Evaluation of current statistical approaches for predictive geomorphological mapping

Predictive models are increasingly used in geomorphology, but systematic evaluations of novel statistical techniques are still limited. The aim of this study was to compare the accuracy of generalized linear models (GLM), generalized additive models (GAM), classification tree analysis (CTA), neural networks (ANN) and multiple adaptive regression splines (MARS) in predictive geomorphological modelling. Five different distribution models both for non-sorted and sorted patterned ground were constructed on the basis of four terrain parameters and four soil variables. To evaluate the models, the original data set of 9997 squares of 1 ha in size was randomly divided into model training (70%, n=6998) and model evaluation sets (30%, n=2999).In general, active sorted patterned ground is clearly defined in upper fell areas with high slope angle and till soils. Active non-sorted patterned ground is more common in valleys with higher soil moisture and fine-scale concave topography. The predictive performance of each model was evaluated using the area under the receiver operating characteristic curve (AUC) and the Kappa value. The relatively high discrimination capacity of all models, AUC=0.85–0.88 and Kappa=0.49–0.56, implies that the model's predictions provide an acceptable index of sorted and non-sorted patterned ground occurrence. The best performance for model calibration data for both data sets was achieved by the CTA. However, when the predictive mapping ability was explored through the evaluation data set, the model accuracies of CTA decreased clearly compared to the other modelling techniques. For model evaluation data MARS performed marginally best.Our results show that the digital elevation model and soil data can be used to predict relatively robustly the activity of patterned ground in fine scale in a subarctic landscape. This indicates that predictive geomorphological modelling has the advantage of providing relevant and useful information on earth surface processes over extensive areas, such data being unavailable through more conventional survey methods.     

Crustal structure of the Lofoten–Vesterålen continental margin, off Norway

By compiling wide-angle seismic velocity profiles along the 400-km-long Lofoten–Vesterålen continental margin off Norway, and integrating them with an extensive seismic reflection data set and crustal-scale two-dimensional gravity modelling, we outline the crustal margin structure. The structure is illustrated by across-margin regional transects and by contour maps of depth to Moho, thickness of the crystalline crust, and thickness of the 7+ km/s lower crustal body. The data reveal a normal thickness oceanic crust seaward of anomaly 23 and an increase in thickness towards the continent–ocean boundary associated with breakup magmatism. The southern boundary of the Lofoten–Vesterålen margin, the Bivrost Fracture Zone and its landward prolongation, appears as a major across-margin magmatic and structural crustal feature that governed the evolution of the margin. In particular, a steeply dipping and relatively narrow, 10–40-km-wide, Moho-gradient zone exists within a continent–ocean transition, which decreases in width northward along the Lofoten–Vesterålen margin. To the south, the zone continues along the Vøring margin, however it is offset 70–80 km to the northwest along the Bivrost Fracture Zone/Lineament. Here, the Moho-gradient zone corresponds to a distinct, 25-km-wide, zone of rapid landward increase in crustal thickness that defines the transition between the Lofoten platform and the Vøring Basin. The continental crust on the Lofoten–Vesterålen margin reaches a thickness of 26 km and appears to have experienced only moderate extension, contrasting with the greatly extended crust in the Vøring Basin farther south. There are also distinct differences between the Lofoten and Vesterålen margin segments as revealed by changes in structural style and crustal thickness as well as in the extent of elongate potential-field anomalies. These changes may be related to transfer zones. Gravity modelling shows that the prominent belt of shelf-edge gravity anomalies results from a shallow basement structural relief, while the elongate Lofoten Islands belt requires increased lower crustal densities along the entire area of crustal thinning beneath the islands. Furthermore, gravity modelling offers a robust diagnostic tool for the existence of the lower crustal body. From modelling results and previous studies on- and off-shore mid-Norway, we postulate that the development of a core complex in the middle to lower crust in the Lofoten Islands region, which has been exhumed along detachments during large-scale extension, brought high-grade, lower crustal rocks, possibly including accreted decompressional melts, to shallower levels.     

Gold partitioning in melt-vapor-brine systems

We used laser-ablation inductively coupled plasma mass spectrometry to measure the solubility of gold in synthetic sulfur-free vapor and brine fluid inclusions in a vapor + brine + haplogranite + magnetite + gold metal assemblage. Experiments were conducted at 800°C, oxygen fugacity buffered at Ni-NiO (NNO), and pressures ranging from 110 to 145 MPa. The wt% NaCl eq. of vapor increases from 2.3 to 19 and that of brine decreases from 57 to 35 with increasing pressure. The composition of the vapors and brines are dominated by NaCl + KCl + FeCl₂ + H₂O. Gold concentrations in vapor and brine decrease from 36 to 5 and 50 to 28 μg/g, respectively, and the calculated vapor:brine partition coefficients for gold decrease from 0.72 to 0.17 as pressure decreases from 145 to 110 MPa. These data are consistent with the thermodynamic boundary condition that the concentration of gold in the vapor and brine must approach a common value as the critical pressure is approached along the 800°C isotherm in the NaCl-KCl-FeCl₂-HCl-H₂O system.We use the equilibrium constant for gold dissolution as AuOH⁰, extrapolated from lower temperature and overlapping pressure range, to calculate expected concentrations of AuOH⁰ in our experimental vapors. These calculations suggest that a significant quantity of gold in our experimental vapors is present as a non-hydroxide species. Possible chloridogold(I) species are hypothesized based on the positively correlated gold and chloride concentrations in our experimental vapors. The absolute concentration of gold in our synthetic vapor, brine, and melt and calculated mass partition coefficients for gold between these physicochemically distinct magmatic phases suggests that gold solubility in aqueous fluids is a function of aqueous phase salinity, specifically total chloride concentration, at magmatic conditions. However, though we highlight here the effect of salinity, the combination of our data with data sets from lower temperatures evinces a significant decrease in gold solubility as temperature drops from 800°C to 600°C. This decrease in solubility has implications for gold deposition from ascending magmatic fluids.     

Macropores as preferential flow paths in meander bends

Macropores are subsurface connected void spaces caused by processes such as fracture of soils, micro‐erosion, and fauna burrows. They are common near streams (e.g. hyporheic and riparian zones) and may act as preferential flow paths between surface and groundwaters, affecting hydrologic and biogeochemical processes. We tested the hydrologic function of macropores by constructing an artificial macropore within the saturated zone of a meander bend (open macropore, ‘OM’) and later filling its upstream end (partially filled macropore, ‘PFM’). For each treatment, we injected saline tracer at an upgradient monitoring well within the meander and monitored downgradient hydraulics and tracer transport. Pressure transducers in monitoring wells indicated hydraulic gradients within the meander were 32% higher perpendicular to and 6% higher parallel to the macropore for the OM than for the PFM. Additionally, hydraulic conductivities measured via falling head tests were 29 to 550 times higher along the macropore than in nearby sediment. We used electrical conductivity probes in wells and electrical resistivity imaging to track solute transport. Transport velocities through the meander were on average 9 and 21% higher (per temporal moment analysis and observed tracer peak, respectively) for the OM than for the PFM. Furthermore, temporal moments of tracer breakthrough analysis indicated downgradient longitudinal dispersion and breakthrough tracer curve tailing were on average 234% and 182% higher for the OM, respectively. This suggests the OM enabled solute transport at overall shorter timescales than the matrix but also increased tailing. Our results demonstrate the importance of macropores to meander bend hydrology and solute transport. Copyright © 2012 John Wiley & Sons, Ltd.     

Diamond in metasedimentary crustal rocks from Pohorje,Eastern Alps: a window to deep continental subduction

We report the first finding of diamond and moissanite in metasedimentary crustal rocks of Pohorje Mountains (Slovenia) in the Austroalpine ultrahigh‐pressure (UHP) metamorphic terrane of the Eastern Alps. Microscopic observations and Raman spectroscopy show that diamond occurs in situ as inclusions in garnet, being heterogeneously distributed. Under the optical microscope, diamond‐bearing inclusions are of cuboidal to rounded shape and of pinkish, yellow to brownish colour. The Raman spectra of the investigated diamond show a sharp, first order peak of sp³‐bonded carbon, in most cases centred between 1332 and 1330 cm^?1, with a full width at half maximum between 3 and 5 cm^?1. Several spectra show Raman bands typical for disordered graphitic (sp²‐bonded) carbon. Detailed observations show that diamond occurs either as a monomineralic, single‐crystal inclusion or it is associated with SiC (moissanite), CO₂ and CH₄ in polyphase inclusions_. This rare record of diamond occurring with moissanite as fluid‐inclusion daughter minerals implies the crystallization of diamond and moissanite from a supercritical fluid at reducing conditions. Thermodynamic modelling suggests that diamond‐bearing gneisses attained P–T conditions of ≥3.5 GPa and 800–850 °C, similar to eclogites and garnet peridotites. We argue that diamond formed when carbonaceous sediment underwent UHP metamorphism at mantle depth exceeding 100 km during continental subduction in the Late Cretaceous (c. 95–92 Ma). The finding of diamond confirms UHP metamorphism in the Pohorje Mountains, the most deeply subducted part of Austroalpine units.