A remote sensing and geospatial statistical approach to understanding distribution and evolution of ignimbrites in the Central Andes with a focus on Southern Peru
by Melanie Brandmeier
Date of Examination:2014-02-24
Date of issue:2014-03-25
Advisor:Prof. Dr. Gerhard Wörner
Referee:Stefan Erasmi
Referee:Prof. Dr. Martin Kappas
Referee:Prof. Dr. Reinhard Gaupp
Files in this item
Name:Mastertable.xlsx
Size:1.31Mb
Format:VND.OPENXMLFORMATS-OFFICEDOCUMENT.SPREADSHEETML.SHEET
Description:Summary table of samples
Name:Dissertation.pdf
Size:55.8Mb
Format:PDF
Description:Dissertation Melanie Brandmeier 2014
Abstract
English
This PhD thesis investigates timing, cause and location of “ignimbrite flare ups” during the evolution of the Central Andean uplift and relates volcano-tectonic structures and calderas with shallow intrusive stocks to mineralization by applying methods of remote sensing, GIS and geospatial statistics together with traditional geological fieldwork, 40Ar/39Ar geochronology and geochemical analysis. Chapters II to IV focus on Southern Peru, while Chapter V puts local results on an Andean scale, investigating compositional differences in ignimbrites and modelling ingimbrite eruptions in space and time.CHAPTER II: In this chapter we present first results from a reconnaissance study using ASTER data in Southern Peru. A combined approach to detect hydrothermal alteration zones and their mineral distribution is proposed for a relatively remote area around the Carhuarazo volcanic complex in Southern Peru encompassing 2,222 km2. In this region, tertiary volcanic structures associated with hydrothermal alteration are well known to host epithermal ore deposits. We make an attempt to detect and to quantify alteration minerals based on spectral analysis using ASTER reflectance data product provided by LP-DAAC. Besides commonly used ratio images, mineral indices (MI) and relative band depth images (RBD), we also extracted end-member spectra using Pixel-Purity-Processing, preceded by minimum noise fraction transformation. These spectra are thought to represent the spectrally purest pixel of the image and show the typical absorption features of the main constituents. Based on this assumption, we used different spectral analysis methods in order to extract the most important alteration minerals for such an environment. These minerals were then used for matched filter processing in areas showing high values in MIs and RBDs. Using this method, we detected and mapped argillic alteration and variations in the distribution of important minerals like alunite, kaolinite or nacrite. There were no indications for the presence of propilitization at ASTER spatial resolutions. Our method can be applied easily to any ASTER scene and provides information about the intensity of alteration and the character of alteration zones. The intensity is highest in the center of the Carhuarazo volcanic complex and is mostly argillic with a high content of alunite, dickite and other clay minerals.
CHAPTER III: This chapter further develops results presented in Chapter II, focusing on mineral and lithological mapping in an extended area in Southern Peru to better characterize and understand the Tertiary volcanic evolution in this region. Our goal was to characterize volcanic regions near Puquio (Ayacucho) by correlating areas of intense alteration and related ignimbrite outflow sheets. In particular, we spectrally and mineralogically mapped different types and intensities of alteration based on remote sensing and ground-truth data. ASTER ratio images, alteration indices and false color composites were used to select ground-training areas for sample collection and field spectrometry. Alteration samples were characterized geochemically, mineralogical and spectrally. Absorption features correlate with chemical properties and an Index of Absorption was proposed as a measure for alteration intensity. Hyperspectral data from field spectrometry allows identification of important alteration minerals such as kaolinite and smectite. Alteration mineral assemblages range from silicic to argillic to “zeolite-type”. Using a support vector machine classification (SVM) algorithm on ASTER data, we mapped the different types and intensities of alteration, along with unaltered ignimbrite and lava flows with an accuracy of 80%. We propose a preliminary model for the interpretation of alteration settings, discuss the potential eruption sites of the ignimbrites in the region and, propose pH and temperature estimates for the respective classes based on the mineral assemblages identified.
CHAPTER IV: In this chapter, we present geological and chronological data for the ignimbrites of the area spectrally characterized in Chapters II and III in order to constrain the Neogene evolution of Southern Peru. We present 31 40Ar/39Ar ages of ignimbrites and related lava samples from three major valleys draining to the Amazon River and one valley west of the present drainage divide in the area. We combine these with drainage and DEM analysis and present a new stratigraphy for the western side of the drainage divide. We discuss timing, extent and possible eruptive centers for the ignimbrites and propose a “Santa Ana Caldera” with an age of ~5 Ma and a diameter of ~20 km. Ignimbrite ages correspond to the ~20 Ma (Nazca age) and four age ranges within the “Formación Andamarca”: 14 Ma (Andamarca 1), 7.5-9.5 Ma (Andamarca 2), 5-6.5 Ma (Andamarca 3) and 3.5-4 Ma (Andamarca 4). Based on the position of the samples in the valleys, we found a minimum incision of ~300 m prior to 14 Ma, of ~500-800 m after ~6 Ma and renewed incision after ~3.8 Ma of 200 to 300 m. For the Visca valley, we know of at least one additional event (~9.4 Ma) that filled the valley and cannot be quantified in terms of re-incision. Causes for incision and changes in incision rates are uplift (mainly between 14 Ma to 3.8 Ma) and a change in climate and drainage system with related base-level changes. Our findings agree with an increase of erosion rates and headwater erosion found by other authors in the Eastern Cordillera at 15-10 Ma that would have shifted the drainage divide in a westerly direction. Uplift on the order of 2000-3500 m found in the Altiplano and Eastern Cordillera since ~ 10 Ma are reflected by river incision of at least 1 km during that time, with at least one more phase of incision (9-6 Ma) that cannot be quantified. Re-incision after 3.8 Ma is probably related to the wetter climate and glaciation history of the area Comparing the stratigraphic record of the three valleys east of the drainage divide to the one valley west of it and stratigraphies found by other authors for the westerns escarpment, we found that the 20 Ma Nazca age (and mostly the 14 Ma age) are completely absent on the eastern part of the drainage divide. This striking difference argues for high rates of incision and denudation in that direction whereas the plateau-forming Nazca and Huaylillas ignimbrites are so well preserved on the western escarpment. Assuming that eruptions of these ignimbrites were not completely asymmetric, we argue that this is due to uplift caused by the arrival of the Nazca ridge at that latitude after ~12 Ma, accompanied by a change in precipitation toward more humid conditions on the eastern side of the orogen.
CHAPTER V: This chapter puts our regional findings on an Andean scale. We analyzed temporal and compositional patterns of large volume ignimbrite magmatism in the Central Andes during the Neogene using geostatistical modeling and tested the hypothesis whether we can use compositional signatures to “fingerprint” ignimbrites. In order to examine the spatiotemporal pattern of so-called “ignimbrite flare-ups”, we mapped 201 ignimbrites, together with 1,602 ignimbrite samples (with geochronological and chemical data) using satellite imagery, available literature, maps and new data presented in Chapter IV, creating a Web Mapping Application (WMA) that is now globally available. Applying cluster analysis to clr-transformed major and trace element data, we grouped ignimbrites according to geochemical characteristics and compared our results to traditional geochemical parameters. Based on major elements, we found a rhyolitic and a dacitic “end-member”. Those “end-members” overlap in REE compositions with more or less pronounced negative Eu anomaly, depletion of MREEs and enrichment/depletion in LREEs. Based on these results, we argue, that a mere distinction between “rhyolitic, crystal-poor, small-volume” and “large-volume, crystal-rich monotonous intermediate” seems to be insufficient to capture differences in ignimbrite evolution and genesis. The large-volume, rhyolitic, ~19 Ma Oxaya ignimbrite, for example, is compositionally distinct from the young, dacitic ~4 Ma APVC Atana ignimbrite, implying a different genesis model than suggested for the large-volume APVC ignimbrites, with accumulation of large bodies of dacitic magma in the upper crust with time, fed by mantle power input. This finding agrees with differences in Sr isotopes, arguing for less crustal assimilation, possibly due to the thinner and colder crust at that time. To test our hypothesis that compositional signatures can be used to “fingerprint” ignimbrites, we applied discriminant analysis to selected ignimbrites. Classification gave an 87.5% overall classification accuracy and we therefore propose to apply this technique more widely on compositional signatures. Spatiotemporal pattern of so-called “ignimbrite flare-ups” were simulated by calculating the cumulative areal extent and volume of ignimbrites over time. We propose minimum estimates for the whole CVZ and for five N-S segments. In total, we estimate eruptive volumes of 31,000 km3, with 2,400 km3 for Southern Peru, 2,700 km3 for Southernmost Peru, 8,400 km3 for the Altiplano, 14,200 km3 for the Northern Puna and 3,100 km3 for the Southern Puna segments. As ignimbrite eruptions represent the surface manifestation of plutonic activity, eruptive volumes can help us to understand processes taking place in the upper crust. Using the same assumptions as De Silva and Gosnold (2007), we calculate a minimum plutonic input of 7,200 km3, 8,100 km3, 25,200 km3, 42,600 km3 and 9,300 km3 for the respective segments during the past 30 Ma and observe a N-S “younging” of eruption ages and “ignimbrite pulses”. Major pulses occurred at 19-24 Ma, 13-14 Ma, 6-10, 3-6 Ma with only minor ignimbrites after 3 Ma. We propose that large-volume ignimbrite eruptions occurred in the wake of the subducting Juan-Fernandez ridge, with compression, uplift, shallow subduction and fluid release in a first stage, upon arrival of the ridge, and melting of the so “conditioned” crust due to renewed asthenospheric mantle flow above a steepening slab after the passing of the ridge. The total estimates for the northern segments 1-3 and the Northern Puna are sub-equal, however, calderas and thus intra-caldera volumes for ignimbrites in these segments are not known and due to higher age, preservation level for the ignimbrites is much lower. Thus, it may be possible, that the latter ignimbrites represent volumes greater than in the Northern Puna. If further studies show that this is the case, the concept of an APVC flare-up should be revised and not viewed as a regionally and temporally restricted event of high-magma flux and batholith construction. Instead, we suggest a paradigm shift towards a dynamic model, with the “flare up” as a moving entity that has progressed across the Andes during the past 25 Ma, probably related to ridge subduction, with the Northern Puna flare up only being the most recent and best preserved remnant, and the Southern Puna Cerro Galán eruption possibly heralding another flare-up.
Keywords: Andes; ignimbrites; remote sensing; ASTER; alteration mapping; compositional data analysis; geospatial statistics; Ar/Ar dating; Peru