Quantitative evaluation of factors controlling groundwater contamination in Jeju Island

Abstract

Among various types of groundwater contamination, nitrate has been a major threat with regard to the deterioration of groundwater resources in many countries. The potential vulnerability of aquifers to nitrate pollution can intensify as human activities increase. Therefore, it is essential to protect water quality levels by establishing proper management plans which consider dominant factors that can be used to control the nitrate dynamics in subsurface aquifer systems. On Jeju Island, as groundwater is the sole water resource, severe nitrate contamination of groundwater is a major concern. The island is characterized by complex hydrogeological systems and various anthropogenic activities

therefore, a systematic assessment that characterizes the complex nitrate dynamics of the island is necessary prior to the formulation of any water quality protection plan. In this study, through a combination of investigative approaches, factors controlling the nitrate dynamics are quantified in an effort to identify the mechanisms of nitrate contamination on Jeju Island. First, the overall spatio-temporal distributions of NO3-N concentrations along with Cl in groundwater throughout Jeju Island were identified to evaluate ongoing and continuing harmful impacts on the water quality. The determined temporal trends of these two elements were then used to assess the effects of land usage changes and groundwater management actions on groundwater quality levels. The results indicated that upward trends in NO3-N were associated with expansions of agricultural lands, whereas the Cl trends were considered to be affected by regulations on groundwater extraction to reduce seawater intrusion. The deterioration of the quality of groundwater by nitrate is a continuous problem on Jeju Island, especially in the western part of the island. Based on the results of the trend analyses conducted as part of the present study, a focused study related to the continuous threat of NO3-N as a contaminant on Jeju Island was conducted in the Gosan agricultural area in the western part of the island. This was done using a combined approach involving a numerical simulation and age-dating methods. In the Gosan study area, which contains a layered aquifer system penetrated by leaky wells, the quantification of well leakage effects was undertaken by utilizing a leaky-well module and a double-domain integration method to compute the degree of nitrate cross-contamination through the layered aquifer system. The numerical results demonstrated that the well leakage flux rapidly degraded the water quality of the underlying regional aquifer by acting as a direct pathway for nitrate-rich shallow perched groundwater. Based on predictions by the developed model with regard to decreases in the NO3-N concentration at regional groundwater wells, the maximum allowable fertilizer amount for Gosan would be 45%–65% of the currently applied fertilizer level, whereas sealing of the regional groundwater wells would rapidly decrease the NO3-N concentration while also mitigating the need to reduce fertilizer usage levels. Given the complex hydrogeological features of the Gosan area, mixing of groundwater ages is expected due to migration through various flow pathways. Therefore, three age-estimation methods were used to interpret groundwater ages (environmental tracer age dating, a lumped parameter model and a numerical model) so as to characterize the nitrate input history and determine which major pathways influence the quality of the groundwater. The results of the three age-estimation methods were compared to establish suitable groundwater management strategies. These results showed that using different age estimation methods leads to variations in the estimated contaminant loading history and, accordingly, different groundwater management strategies. The discrepancies in the age estimations produced by the different models were more prominent in the complex hydrogeological system. For this reason, it is necessary to apply multiple age estimation methods and compare the results based on an interpretation of the full age distributions.