampled is well foliated, the intrusive age gives an upper limit for the onset of extensional fabric development in this location. The presence nearby of unfoliated G4 and G5 plutons intruded only a few m.y. later restricts collapse to a relatively short period.

AGU 2001 poster (gif, 500 kb)

(vii) Other things (what I did on my holidays)

After some nights at the telscopes on Mauna Kea, we took a helicopter ride over Pu'u O'o (2002)