Application of 3D hydrogeochemistry and particle tracking in detecting groundwater flow patterns within an aapa mire–outwash plain system in a boreal environment at a mining development site

Abstract

A combination of statistical hydrogeochemical methods, PCA, correlation calculations and SiroSOM clustering, as well as groundwater flow modelling with particle tracking, was applied in the 3D hydrogeochemical characterization of the Sakatti mining development site, located in an area of complex Quaternary sediments and weathered/fractured bedrock in northern Finland. This study enhanced the understanding of the groundwater flow paths through the subsurface–mire complex, recharge areas of the study site, and chemical signatures of different water types.

The flow paths of the site are variable and occur at three scales: large-scale (>1000 m), medium-scale (1000–100 m) and small-scale flow (<100 m). Large-scale flow paths start from mire islands, continuing through subpeat sediments and the weathered and fractured bedrock, ending in the River Kitinen. Medium-scale flow paths go from the outwash plain areas and adjacent finger-like mire areas via the subsurface to the river. Small-scale flow occurs in the banks of the River Kitinen and in the Viiankiaapa mire and is related to variation in the topography, as well as the string and flark pattern of the mire. The average backward (BW)-tracked residence time is ca. 150, 25, 6 years for the large, medium, and small-scale flow paths, respectively.

The d-excess distribution displays areas with evaporated (<5), meteoric (>8), and mixed water signals. Particle tracking and d-excess values indicate that groundwater recharges in outwash plain areas and in the Viiankiaapa mire area. The low d-excess values in springs of the river bank and shallow bedrock wells indicate water flow from the Viiankiaapa mire.

The simulated BW tracking in outwash plain areas indicated average residences times corresponding to CFC-12 and ³H/³He ages. Evaluation of the likely recharge area with BW tracking based on a d-excess limit (<5 for evaporated waters) showed the best correspondence in GW wells and mire porewaters and was related to flow paths longer than 20 m. Surface water samples and spring samples displayed intermediate or low correspondence.

Most of the waters represent Ca–HCO₃ water types. SiroSOM analysis grouped the water samples into seven hydrogeochemical clusters including groundwaters with elevated concentrations to diluted surface waters. In general, the concentrations of elements were lowest in groundwater travelling through peat unit and highest in groundwater flowing in bedrock due to the longer residence time and flow path. The hydrogeochemical signal of anoxic, Fe- and Mn-rich mire (pore) waters is not evident in spring samples due to the oxidation of water in the springs and the precipitation of Fe and Mn.

Understanding of the groundwater flow patterns, as well as the recharge and discharge areas in a mining development site enables successful planning and positioning of the mining activities. Groundwater flow modelling can be used to plan and optimize groundwater monitoring in future mining sites.