RICERCA ITALIANA     

Research program

CHAOTIC DYNAMICS AND FRACTAL GEOMETRY IN THE GENESIS AND MIXING OF MAGMAS

University Co-ordinator

Università degli Studi di PERUGIA - SCIENZE DELLA TERRA - PERUGIA(PG)

Research Unit Leader

Giampiero POLI

Description

The aim of this research project is to study in details different aspects of magma mixing processes including interaction of basaltic melts during their migration towards the Earth surface, the initial stages of intrusion of a mafic magma into a felsic magma chamber, and more advanced interaction stages where magmas interact in the magma chamber and in crustal conduits.
By starting from acquired knowledge about magma mixing processes and experience in development of new methods based on Fractal Geometry and Chaos Theory [e.g. 4,5,9,10,11,12,13,14] the main targets of this project are:
i) to understand mechanisms acting during interaction of magmas with similar rheological properties by studying basaltic magmas, generated from an heterogeneous mantle, which interact during their ascent to the Earth surface and to develop new analytical method based on Fractal Geometry and Chaos Theory to reconstruct the geochemical composition of end-members;
ii) to investigate the dynamics characterising the initial stages of intrusion of mafic magmas into felsic magma chambers in relationship to rheological contrasts between magmas;
iii) to understand in detail the role played by the different dynamic regions coexisting in the same system (AMR and CR; see 2.4) in the development of chemical diffusion and to establish the influence of mixing processes at the micrometric scale ("micro-mixing") on the chemical composition of melt inclusions;
iv) to establish the influence of mineralogical phases on the spatial and temporal evolution of magma mixing processes and their role in restructuring flow fields (AMR and CR) within the mixing system;
v) to understand the factors leading magma mixing systems to generate chaotic dynamics, and to develop new analytical methods to exploit such dynamics with the aim of reconstructing petrogenesis of rocks generated by magma mixing processes.
The research will be carried out by detailed studies of plutonic and volcanic rocks bearing evidence of magma mixing processes using conventional analytical techniques, experimental petrology in collaboration with Prof. Dingwell group (Ludwig-Maximilians-Universität, Munich, Germany), numerical simulations by finite element technique, and new methods based on Fractal Geometry and Chaos Theory.
In particular, regarding the plutonic environment, researches will be focused on key outcrops of granitoid rocks constituting the Adamello pluton (Italian Alps) and plutons of coastal Maine [USA; e.g. Vinalhaven Pluton; 19,20] where the initial stages of intrusion of a mafic magma in a felsic magma chamber are "fossilized" and, hence, well analyzable. Regarding the volcanic environment attention will be focused on mingled/mixed lavas with variable crystal contents cropping out at Monte Arci [Sardinia, Italy; 21,22,23,24] and on the islands of Capraia [Italy; 40,41,42,12] and Lesbos [Greece; 4,5,25,26] where mixing structures are well preserved.
The project is divided into two main Activities (Activity 1 and 2), which are described in detail as follows.

Activity 1
Activity 1 is focused on studies of mixing processes between magmas with similar geochemical properties and on the initial stages of intrusion of a mafic magma into a felsic magma chamber.
Regarding the study of magma mixing between rheologically similar magmas, attention will be focused on balsatic magmas since mixing processes can explain the strong geochemical and isotopic variability which is commonly observed even within the same volcanic centre [e.g. 28,39]. In these cases there is not macroscopic evidence of interaction processes because magmas have similar characteristics and, therefore, indirect methods have to be used in order to identify if such rocks suffered any mixing process. Chemical transects across basaltic samples from Mt. Etna, in collaboration with the Unit of Pisa, will be analyzed by EMPA and LAM-ICP-MS and the degree of correlation between trace elements will be studied; this method has been demonstrated to be very efficient to recognise the presence of chaotic mixing in rock samples [see 2.4; e.g. 9]. These analyses will allow us to study the style and efficiency of mixing processes between magmas with similar rheologies and will give the opportunity to establish if basaltic samples are to be considered "primary" magmas or if they recorded mixing processes, and to identify the best criteria to reconstruct the composition of end-member magmas.
Regarding the initial stages of intrusion of mafic magmas into felsic magma chambers, recent studies evidenced the fractal nature of contacts between the two magmas ("viscous fingering") whose degree of irregularity would depend on the rheological contrast between magmas [e.g. 15,16; Fig. 2.5-1]. Since Fractal Geometry is the "geometry of chaos" it is reasonable to hypothesise that the fingering process is chaotic and, hence, it obeys the "sensitivity to initial conditions" (APPENDIX). This could imply that small differences in the initial rheologies of magmas could lead to mixing processes with strongly different efficiencies. Therefore, understanding the dynamics which characterise these initial stages of magma interaction is crucial in order to evaluate the possibility to generate large amounts of hybrid magmas, a hotly debated topic in modern igneous petrology [e.g. 29,30].
The study of initial stages of the intrusion of a mafic magma into a felsic magma chamber requires the availability of outcrops where these initial stages have remained "fossilised", and that only the plutonic environment can furnish. In the Adamello pluton [Italian Alps; 31,32] and some plutons of coastal Maine [USA; e.g. Vinalhaven Pluton; 19,20] magma mixing processes between mafic and felsic magmas are common [e.g. 19,20,31,32] and, in some cases, there is availability of geological sections showing the initial stages of intrusion of a mafic magma into a felsic magma "frozen" in time. Samples for performing fractal analysis of mafic/felsic interfaces are already available regarding Maine plutons and other samples will be soon available thank to the collaboration with American colleagues. All samples will be analysed in detail for their major and trace element composition on both whole rock (XRF e ICP-MS) and mineral phases (LAM-ICP-MS) in order to correlate the complexity of mafic/felsic interfaces with mass transfer processes (both chemical diffusion and transfer of mineral phases from one magma to the other).
The study of rocks does not give the opportunity to follow the dynamic evolution of the injection process in time and, since in chaotic system the time dynamic is of fundamental importance for the evolution of the entire system, it is necessary to perform experimental petrology runs to follow the process in time. In collaboration with Prof. Dingwell research group (Ludwig-Maximilians-Universität, Munich, Germany) experiments of injection of a mafic magma into a felsic magma will be performed. The injection of the mafic magma will be performed by using a centrifuge, a method which has been demonstrated extremely efficient to reproduce in short times several igneous processes [e.g. 33,34,35; Fig. 2.5-2]. The sketch of Fig. 2.5-3 shows the main features of experiments that will be carried out. In particular, it is planned to introduce a mafic and a felsic magma in a capsule separated by a platinum piston with a central hole; rotation of the capsule by the centrifuge will provoke the sliding of the piston and the intrusion of the mafic magma into the felsic one; several experiments will be performed by considering different run times in order to follow the time evolution of the process; experiments will be performed by considering magma with variable geochemical compositions and, hence, variable rheologies. Experimental samples will be studied by Fractal Geometry to quantify the irregularity of mafic/felsic interfaces, and analysed by EMPA and LAM-ICP-MS in order to correlate the morphology of the interface with chemical exchanges (both major and trace elements) between magmas.
The comparison between results obtained on experimental and natural samples will allow us to characterise rigorously the dynamics associated to the initial stages of intrusion of a mafic magma into a felsic magma chamber, and to clarify its petrological significance.

Activity 2
Activity 2 will be focused on a detailed study of magma mixing processes by considering end-members with variable geochemical features, and hence variable rheologies, and different crystal contents.
Part of this Activity will be concentrated on rocks bearing evidence of magma interaction processes between mafic and felsic magmas with the aim of studying chemical exchanges during mixing processes at the micrometric scale ("micro-mixing"). This will give the opportunity to understand the influence of the different dynamic regions (AMR and CR) at such a length-scale and, hence, to evaluate the potential pitfalls in the usability of melt inclusions in petrological system characterised by mixing processes [e.g. 9]. Analyses will be carried out on lavas cropping out at Monte Arci (Sardinia, Italy) and on the islands of Capraia (Italy) and Lesbos (Greece) where magmas having variable geochemical compositions and rheologies interacted. In particular, compositional transects on samples with variable crystal contents will be analysed along with melt inclusions in mineral phases. The comparison between the degree of correlation of chemical elements across the transects and in the melt inclusions will allow us to understand the representativeness of melt inclusions in magma mixing systems and to set up diagnostic methods to establish their applicability in petrogenetic studies.
In order to study in detail this crucial aspect of magma mixing, experiments will be performed in collaboration with Prof. Dingwell research group by mixing natural magmas from both Monte Arci and Lesbos island. On this respect, recent studies of Munich research group demonstrated that it is possible to mix natural magmas by keeping into account a large number of parameters and it is, hence, possible to follow in great detail the spatial and temporal evolution of the process [e.g. 17; Fig. 2.5-4]. In the present research project experiments will be carried out by mixing magmas following chaotic mixing experimental protocols. In particular, the basic configuration that will be utilised in the so-called "Journal Bearing Flow" (JBF) which has been demonstrated extremely useful in the study of chaotic fluid mixing [e.g. 36,37,38]. This system is constituted of two eccentric cylinders and mixing is induced by rotating the cylinders in a time-periodic fashion [Fig. 2.5-4A]. This system is solvable numerically and chaotic configurations can be reproduced with great detail [e.g. 36,37,38; Fig. 2.5-5]; these configurations will be utilised for magma mixing experiments. It is noteworthy that the system possesses all the fluid-dynamic "ingredients" of natural chaotic mixing systems since are present both AMR and CR [e.g. 18,36] which are identical, from a dynamic point of view, to those observed in natural samples (compare Fig. 2.4-1 and Fig. Fig. 2.5-5).
Experimental samples will be analysed by using the same techniques used to study natural samples and comparison among results will allow us to understand in detail the influence of AMR and CR on the development of "micro-mixing" processes.
The study of samples of mixed volcanic rocks with high crystal contents will allow us to understand the role played by mineralogical phases in the development of the different dynamical regions (AMR and CR) and their influence on mixing dynamics. By using experimental mixing protocols developed in the previous research stage, experimental petrology runs will be performed by mixing magmas with variable crystal contents in order to follow in space and time the influence of mineral phases on mixing dynamics.
In both above Activities numerical simulations will be carried out by using the finite element technique with the aim to understand the causes which are responsible for the onset of chaotic dynamics in magma mixing systems. In particular, numerical simulations will be performed keeping into account variable boundary conditions valid for magmatic systems by considering convection in magma chambers with different aspect ratio and in conduits, flow in crustal conduits, and in fracture networks.
Researches carried out during this project will allow to comprehend in detail the petrogenetic importance of chaotic mixing processes in magmatic systems and to understand their causes. It is to note that for the first time it is attempted to integrate methods developed in the fields of Chaos Theory and Fractal Geometry, and experimental petrology techniques. Such an approach could be extremely fruitful in order to clarify the several problems that still remain to be explained in the field of magma mixing, a process that, generally speaking, has to be regarded as ubiquitous during genesis, transport, and evolution of magmatic systems.

The above illustrated researches will be carried out through the following stages:
First Year.
Stage #1 - months 1-3: sampling of basaltic rocks on Mt. Etna, in collaboration with the Unit of Pisa. Sampling of granitoid rocks bearing evidence of the initial stages of the intrusion of a mafic magma into a felsic magma chamber occurring at the Adamello pluton and acquisition of data from similar outcrops (coastal Maine; USA) from American colleagues.
Stage #2 - months 4-8: image and geochemical analyses. Experimental petrology runs of injection of a mafic magma into a felsic magma by using the centrifuge apparatus.
Stage #3 - months 9-12: analysis of data obtained from the application of Chaos Theory and Fractal Geometry and reconstruction of a model for the initial stages of magma interaction. Starting of numerical simulations of magma mixing processes in magma chamber.
Second Year.
Stage #4 - months 13-16: on the basis of knowledge acquired during the first year, the research will continue by sampling mixed lavas with variable crystal contents occurring at Monte Arci (Sardinia, Italy) and on the islands of Capraia (Italy) and Lesbos (Greece).
Stage #5 - months 17-20: geochemical analysis on rocks and melt inclusions. Experimental petrology runs of magma mixing processes by using the JBF protocol considering magmas with variable crystal contents.
Stage #6 - months 21-24: analysis of data obtained from the application of Chaos Theory and Fractal Geometry techniques and reconstruction of a complete model of magma mixing processes at the micrometric scale, their influence on melt inclusion composition, and of the role played by mineral phases on mixing dynamics. Completion of numerical simulations of magma mixing processes in crustal conduits and in fracture networks.
Data diffusion
Publication and presentation of results on national and international journals, at least 4 papers on international journals, and at national and international meetings (e.g. FIST, EGS-AGU, EGU), also in collaboration with the Unit of Pisa.