Conceptualizing a multi‐layered shingle aquifer model based on volcanic stratigraphy and water inflow to lava caves in Jeju Island, Korea

Abstract

AbstractVolcanic aquifers supply a substantial portion of water resources in many parts of the world, including islands, and their productivity depends strongly on volcanic stratigraphy, which exhibits considerable heterogeneity. We investigated water inflow to lava tube caves formed from numerous basaltic lava flows in the northeastern coastal area of Jeju Island after storm events and monitored relative inflow rates monthly over 1 year to characterize groundwater flow processes in the upper parts of volcanic aquifers, and to evaluate the applicability of the previous hydrogeological models proposed for the island. Considerable water inflow arose shortly after storms from exposed palaeosol layers on the walls of the caves. The monthly monitoring results showed that wall inflow associated with these palaeosol layers is substantial. In both cases, discharge from ceiling drips was much less and more temporally variable compared to wall inflow discharge. Water flowing into the caves was rapidly drained through the floor at all monitoring sites. The lateral extent of the palaeosol layers was identified using drill core logs near the cave and outcrops in the coastal area. Based on these results, we inferred that multiple perched aquifers are formed by low‐permeability palaeosol layers between lava flows, which are connected by vertical flows at discontinuities in the palaeosol layer, eventually reaching the basal aquifer. This study revealed the water inflow processes observed in lava tube caves constrained by palaeosol layers, and established a hydrogeological conceptual model incorporating multiple perched aquifers in both coastal and mountainous areas associated with extensive palaeosol layers formed during volcanic hiatuses. This finding would help elucidate recharge, groundwater flow, and contaminant transport processes in many volcanic aquifers that are not adequately represented by the previous models, and contribute to better management of groundwater in those areas.