Pyrite oxidation in river basins of India: Spatio-temporal study of sulfur and oxygen isotopes of dissolved sulfates

Abstract

Oxidation of pyrites, despite being a trace lithology, plays dominant role in key continental (acid-alkalinity balance), oceanic (Fe, S and trace elemental supply) and atmospheric (oxygen and CO₂ cycling) processes. This thesis work has aimed to investigate the spatial and temporal variations in the sulfide oxidation pattern of Indian river basins by using water chemistry and dissolved sulfur (δ³⁴S) and oxygen (δ¹⁸O) isotopes of sulfate fractions. The analysed samples for this work include (i) weekly collection of water from the upper Indus (at Leh) and Mahanadi (at Kusunpura) rivers for one-year duration, and (ii) spatial collection of water from Indus headwater basin. Forward and inverse modeling of these datasets were used to apportion the cationic and sulfate sources, which helped to assess their effect on carbon cycle. Time-series chemical and isotopic data for the Mahanadi river from the Peninsular India (from core-monsoon zone) were employed to constrain the seasonality in weathering processes in a transport-limited basin. Concentrations for most of the elements and also δ³⁴Sso₄ values show significant seasonal changes, with relatively higher values being observed for the lean-flow stages. The δ³⁴Sso₄ varies between 8.8 ‰ and 17 ‰, with a discharge-weighted average of 12 ± 4 ‰ (global riverine value ~ 4.4 ± 4.5 ‰). Calculations involving δ³⁴Sso₄ values and its possible sources compute the relative supply of sulfates from atmospheric deposition (38 %), gypsum (46 %), and sulfides (16 % (a lower limit)) to the river. The minimal contribution via sulfide oxidation confirms low CO₂ release (0.1 ± 0.1 tC/km²/yr) in this basin through sulfuric-acid mediated carbonate weathering. Mass balance calculations involving Ca, Mg and Si values confirm that these seasonal changes are mainly due to increase (by 18 ± 9 %) of groundwater supply to the stream during non-monsoon seasons. The silicate-derived cations (Cats) for the Mahanadi show minimal changes during monsoon (36 ± 5 %) and non-monsoon (33 ± 8 %) seasons. These data correspond to a CO₂ consumption rate of 2.4 ± 1.6 tC/km²/yr, which is about two times higher than that reported for global rivers. Low sulfide oxidation in this supply-limited regime is possibly linked to limited exposure of fresher minerals, whereas high silicate erosion in the basin is consistent with its silicate-dominated lithology and tropical climate. Further, efforts were also made to establish seasonal changes in pyrite oxidation and related controlling factors using time-series (weekly to biweekly) data of water chemistry and δ³⁴Sso₄ - δ¹⁸Oso₄ values for the Indus headwater at Leh, India. Data on δ³⁴Sso₄ (-2.6 to 2.9 ‰) and δ¹⁸Oso₄ (-5.9 to 2.6 ‰) exhibit significant variations, with relatively depleted values for the lean-flow samples. Apportionment of the cationic and sulfate sources using an inverse model also confirms seasonal changes. Fluctuations in Ca/Na* ratio and Cats indicate relatively higher silicate weathering during the early phase of summer than other seasons. This trend is different than earlier-reported enhanced carbonate weathering during the whole summer season for other Himalayan basins, and is linked to strong control of temperature on rock dissolution in early summer for the Indus basin. Further, the fraction of sulfide-derived sulfates is systematically higher during lean-flow than high-flow stages, attributable to diffusion of atmospheric oxygen to deeper weathering fronts with pyrite availability. This diffusion is promoted via microfractures, possibly produced by tectonic stresses along the shear zone of the basin, and limited by interflow of groundwater in this highly-elevated catchment. Outcomes of this study predict increase in subsurface pyrite oxidation and hence, river acidity in response to any future change (fall) in groundwater table. We also present new dataset on dissolved major ions, and δ³⁴Sso₄ and δ¹⁸Oso₄ for the Indus headwater basin. This dataset and its modeling were used to evaluate the net effect of Himalaya weathering on the global Cenozoic cooling event. The δ³⁴Sso₄ of these samples vary between - 11 ‰ and 5 ‰, with depleted δ³⁴Sso₄ values being observed for the mainstream samples (- 4 ± 2 ‰; n = 12) than that reported earlier for the Indus outflow (0.8 ‰). The δ18O of water samples (- 10 ± 3 ‰) and their sulfate fractions (- 1 ± 5 ‰) exhibit strong linear correlation, with their regression slope being a function of fpy (fraction of sulfide-derived sulfate). Inversion and Monte-Carlo simulation of the chemical and isotopic datasets were used to estimate the fraction of silicate-derived cations (0.2 ± 0.1 %; n = 61), and fpy (0.6 ± 0.2; n = 50) values. The low silicate weathering rate ((2.0 ± 0.1) 10⁵ mol/km²/yr) for this high-elevated and low relief terrain (at Batalik) is attributable to strong temperature control, whereas the high sulfide oxidation ((3.7 ± 0.4) 10⁵ mol/km²/yr) for this basin is consistent with the basin lithology dominated by Palaeozoic carbonates and organic-rich shales, and high glacial coverage. These rates confirm net release of CO₂ from the upper Indus basin, whereas modeling of literature data points to net CO₂ consumption at the Indus outflow. The net supply of CO₂ from the mountainous regions is in clear contrast to previous suggestion of significant CO₂ removal during Himalaya weathering and hence, challenges the role of land surface processes in the NW Himalaya in regulating the Cenozoic cooling event.