A machine learning approach for regional geochemical data:Platinum-group element geochemistry vs geodynamic settings of the North Atlantic Igneous Province

Whilst traditional approaches to geochemistry provide valuable insights into magmatic processes such as melting and element fractionation, by considering entire regional data sets on an objective basis using machine learning algorithms(MLAs), we can highlight new facets within the broader data structure and significantly enhance previous geochemical interpretations.The platinum-group element(PGE) budget of lavas in the North Atlantic Igneous Province(NAIP) has been shown to vary systematically according to age, geographic location and geodynamic environment.Given the large multi-element geochemical data set available for the region, MLAs were employed to explore the magmatic controls on these shifting concentrations.The key advantage of using machine learning in analysis is its ability to cluster samples across multi-dimensional(i.e., multi-element)space.The NAIP data set is manipulated using Principal Component Analysis(PCA) and t-Distributed Stochastic Neighbour Embedding(t-SNE) techniques to increase separability in the data alongside clustering using the k-means MLA.The new multi-element classification is compared to the original geographic classification to assess the performance of both approaches.The workflow provides a means for creating an objective high-dimensional investigation on a geochemical data set and particularly enhances the identification of metallogenic anomalies across the region.The techniques used highlight three distinct multi-element end-members which successfully capture the variability of the majority of elements included as input variables.These end-members are seen to fluctuate in prominence throughout the NAIP, which we propose reflects the changing geodynamic environment and melting source.Crucially, the variability of Pt and Pd are not reflected in MLA-based clustering trends, suggesting that they vary independently through controls not readily demonstrated by the NAIP major or trace element data structure(i.e., other proxies for magmatic differentiation).This data science approach thus highlights that PGE(here signalled by Pt/Pd ratio) may be used to identify otherwise localised or cryptic geochemical inputs from the subcontinental lithospheric mantle(SCLM) during the ascent of plume-derived magma, and thereby impact upon the resulting metallogenic basket.     

A 520 year record of summer sunshine for the eastern European Alps based on stable carbon isotopes in larch tree rings

A 520-year stable carbon isotope chronology from tree ring cellulose in high altitude larch trees (Larix decidua Mill.), from the eastern European Alps, correlates more strongly with summer temperature than with summer sunshine hours. However, when instrumental records of temperature and sunshine diverge after AD1980, the tree ring time series does not follow warming summer temperatures but more closely tracks summer sunshine trends. When the tree ring stable carbon isotope record is used to reconstruct summer temperature the reconstruction is not robust. Reconstructed temperatures prior to the twentieth century are higher than regional instrumental records, and the evolution of temperature conflicts with other regional temperature reconstructions. It is concluded that sunshine is the dominant control on carbon isotope fractionation in these trees, via the influence of photosynthetic rate on the internal partial pressure of CO₂, and that high summer (July–August) sunshine hours is a suitable target for climate reconstruction. We thus present the first reconstruction of summer sunshine for the eastern Alps and compare it with the regional temperature evolution.     

Applying neighbourhood classification systems to natural hazards: a case study of Mt Vesuvius

Mt Vesuvius is regarded as one of the most deadly volcanoes on earth. With over 1 million people living on its flanks and in its periphery, there is little doubt that an eruption of sub-Plinian magnitude would be catastrophic to the livelihood and well being of contemporary Neopolitans. Such a large scale eruption would have wide ranging and differential effects on the surrounding population. Whereas previous studies of social vulnerability have focused on individual demographic factors (such as age, income or ethnicity), this research proposes the application of a general neighbourhood classification system to assess natural hazard vulnerability. In this study, Experian’s Mosaic Italy is used to classify and delineate the most vulnerable neighbourhood types around the province of Naples. Among the neighbourhoods considered most at risk, those areas with high proportions of elderly and low income families are deemed particularly vulnerable. With current evacuation plans deemed outdated and poorly communicated to the locals Rolandi (2010), Barberi et al. (2008), this methodology could prove to be a useful input to both town planners and civil protection agencies. A range of statistical measures and geophysical risk boundaries are employed here to assess the different areas of human resilience.     

Alteration of platinum-group minerals and dispersion of platinum-group elements during progressive weathering of the Aguablanca Ni–Cu deposit, SW Spain

The distribution, mineralogy and mobility of the platinum-group elements (PGE) in the surface environment are poorly understood. This study of the lower, less altered and upper, more altered gossan, overlying the Aguablanca Ni–Cu-(PGE) magmatic deposit (Spain), has shown that the platinum-group minerals (PGM) are progressively oxidised and dispersed into iron oxides that form the gossan. A combination of the characterization of PGE in host PGM, using a scanning electron microscope, and measurement of PGE at lower concentrations in host iron oxides, using laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS), has for the first time allowed the total distribution of the PGE within a gossan to be documented. This study has revealed a complete in situ alteration and dispersion sequence of the PGM including (1) breakdown of both the more stable Pt-arsenides, Pt/Pd-tellurides and the less stable bismuthotellurides, (2) formation of partially oxidised PGM, (3) development of a wide range of oxidised Pt- and Pd-bearing phases, (4) subsequent formation of Fe–PGE-oxides and PGE-hydroxides, (5) incorporation of PGE into ferruginous supergene products and lastly (6) concentration of PGE at the edges of veins and iron oxides. Dispersion of Pd is greater than for the other PGE with Pd being widely distributed throughout the iron oxides. This oxidising environment produced PGE-oxides rather than PGE-alloys, also commonly found in the surface environment, especially in placers. These results provide critical evidence for the stages of mineralogical change from PGE host mineralogy in magmatic ores to surface weathering producing PGE-oxides.     

Impact melt rocks from the Late Paleocene Hiawatha impact structure,northwest Greenland

Impact melt rocks formed during hypervelocity impact events are ideal for studying impact structures. Here, we describe impact melt rock samples collected proximal to the 31 km wide 58 Ma Hiawatha impact structure, northwest Greenland, which is completely covered by the Greenland Ice Sheet. The melt rocks contain diagnostic shock indicators (e.g., planar deformation features [PDF] in quartz and shocked zircon) and form three groups based on melt textures and chemistry: (i) hypocrystalline, (ii) glassy, and (iii) carbonate-based melt rocks. The exposed foreland directly in front of the structure consists of metasedimentary successions and igneous plutons; however, the carbonate-based impactites indicate a mixed target sequence with a significant carbonate-rich component. Well-preserved organic material in some melt rocks indicates that North Greenland at the time of impact was host to abundant organic material, likely a dense high-latitude temperate forest. Geochemical signatures of platinum-group elements in selected samples indicate an extraterrestrial component and support previous identification of a highly fractionated iron impactor in glaciofluvial sand. Our results illustrate the possibility to study impact structures hidden beneath a thick ice sheet based on transported samples and this opens a new avenue for identifying other potential impact craters in Greenland and Antarctica.