Geo- and Environmental chemistry
Water, soft rock & hard rock
Banner photo is from fieldwork in western Arabian where large wadis incise basalt flows.
Water: Source apportionment of chemicals of emerging concern
Recently we have been exploring how novel inverse (a form of AI) models can be used to identify the sources of chemicals in waterways. We think that doing so provides a step-change in capabilities to identify sources of contaminants of emerging concern, e.g., pesticides, illict drugs and pharmaceuticals in waterways. We have published a proof-of-concept study of a tributary of the River Thames in Chrapkiewicz et al. (2024).
Figure: Graphical abstract from Chrapkiewicz et al. (2024).
Soft rock: Composition of eroding landscapes and material fluxes
We have established that concentrations of geologically-relevant chemicals in sediments within drainage networks can be used to identify their sources by combining new and existing observations with forward and inverse modelling of river bed sediments (e.g. Lipp et al., 2020, 2021). Conservative mixing has been shown to provide reliable predictions of compositions along rivers.
Figure: Forward model predictions of magnesium concentrations along rivers draining the Cairngorms and their residuals with observations. (a) Circles = measurements of magnesium concentration in Scottish rivers. Coloured curves = predicted concentration calculated by forward modelling using independent observations of source region geochemistry and mixing calculated using Scotland's drainage patterns. (b & c) Comparison of observed (circles in panel a) and predicted (curves in panel a) chemistry. See Lipp et al. (2020) for more details.
Figure: Inverting downstream sediment samples (circles in previous figure) for concentration of magnesium in source regions. (a) Optimum upstream concentration of magnesium generated by inverting the magnesium concentration of 67 samples gathered downstream (see Lipp et al., 2021 for details). (b) Independent Geochemical Baseline Survey of the Environment (G-BASE) stream sediment concentration of magnesium gridded to same resolution as panel (a). (c) Cross-plot of observed (G-BASE) and predicted concentrations for each grid cell (5 km resolution). Colors show misfit discretized at intervals equal to global root-mean-square (RMS) misfit (0.195). Gray dashed line = 1:1 relationship; black line = linear regression. (d) Misfit between observed magnesium concentration and best-fitting inverse model. Inset indicates distribution of residuals and normal distribution; bin-width = global RMS misfit (0.195).
Hard rock: Mantle melting
We have sought to identify the origins of topographic support and the history of mantle convection using hard rock geochemistry (e.g. thermobarometry of mafic melts), geophysical and surface observations in concert. This figure shows an example of doing so for Borneo (see Roberts et al., 2018, for more details).
Figure: Deep structure of Borneo. (a) NW-SE-NE transect, x–x′, showing long wavelength free-air gravity anomaly (see inset). (b) Uplift along transect from sub-plate temperature variation determined by calibrating seismic tomographic model shown in panel (d). Solid circles with vertical bars = air-loaded uplift, assuming that asthenosphere has ambient temperature of 1333+/-30 ◦C. (c) Topographic height along transect. Dashed line = mean sea level. (d) Vertical slice through seismic tomographic model of Schaeffer and Lebedev (2013). Gray polygon = regions where calculated temperature <1333 ◦C; dashed line = position of 1333 ◦C isothermal surface. Gray/black arrows = loci of two calculated geothermal profiles shown in panels (d) and (e). (e) Contour map = temperature as a function of shear wave velocities, Vs(z), and depth (Priestley and McKenzie, 2006). Gray/black circles = Vs(z) beneath central Borneo and Sunda Shelf, respectively (see arrows in panel d). (f) Pressure–temperature calculations. Gray/black circles = geothermal profiles beneath central Borneo and Sunda Shelf from Vs conversion scheme shown in panel (d). White circle = pressure and temperature estimates of primary melt for sample SBK13 with MgO number of 7.66 wt.% from Semporna peninsula (Macpherson et al., 2010). Pressures and temperatures were calculated using mafic thermobarometric method (Plank and Forsyth, 2016). Uncertainties in pressure and temperature estimates were determined for range of sample H2O and Fe oxidation contents (McNab et al., 2018; see body text). Dotted line projected to surface indicates mantle potential temperatures (T p ≈ 1380 ◦C). Pressure–temperature estimates are compared to anhydrous melt paths (gray polygon) calculated using Katz et al. (2013)’s formulation. Dashed line is their anhydrous solidus.