Abstract
Numerical groundwater models were used to assess groundwater sustainability on Jeju Island, South Korea, for various climate and groundwater withdrawal scenarios. Sustainability criteria included groundwater-level elevation, spring flows, and salinity. The latter was studied for the eastern sector of the island where saltwater intrusion is significant. Model results suggest that there is a need to revise the current estimate of sustainable yield of 1.77 × 106 m3/day. At the maximum extraction of 84 % of the sustainable yield, a 10-year drought scenario would decrease spring flows by 28 %, dry up 27 % of springs, and decrease hydraulic head by an island-wide average of 7 m. Head values are particularly sensitive to changes in recharge in the western parts of the island, due to the relatively low hydraulic conductivity of fractured volcanic aquifers and increased groundwater extraction for irrigation. Increases in salinity are highest under drought conditions around the current 2-m head contour line, with an estimated increase of up to 9 g/L under 100 % sustainable-yield use. The study lists recommendations towards improving the island’s management of potable groundwater resources. However, results should be treated with caution given the available data limitations and the simplifying assumptions of the numerical modeling approaches.
Zusammenfassung
Numerische Grundwassermodelle wurden verwendet, um Grundwassernachhaltigkeit für verschiedene Klima- und Grundwasserentnahmeszenarien auf der Insel Jeju, Südkorea, zu beurteilen. Nachhaltigkeitskriterien umfassen Grundwasser-Pegel-Messung, Quellfluss und Salzgehalt. Letzterer wurde für den Ostteil der Insel untersucht, wo das Eindringen von Salzwasser bedeutend ist. Modellergebnisse deuten darauf hin, dass es notwendig ist, die aktuelle Schätzung des nachhaltigen Ertrags von 1.77 × 106 m3/day zu überarbeiten. Bei der maximalen Entnahme von 84 % des nachhaltigen Ertrags würde sich beim zehnjährigen Dürreszenario der Quellfluss um 28 % verringern, 27 % der Quellen vertrocknen und die hydraulische Höhe der Insel im Durchschnitt um 7 m verringert werden. Hydraulische Höhewerte sind besonders anfällig auf Veränderungen der Grundwasserneubildung im westlichen Teil der Insel, aufgrund der relativ geringen hydraulischen Leitfähigkeit des gebrochenen vulkanischen Grundwasserleiters und der erhöhten Grundwasserentnahme zur Bewässerung. Erhöhter Salzgehalt ist am höchsten unter Dürre um die gegenwärtige 2-m hydraulische Höhe-Konturlinie, mit einem geschätzten Anstieg von bis zu 9 g/l bei 100 % Gebrauch des nachhaltigen Ertrags. Die Studie nennt Empfehlungen zur Verbesserung der Verwaltung des Trinkwasservorkommens der Insel. Allerdings sollten die Ergebnisse mit Vorsicht behandelt werden, angesichts der eingeschränkten verfügbaren Daten und der vereinfachenden Annahmen der numerischen Modellierungsansätze.
Resumé
Des modèles numériques ont été utilisés pour évaluer la durabilité des eaux souterraines sur l’île de Jeju, Corée du Sud, pour différents scénarios climatiques et de prélèvements des eaux souterraines. Les critères de durabilité comprennent l’altitude du niveau piézométrique, les débits des sources et la salinité. Cette dernière a été étudiée dans le secteur est de l’île où l’intrusion saline est forte. Les résultats des modèles suggèrent qu’il est nécessaire de réviser l’estimation actuelle du débit durable de 1.77 × 106 m3/j. Au maximum de prélèvement de 84 % du débit durable, un scénario de sécheresse de période de retour 10 ans réduirait le débit des sources de 28 %, assècherait 27 % des sources et diminuerait les niveaux piézométriques de 7 m en moyenne à l’échelle de l’île. Les niveaux piézométriques sont particulièrement sensibles aux variations de la recharge dans les parties ouest de l’île, en raison de la perméabilité relativement faible des aquifères volcaniques fracturés et de l’augmentation des prélèvements d’eau souterraine pour l’irrigation. Sous conditions de sécheresse, les augmentations de la salinité sont maximales aux alentours de la ligne isopièze 2 m, avec une hausse estimée atteignant 9 g/L pour l’usage de 100 % du débit durable. L’étude dresse la liste des recommandations destinées à améliorer la gestion des ressources en eau potable souterraines. Cependant, les résultats doivent être considérés avec prudence compte tenu des limites des données disponibles et des hypothèses simplificatrices des approches de modélisation numérique.
Resumen
Se usaron modelos numéricos de agua subterránea para evaluar la sustentabilidad del agua subterránea en la isla Jeju, Corea del Sur, para varios escenarios de clima y de extracción de agua subterránea. Los criterios de sustentabilidad incluyeron la elevación del nivel de agua subterránea, los flujos de manantiales, y la salinidad. Esta última se estudió para el sector este de la isla donde es significativa una intrusión de agua salada. Los resultados sugieren que hay una necesidad de revisar las estimaciones actuales de rendimiento sustentable de 1.77 × 106 m3/d. Con una extracción máxima del 84 % del rendimiento sustentable, un escenario de sequía de diez años reduciría los flujos hacia los manantiales en un 28 %, secándose hasta un 27 % de los manantiales, y una disminución de la carga hidráulica en toda la isla en un promedio de 7 m. Los valores de carga hidráulica son particularmente sensibles a los cambios en la recarga en las partes oeste de la isla, debido a la relativamente baja conductividad hidráulica de los acuíferos volcánicos fracturados y a un incremento de la extracción de agua para irrigación. Los incrementos en la salinidad son más altos bajo condiciones de sequía alrededor de la línea de carga hidráulica actual de 2 m, con un incremento estimado de hasta 9 g/L bajo un rendimiento sustentable de 100 %. El estudio lista las recomendaciones para una mejora del manejo de los recursos de agua subterránea potable de la isla. Sin embargo, los resultados deben ser tratados con precaución dadas las limitaciones en la disponibilidad de datos y los supuestos simplificados de las aproximaciones del modelado numérico.
الملخص
استخدمت النماذج العددية لمحاكاة حركة الجريان الجوفي لتقييم استدامة المياه الجوفية في جزيرة جي جو بكوريا الجنوبية في ظل سيناريوهات مختلفة للمناخ ومعدلات سحب المياه الجوفية . شملت معايير الاستدامة على منسوب مستوى المياه الجوفية، وجريان الينابيع ودرجة الملوحة . وتم دراسة درجه الملوحة في القطاع الشرقي من الجزيرة حيث يتعاظم تداخل مياه البحر المالحة . أوضحت نتائج النمذجه الحاجة إلى تعديل القيمة الحالية لمعامل الانتاج المستمر (1.77 × (106 × 1.77 m3/day). عند اقصى سحب بقيمة %84 من معامل الانتاج المستمر، سيناريو عشر سنوات جفاف سيقلل جريان الينابيع بنسبة %28 ويجفف 27 % من الينابيع ويقلل الضاغط المائي بقيمة متوسطة 7 أمتار بالجزيرة . بالإضافة الى ان قيم الضاغط المائي حساسة، على وجه الخصوص، للتغيرات في شحن الخزان الجوفي في الأجزاء الغربية من الجزيرة وذلك نتيجة للانخفاض النسبي لقيم معامل التوصيل الهيدروليكي للخزانات الجوفية الفولكانيه ذات التشققات وأيضا زياده سحب المياه الجوفية لغرض الري . وأوضحت النتائج أيضا أن زيادة الملوحة للمياه تصل لأعلى قيمة في ظل ظروف الجفاف عند خط الكنتور الحالي ذو القيمة 2 متر للضاغط المائي مصحوبا بزياده متوقعه إلى 9 g/Lعند استخدام معامل الانتاج المستمر بقيمة 100 % . ويجب الإشارة إلي توخى الحذر عند تطبيق ما انتهت اليه الدراسة من توصيات لتحسين إدارة الجزيرة لمصادر المياه الجوفية المستخدمة للشرب وذلك لمحدودية البيانات المتوفرة إلي جانب الافتراضات المبسطة التي اسس عليها النموذج العددي المستخدم في الدراسة .
摘要
数值地下水模型用于评价韩国Jeju岛各种气候和地下水抽取情况下的地下水的可持续性。可持续性标准包括地下水位海拔、泉流和盐度。对岛的东部进行了盐度的研究,因为那里的海水入侵很严重。模型结果显示需要修订目前可持续出水量为1.77 × 106 m3/day的估算值。在最大抽取量为可持续出水量的84%时,10年干旱方案泉流将降低28%,27%的泉将干涸,全岛范围内的水头将降低7米。水头值对岛西部补给变化特别敏感,因为断裂火山岩含水层导水系数相对较低及农业用地下水抽水量增加。在干旱条件下盐度在目前2 米水头等高线附近增加最高,在100%可持续出水量利用情况下,估计增加到每升9克。本研究列出了改进岛上饮用水管理的建议。然而,研究结果应该谨慎对待,因为现有的资料有限,数值模拟方法假定简化。
초록
대한민국 제주도에서 지하수 유동 모델링 기법을 이용하여, 다양한 기후변화 및 지하수 이용 시나리오에 따른 지하수 지속가능성을 평가하였다. 지속가능성을 평가하는 기준으로는 지하수위, 용천수량, 염도 등을 포함시켰다. 염도 기준은 지하수 내 염수침입이 빈번한 제주도 동부지역에 적용되었다. 모델링 결과, 현재의 지속가능 개발량인 1.77 × 106 m3/day가 수정될 필요가 있음을 알 수 있었다. 지속가능 개발량의 최대 84%의 지하수를 사용하고 10년 빈도 가뭄 시나리오를 적용했을 경우, 용천수량은 28%정도 줄어 들어, 전체의 27%가 말라버리고, 지하수위는 평균 7m정도 감소하는 것으로 평가되었다. 지하수위는 제주도 서부지역에서 지하수 함양량에 특히 민감하였는데, 이는 이 지역의 대수층이 상대적으로 수리전도도가 낮고 지하수가 농업용 관개용수로 많이 사용되기 때문이다. 염도는 가뭄시 지속가능 개발량의 100%의 지하수를 사용하였을 때 가장 높았으며, 현재 지하수위 2 m 지점에서는 9 g/L까지 증가하였다. 또한, 본 연구에서는 모델링 결과를 토대로 섬지역에서 가용한 지하수 자원을 효과적으로 관리하기 위한 대안들을 제시하였으나, 모델링 결과는 입력자료가 불충분하고 모델링에 필요한 간략화된 가정들 때문에 한계를 가지고 있어, 주의깊게 다루어 져야 한다.
مقدمه
جهت بررسی پایداری آبهای زیرزمینی جزیره جیجو، کره جنوبی، با توجه به سناریوهای برداشت آب زیرزمینی و اقلیم گوناگون، مدلهای عددی آب زیرزمینی به کار گرفته شدهاند. معیارهای پایداری، تراز سطح آب زیرزمینی، دبی چشمهها و شوری در نظر گرفته شده است. معیار شوری برای بخش شرقی جزیره که نفوذ آبهای شور قابل توجه است مورد مطالعه قرار گرفته است. نتایج مدل بیانگر نیاز به اصلاح برآورد حاضر از جریان آورد پایدار به میزان m3/day 107 × 77.1 میباشد. با تخلیه بیشینه 84 % از جریان آورد پایدار، و در نظر گرفتن یک سناریوی خشکسالی ده ساله، 28 % از دبی جریان چشمهها کاهش خواهد یافت، 27 % چشمهها خشک شده، و هد هیدرولیکی به طور میانگین 7 متر در سطح جزیره کاهش مییابد. مقادیر هد به علت هدایت هیدرولیکی نسبتاً پایین آبخوانهای آتشفشانی گسسته و افزایش استخراج آب زیرزمینی برای کشاورزی، خصوصاً به تغییرات سطح تغذیه در بخشهای غربی جزیره، حساس میباشد. میزان سطح شوری تحت شرایط خشکسالی، پیرامون خط تراز هد 2 متر، با یک سطح افزایش تخمینی تا حدود g/L9 با 100 % استعمال از آورد پایدار، به بالاترین میزان خود میرسد. این تحقیق پیشنهادهایی را جهت بهبود مدیریت منابع آب زیرزمینی آشامیدنی جزیره ارائه میکند. هر چند که این نتایج باید با توجه به محدودیت دادههای موجود و فرضیات سادهسازی روشهای مدلسازی عددی، با احتیاط لازم مورد بحث قرار گیرند.
Resumo
Foram utilizados modelos hidrogeológicos numéricos para avaliar a sustentabilidade das águas subterrâneas na Ilha de Jeju, Coreia do Sul, para vários cenários climáticos e de extração de água subterrânea. Os critérios de sustentabilidade incluíram os níveis de água subterrânea, a produção de nascentes e a salinidade. Esta última foi estudada para o sector oriental da ilha onde é significante a intrusão salina. Os resultados do modelo sugerem que há necessidade de rever a atual estimativa de caudal sustentável de 1.77 × 106 m3/d. À taxa de extração máxima de 84% do caudal sustentável, um cenário de seca de 10 anos de duração provocaria uma redução de 28% no caudal das nascentes, secaria 27% das nascentes e reduziria o nível da água subterrânea médio à escala da ilha em 7m. Os níveis de água subterrânea são particularmente sensíveis a alterações na recarga na parte ocidental da Ilha devido à relativamente baixa condutividade hidráulica dos aquíferos de rocha vulcânica fraturada e ao aumento da extração de água subterrânea para irrigação. O aumento da salinidade é máximo em condições de seca em torno da atual linha de nível de 2m, com um incremento estimado de até 9 g/L em condições de 100% de uso do caudal sustentável. O estudo enumera recomendações destinadas à melhoria da gestão dos recursos de água potável da ilha. No entanto, os resultados deverão ser tratados com cautela atendendo às limitações dos dados disponíveis e às condições simplificadas adotadas nas abordagens por modelação numérica.
Öz
Sayısal yeraltısuyu modelleri kullanılarak Güney Kore’nin Jeju adasında değişik iklim ve yeraltısuyu çekim senaryoları için yeraltısularının sürdürülebilirliği değerlendirilmiştir. Sürdürülebilirlik kriterleri yeraltısuyu seviyesi, pınar debileri ve tuzluluk oranlarını içermektedir. Tuzluluk kriteri, tuzlusu girişiminin etkin olduğu adanın doğu kısımları için çalışılmıştır. Model sonuçları, 1.77 × 106 m3/gün olan güncel tahmini sürdürülebilir yeraltısuyu veriminin gözden geçirilerek yeniden hesaplanması gerektiğini önermektedir. Maksimum yeraltısuyu çekimi sürdürülebilir verimin %84’ü olduğu takdirde, on-yıllık kuraklık senaryosunda pınar debileri %28 azalmakta, pınarların %27 si kurumakta ve adanın tamamında hidrolik su seviyeleri ortalama 7 m azalmaktadır. Çatlaklı volkanik akiferlerdeki nispeten düşük hidrolik iletkenlik ve sulama sebebiyle artarak devam eden yeraltısuyu çekiminden dolayı, su seviyesi değerleri, adanın batı kısımlarında yeraltısuyu beslenimindeki değişiklere karşı özellikle hassastır. Sürdürülebilir yeraltısuyu veriminin %100’ünün kullanıldığı kuraklık senaryosunda, mevcut 2-m’lik su seviyesi eşyükselti eğrisi etrafındaki yerlerde tuzluluktaki artış en yüksek değeri olarak tahmin edilen 9 g/L’ye ulaşmaktadır. Bu çalışmada, adanın kullanılabilir yeraltısuyu kaynakları yönetiminin geliştirilmesine yönelik öneriler sıralanmıştır. Ancak, mevcut dataların sınırlı olması ve sayısal modelleme yaklaşımında kullanılan basitleştirici varsayımlardan dolayı, model sonuçları ihtiyatla kullanılmalıdır.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
In groundwater-resource assessments, terms such as sustainability, sustainable yield, sustainable pumping, sustainable development, and safe yield are often used interchangeably or with little distinction between the concepts. Definitions and approaches to assess sustainability of potable drinking-water resources are summarized by Bredehoeft (2002), Kendy (2003), Alley and Leake (2004), Maimone (2004), Devlin and Sophocleous (2005), Liu et al. (2006), and Bredehoeft (2007), among others. Sophocleous (2000), for instance, defines sustainable development, in general, as development that “meets the needs of the present without compromising the ability of future generations to meet their own needs.” Specifically, optimal or sustainable groundwater-resource use requires setting upper limits on water withdrawal (or sustainable yield) to avoid compromising the source (Mink 1992; Mink and Lau 1990; Takasaki and Mink 1985). Although the concept is widely practiced by state and federal government agencies, sustainable yield has come under heavy criticism for both theoretical and practical reasons. Sustainable yield is not always easy to estimate or utilize in practice. Estimation problems arise due to poor assumptions in models and the lack of reliability of available data. For example, although water use in Hawaii is subjected to regulations regarding sustainable yield, studies have shown rising salinity and decreasing freshwater resources (Oki 2005).
On islands where potable water resources are limited, an accurate and reliable estimate of sustainable use or yield is critical. Estimates have to address not only decreasing water availability, as manifested by declining water levels, but also deterioration of water quality caused by, e.g., seawater intrusion (Kim et al. 2006) or agro-chemical use (Koh et al. 2007).
On Jeju Island, as is the case for many islands, sustainability of groundwater resources is especially important given the increasing water demands from urban and agricultural development and the island’s lack of reliable surface-water resources (Hagedorn et al. 2011; Won et al. 2006). The island has experienced tremendous growth over the past 30 years and now supports a residential population of 583,000 and 8.7 million visitors annually. Groundwater aquifers supply 92 % of freshwater consumed on Jeju (Kim et al. 2003a, b). Because of the rapid growth and increased demand for potable water, saltwater intrusion has become a persistent problem, particularly in the eastern portion of the island (Kim et al. 2003a, b, 2006).
Examples of approaches to define and estimate sustainable aquifer yield are those implemented in Hawaii and Jeju. On the Hawaiian Islands, the sustainable yield of an aquifer is defined by the Hawaii State Water Commission (1979) as the “water supply that may normally be withdrawn from a source at the maximum rate, which will not severely impair source utility.” Mink (1980) notes that the clearest expression of sustainable yield is that of allowable net draft for a selected equilibrium head. In a subsequent publication, Mink (1981) developed the robust analytical model (RAM) for estimating the sustainable yield of a basal aquifer. However, the model does not take into account the effects of transport processes due to saltwater–freshwater interaction. The analytical RAM2 model was developed as an extension of the original RAM model to include the simulation of the salinity distribution in a transition zone (Liu 2007). The RAM and RAM2 models are currently used to estimate the sustainable yield of basal aquifers in the Hawaiian Islands (Wilson Okamoto Corporation 2008). However, these analytical models are limited in their ability to evaluate spatio-temporal effects of climate and withdrawal scenarios on groundwater resources.
Based on detailed analysis of aquifer testing data on Jeju Island, Hahn et al. (1997) estimated sustainable yield by combining mass balance with Darcy’s law. In this regard, the authors set an upper limit for the ratio between the equilibrium water level to be maintained and the initial water level in the pre-development time for each area under consideration. While this approach is very useful for providing a cursory estimate, it is subjected to significant uncertainty due to data limitations, considering that initial water-level data for pre-development times are rarely available. A need exists for a more comprehensive, physically based approach that utilizes site-specific data.
Hahn et al. (1997) and the Korean Water Resources Corporation (KOWACO 2003a) used a set ratio of 80 % between the equilibrium water level to be maintained and the initial water level in the pre-development time to estimate the sustainable yield of Jeju’s aquifers. Koh et al. (2006b) reported the most recent estimate of sustainable yield by KOWACO as 1.77 × 106 m3/day. KOWACO also specifies a maximum allowable aquifer withdrawal rate (MAWR) of 1.48 × 106 m3/day for domestic, industrial, and irrigation purposes (KOWACO 2003a), which is roughly 84 % of the estimated sustainable yield. Estimates of groundwater recharge on Jeju range from 4.09 × 106 to 4.80 × 106 m3/day (Hahn et al. 1997; KOWACO 2003b; Koh et al. 2006b; Won et al. 2006). Thus, the limits on aquifer pumping as a percentage of groundwater recharge range from 37 to 43 % for the sustainable yield and from 31 to 36 % for the MAWR. By 1993, total daily pumping averaged 0.66 × 106 m3/day from all of Jeju’s domestic, municipal, and irrigation water-supply wells (Hahn et al. 1997). The data implies that pumping from Jeju’s aquifers by the mid-1990s was already at 37 % of the sustainable yield and 45 % of the MAWR. Although accurate estimates of pumping rates from all sources on Jeju in recent years are lacking, it is highly likely that total pumping now exceeds the rate in 1993 of 37 %. Based on discussion with various government groups, a value of total use of 64 % of the sustainable yield was adopted in the current study based on a best estimate.
In recent years, hydrological models, with various degrees of sophistication, have been used in various settings to evaluate water budgets under various climate or land-use scenarios (e.g., Mink 1981; Gingerich 2008). The report by the Minnesota Department of Natural Resources (2010) evaluates models and tools needed for assessing water availability and sustainability. In general, a water-budget model is used in estimating recharge, which is used as input to groundwater models. Sustainability indicators such as acceptable head or water-quality levels, are examined based on climate and management decisions affecting water extraction and land use. Due to limitations in models and data availability for model calibration, field monitoring of sustainability indicators should be an important part of any sustainability planning effort.
The objective of this study is to assess sustainable groundwater use for Jeju Island under projected climate, land use, and increased-pumping conditions. Numerical modeling is used in assessing the sustainability by evaluating the response of the groundwater system to different climate, land use, and pumping scenarios. The results for each scenario are evaluated by assessing negative effects on groundwater sustainability indicators, namely, hydraulic head, spring flow, and salinity.
Study site
Jeju Island is a mid-latitude volcanic island located about 140 km south of the coast of South Korea in the transitional region between the temperate and subtropical climate zones (Park 2010) (Fig. 1). About half of the mean annual rainfall occurs in the summer monsoon between June and September, with average monthly rainfall higher than 200 mm (KOWACO 2003b). Winters are dry and cold, with occasional snowfall at the higher elevations. About half of the island is covered with natural forest vegetation, while other areas are covered by pasture, grassland, cropland, and increasing urban development along the coastline (KOWACO 2003b). As a Quaternary shield volcano, Jeju is mainly composed of basaltic and trachytic lavas and associated sediments (Won et al. 2006). In addition, nearly 400 scoria cones and tuff rings are scattered across the island. According to Sohn (1996), the major geologic units for Jeju Island are subdivided from top to bottom as follows (Fig. 2):
-
1.
The main volcanic lithologies consisting of permeable basalt, andesite, and trachyandesite lavas, with numerous flow units and abundant fractures
-
2.
The Seogwipo Formation, comprising low permeability volcanic sediments with an average thickness of about 100 m (Nam et al. 2009)
-
3.
The U-Formation, including low permeability unconsolidated silt and sand deposits with an average thickness of approximately 150 m (Nam et al. 2009)
-
4.
The relatively impermeable basement, which is mainly composed of welded tuffs and granite, occurring at depths of about 250–300 m below sea level (Nam et al. 2009)
Recharge occurs predominantly in the high rainfall inland areas of the island (Mair et al. 2013a, b), whereas volcanic lithologies (i.e., basalts, andesites, and trachytes) comprise the main aquifers (Hahn et al. 1997; Won et al. 2005). Groundwater flow occurs mainly through rock fractures and along interflow structures (Hamm et al. 2005). Groundwater level contours roughly follow the topography with highest values in the interior (about 300 m asl) and lowest levels towards the coast (about 2 m asl).
Methods
The numerical groundwater models MODFLOW (Harbaugh et al. 2000) and SEAWAT (Langevin et al. 2007) were applied to assess groundwater-resources sustainability criteria under various climate and groundwater-withdrawal scenarios. MODFLOW was used to develop an island-wide model to evaluate water levels and spring flows, assuming saturated flow and constant water-density conditions. Applicability of this widely used model in island modeling is limited due to the lack of density-change accountability; however, the model can be reasonably calibrated for sites that are farther away from the shoreline (e.g., Whittier et al. 2010). More recently, the SWI2 Package (Bakker et al. 2013) has been added to MODFLOW, which makes the model more appropriate by including variable-density groundwater flow.
SEAWAT combines MODFLOW with MT3DMS (Zheng 2006; Zheng and Wang 1999) in simulating three-dimensional (3D) variable-density groundwater flow. SEAWAT has been applied to saltwater intrusion and up-coning (Shoemaker and Edwards 2003; Masterson 2004; Masterson and Garabedian 2007; Dausman and Langevin 2005; Heywood and Pope 2009; Sanford et al. 2009; Hughes et al. 2009), and it has been also utilized in assessing submarine groundwater discharge (Langevin 2001, 2003; Dausman et al. 2007).
In this study, SEAWAT was used to incorporate the effects of density-driven groundwater flow to assess salinity for the eastern sector of the island where significant saltwater intrusion has been reported (Koh et al. 2006a). MODFLOW and SEAWAT were applied within the groundwater modeling system (GMS) graphical user interface (Aquaveo 2011).
Spatially distributed estimates of groundwater recharge computed by Mair et al. (2013b) were input into the numerical groundwater model. These authors used a deterministic, quasi-3D physically based soil-water balance model developed by Westenbroek et al. (2010) and inputs of soil, land cover, topographic, and climatic data to estimate potential groundwater recharge at a daily time step and 100-m-grid cell size. They estimated potential recharge for baseline, drought, and future climate–land-use scenarios. The baseline scenario was comprised of an 18-year period during 1992–2009, while the drought scenario was developed for a 10-year historical drought during 1963–1972 as identified by Mair et al. (2013a). A future climate–land-use scenario was created to simulate projected changes in surface temperature and land use. Projected future increases in average temperature during 2070–2099 under an A2 emission scenario as reported by Min et al. (2004) were simulated by adjusting the 18-year-baseline-scenario temperature dataset using a seasonally variable change factor approach. Based on a comparison of historical trends with projected changes in rainfall, Mair et al. (2013b) made no changes to the 18-year precipitation dataset for the future climate–land-use-change scenario. Land-use change was incorporated by increasing the urban–residential land-use type by 46 %; henceforth, the term climate scenario will be used to describe future climate–land-use changes.
The models were calibrated for the baseline recharge scenario to match average island-wide water levels, average spring flows at selected locations near the coast, and groundwater salinity. The calibration was done under a pumping rate at the 64 % of sustainable yield. Following calibration, simulations were run for the three groundwater-recharge scenarios—namely, baseline, drought, and climate—each under a number of pumping scenarios for MODFLOW. A total of seven pumping scenarios were considered for each recharge scenario by adjusting pumping to 18, 34, 51, 64, 67, 84, and 100 % of the estimated sustainable yield. The pumping rate at the 84 % level approximates the reported groundwater MAWR, which is the maximum amount of pumping allowed on Jeju (KOWACO 2003a). For this specific study, SEAWAT was not practical for a full-scale analysis, due to its computational burden. Hence, the model was only run for a worst-case scenario of withdrawals at 100 % of sustainable yield.
Conceptual model
Groundwater on Jeju is classified into four general categories: high level, parabasal, basal, and basement (Won et al. 2006 and references therein; Fig. 3). It is postulated that high-level groundwater is perched by low-permeability volcanic layers at high levels in the mountainous interior of the island without a direct hydraulic connection to the other types of groundwater. Modeling the dynamics of this type of groundwater is complicated by the following factors.
-
1.
The lack of understanding of the spatial distribution of low permeability volcanic sequences that may impede vertical flow and perch groundwater at high levels
-
2.
The lack of information on the subsurface distribution of the Seogwipo Formation in that region
-
3.
The lack of detailed water elevation data needed for calibration
Given these issues, the analysis was simplified by excluding the area of the island above the 600 m asl from modeling (Fig. 4). The choice of this 600 m elevation line was somewhat arbitrary; however, generally, no perched groundwater has been observed beyond that contour line (Koh et al. 2006c). Total recharge over this area as computed by Mair et al. (2013b) within each hydrologic district (Fig. 1) was integrated and routed to the modeled area through the boundary of the excluded area where a specified flux condition was applied. Gingerich (2008) adopted a similar approach for an area on the island of Maui (Hawaii) for a similar situation.
Basal groundwater occurs near the coast and floats on the denser seawater, whereas parabasal groundwater is separated from the seawater interface by the low permeability Seogwipo sediments (Fig. 3). Because the Seogwipo Formation occurs at the greatest depths in the eastern section of the island (Koh et al. 2006b), basal groundwater extends farthest inland along the eastern coast (Hagedorn et al. 2011) where typical mixing and water-circulation patterns develop (Kim et al. 2006).
The ratio of head above sea level versus the depth of the freshwater lens below sea level at the midpoint of the transition zone averages 1:23 in the eastern area of Jeju (Won et al. 2006), which is significantly different from the typical 1:40 Ghyben-Herzberg ratio (Freeze and Cherry 1979). Such a deviation can be due to many factors including the effects of anisotropy and vertical flow components (Gingerich 2008). The transition zone is thin in this area and is virtually non-existent in other parts of the island (Won et al. 2006). The relatively minor observed effects of saltwater intrusion and mixing for most of the island justified the use of MODFLOW for an island-wide model and the application of SEAWAT to the eastern sector of Jeju.
Basement groundwater occurs in fractures and joints of the basement rocks at depths of −300–400 m or more. The granitic basement is relatively impermeable with a uniform estimated conductivity of 0.01 m/day (Kim et al. 2009). Basement groundwater does not currently comprise a significant potable water source for Jeju, and no basement specific groundwater data was available for model calibration in this study.
Groundwater discharge mainly takes place as submarine groundwater discharge (SGD) to the ocean, spring flow above sea level, water withdrawal from wells, and evapotranspiration at the water table (Hahn et al. 1997; Koh et al. 2006a; Won et al. 2005). Kim et al. (2003a) assumed that freshwater discharge is approximately equal to recharge and estimated a submarine discharge rate at about 1.5 × 109 m3/year. Also according to Kim et al. (2003a), in the eastern region, almost all groundwater below a few meters of the water table is essentially comprised of re-circulated seawater, whereas in the west, freshwater accounts for 20 % of the SGD. In the north and south, there is less SGD, but rather there are a large number of coastal springs (Koh et al. 2006a). Obviously ignoring density driven and saltwater flow in the MODFLOW model is expected to introduce errors in estimating SGD.
A total of 911 springs have been mapped on Jeju Island (Koh et al. 2006a) and spring flow commonly occurs along streambeds, lava flow boundaries, and scoria cones. Hahn et al. (1997) provide estimates for average discharge from the springs around the coast.
MODFLOW
Spatial discretization and layering
The flow domain is 78-km long, 37-km wide, and up to 800-m deep, rotated counter-clockwise by 17.6° to minimize the number of inactive numerical cells (Fig. 4). Spatially, the model cells are uniform at 500 m by 500 m, resulting in 156 columns and 74 rows. Vertically, the domain was divided into 11 layers: (1) volcanics material assigned as a single layer, (2) the Seogwipo and basement granite each was divided evenly into three layers, and (3) the U-Formation was evenly divided into four layers (Fig. 4). Since the model is only simulating interactions from the water-table downwards, the elevations were truncated at around 325 m asl, which is above the maximum water level reported by Won et al. (2006). The elevations for the top of the layer representing the Seogwipo Formation were based on a co-kriging interpolation of borehole data compiled throughout the island (K. Ha, KIGAM, personal communication, 2010).
Boundary and initial conditions
Bathymetric data are typically used to extend the model boundary beyond the coastline and assign specified head or head-dependent flux conditions. However, in this study, the external model boundary was set at the coastline of the island to reduce the computational burden due to the large area of an island-wide model. A constant head boundary of zero meters was set at the boundary for the basalt layers and a no-flow boundary was assigned for the other layers. The centerline of the saltwater–freshwater mixing zone was set as a no-flow boundary representing the bottom of the model. The location of such a boundary was based on available data on vertical salinity profiles (Kim et al. 2006; Fig. 4).
Because the models were not used to simulate water flow in the unsaturated zone, it was assumed that potential recharge computed from the soil-water balance modeling represents actual recharge at the water table. For the internal model boundary, the areal recharge above the 600-m elevation contour was accumulated and distributed as constant fluxes along that contour.
MODFLOW modeling was completed for steady-state conditions for the baseline and climate scenarios, representing presumed permanent management decisions. Although not required for the steady-state simulations, but to achieve faster convergence, the initial conditions of head for the models were assigned based on average water levels reported by Won et al. (2006). Drought was assessed for a transient condition at the end of 5- and 10-year duration periods, with the baseline-calibrated head-values representing the initial condition. The transient simulations were done for variable time steps starting at about 2 days with an expansion by a factor of 1.05 for the duration of the simulation. All cases were run with constant recharge at the pertinent average daily rate computed for each of the three recharge scenarios.
Recharge conditions
Grid-cell values of recharge for each of the three recharge scenarios as computed by Mair et al. (2013b) were transformed into an average recharge rate within the hydraulic-conductivity calibration zones (Fig. 5). The greatest recharge occurs under a baseline scenario during 1992–2009 (883 or 1,605 × 106 m3/year) followed by the climate scenario (788 or 1,432 × 106 m3/year) and drought scenario (591 or 1,073 × 106 m3/year).
Parameters and calibration
For calibration, which was done through an iterative procedure, volcanic hydraulic conductivity (K) values were constrained in the range of 0.5–2,000 m/day (Won et al. 2006). The calibrated values are generally higher towards the eastern coast (K = 75–1,275 m/day) and are lower in the interior of the island (K = 0.75–75 m/day). This trend relates to the spatial distribution of the volcanic lithologies on the island with more conductive basalts occurring along the western and eastern coastlines and less conductive trachytic rocks prevailing in the inner portions of the island (Won et al. 2005, 2006). Relatively low conductivities also correlate with the locations of tuff cones on the island (Kim et al. 2007). Vertical conductivity for the basalt has been reported to vary between 3 and 28 m/day (Kim et al. 2003a). The Seogwipo Formation was assigned a uniform horizontal conductivity of 1.0 m/day, while the U-Formation was assigned a slightly lower conductivity of 0.1 m/day (Won et al. 2006).
MODFLOW calibration was done through matching the simulated hydraulic heads to contour values taken from Won et al. (2006), and matching average spring flows at the coast, for which each spring should be treated as a point drain. Modeling parameters for a drain include drain-bed elevation and a conductance parameter, which is combined with the groundwater elevation to compute the amount of spring flow. Hence, the drain acts as a head-dependent source that will dry up if the water table declines below a certain level. Calibration was done by the tedious process of adjusting the drain-bed elevation of each drain until the resulting simulated spring flow approximately matched the average value. However, due to the large number of springs, only springs with flows larger than 1,000 m3/day were treated as drains. The springs with flows less than 1,000 m3/day (with total flow of about 10 % of the total) were simulated as constant flux sources, which always withdraws their specified amount of water from the system.
SEAWAT
Spatial discretization and layering
The inland boundary of the eastern sector of Jeju was delineated from MODFLOW’s baseline run by using the model MODPATH (Pollock 1994). Assigning particles at the center of the island, MODPATH delineated streamlines, or groundwater divides, which were used as the sector’s boundary (Fig. 1). The SEAWAT model was divided into 50 × 30 uniform cells, each of the size of about 900 m by 600 m, with a grid that was rotated 45°. The layer thickness ranged between 9 and 37 m near the ocean where saltwater and freshwater interact (Fig. 6). The domain was divided into three layers—basalt, Seogwipo, and U-Formation—considering that there is no significant flow in the basement; these layers were further divided into sub-layers of 5, 3, and 3 sub-layers, respectively.
Boundary and initial conditions
As was the case with MODFLOW, the ocean-side boundary corresponded to the coastline to reduce the computational burden caused by the need for a refined grid for the model and by non-linearity of the problem. Salinity at the shoreline was set as 19 g/L. A specified-head condition was set for the top layer (above the mean sea level), while a head-dependent flux was assigned to cells below the mean sea level. The latter is a general boundary condition that allows accounting for a variable flux, and thereby is more appropriate in the absence of an accurately defined physical boundary. Conductance, a parameter needed at the head dependent flux boundary, was estimated through an iterative trial and error approach in the calibration process. The bottom of the U-Formation was assigned a no-flow boundary, assuming negligible flow in the basement rocks.
As was the case with MODFLOW, it was assumed that potential recharge represents actual recharge at the water table. However, and due to the use of groundwater divides, a no-flow condition was assumed at the upper boundary (i.e., flux condition approach for recharge used at inland MODFLOW model boundary was eliminated).
All scenarios were simulated as transient and the results were assessed after 5 and 10 years. A pseudo steady-state simulation was completed first for the baseline case and used as the initial condition for all scenarios. SEAWAT simulations were done for variable time steps starting at about 2 days with an expansion by a factor of 1.05 for the duration of the simulation. All cases were run with constant recharge at the pertinent average daily rate computed for each of the three recharge scenarios.
Recharge conditions
The SEAWAT model utilized the recharge rates used in the MODFLOW model as described in the preceding. The recharge maps were clipped to cover the eastern sector’s area.
Parameters and calibration
Minimum calibration was done regarding MODFLOW parameters to test whether their values, especially hydraulic conductivity, were independent of the model used. Calibration was mainly related to parameters controlling solute transport and the transient condition—that is, dispersion and storage parameters. The values of these parameters were adjusted for the volcanics within ranges set by Oki (2005). Results were not sensitive to values assigned for other layers considering their relatively low conductivity. Final values used for the longitudinal and transverse dispersivities were 20 and 2 m, respectively, while values for the specific yield and specific storage were 0.1 and 0.001 m−1, respectively.
Results and discussion
Baseline case
MODFLOW
The baseline case, representing the groundwater withdrawal and other related conditions for 2009, was used to calibrate the models and as a basis for assessing the likely effects of various scenarios. Changes in values of various sustainability indicators, i.e., spring flow, head, and salinity, were compared against the respective values for the baseline case.
The calibrated MODFLOW head results are shown in Fig. 7 compared against measured values reported by Won et al. (2006). Reasonable match can be seen, despite the lack of closer match around the 2 m contours in some locations. Such deviations are most likely due to variability of hydraulic conductivity values at a smaller scale than that represented in the simulations. However, it is believed that the model was reasonably calibrated as evident by the reasonable head match and the spatial distribution of hydraulic conductivity values that correlated with various geological features of the island. As shown in Fig. 8, relatively lower values for hydraulic conductivity appear to coincide with the locations of the cinder cones found on Jeju that may represent lower permeability barriers to groundwater flow.
Island wide, the total measured average spring flow is 992,512 m3/day. Among these, there are about 200 springs with flows that are larger than or equal to 1,000 m3/day, with an average flow of 789,075 m3/day. Based on calibration, the model estimated an average flow of 749,621 m3/day, with a calibration error of about −5 %.
SEAWAT
SEAWAT was run first to simulate 2,000 years to obtain a pseudo steady-state solution, which required a relatively large execution time due to non-linearity of the problem. Figure 9 compares the simulated hydraulic-head values against the measured values reported by Won et al. (2006). Overall, the match is reasonable, but the model underestimates the head values at the 2-m contour lines, compared to the MODFLOW results in Fig. 7. Simulations showed sensitivity of the solution regarding numerical scheme concerning advection. Such an assessment is beyond the scope of this study.
Inaccuracies also are evident in the high-elevation area due to the assumption of no-flow conditions along the boundary treated as divides. Certainly, as was the case with the MODFLOW’s simulation, additional calibration could have produced closer results. However, in the absence of a clear basis for parameter adjustments, calibration by refining spatial distribution of parameters is merely a fitting process. Factors contributing to uncertainty of results are related to the limited information regarding (1) the elevation of the top of the Seogwipo Formation, (2) the complex flow paths associated with fractured aquifer settings, (3) detailed lithology information, especially at the coast, (4) groundwater flow in the interior of the island, and (5) the large range of hydraulic conductivities of the volcanic lithologies (Hahn et al. 1997; Won et al. 2006). The discrepancy between the simulated and measured head values in some areas clearly emphasizes the need for more detailed hydrologic and lithologic data.
Figure 10(a) illustrates SEAWAT’s salinity results for the bottom layer of the basalt formation compared against measured maximum values (KIGAM, personal communication, 2010), which are expected to occur in such a layer. Reasonable match can be seen in most areas. Exceptions are evident in the three marked locations, which are consistent with areas of inaccuracies in hydraulic head estimates (Fig. 9). Figure 10b illustrates the simulated vertical profile of salinity, which is typical of a saltwater intrusion in an aquifer bounded at the bottom by a no-flow condition (e.g., Fetter 2001). Saltwater intrudes into the aquifer with distance corresponding to the position of an intrusion toe, with a location that depends on recharge rates, aquifer parameters, and well pumpage (e.g., Izuka and Gingerich 1998), with highest concentrations near the bottom of the formation. Effects of aquifer stratification and the low conductivities of the Seogwipo and U-Formations are evident.
Scenario simulations
In this study, sustainability of groundwater resources was assessed for various recharge scenarios, each under a number of pumping–sustainable yield (P–SY) ratios ranging between 18 and 100 %. The results are compared to those predicted at the baseline pumping (64 % P–SY) and baseline recharge, which will be termed the baseline pumping–recharge (BPR) case. MODFLOW was used to evaluate water levels and spring flows, while SEAWAT was used to assess salinity.
Spring flow results
Table 1 lists the values and the respective percentage changes in the larger spring flows for various scenarios. The percentage changes were estimated relative to the baseline pumping–recharge (BPR) value of 749,621 m3/day. Figure 11 illustrates the percentage change of spring flows versus the P–SY ratio, and shows strong linear relationships, with R 2 exceeding 0.99 for all cases. The results show that, overall, drought has the most negative effect on spring flows, except for the 5-year case, which is exceeded by the climate change case. The slope of the trend-lines indicates that higher P–SY ratios speed up spring-flow reduction due to imposing a steady-state condition in the model. At the BPR, there is an expected reduction in spring flows by up to 19 % under a 10-year drought scenario, whereas at the admittance level, the likely reduction is 28 % under either the climate or the 10-year drought scenario.
Figure 12 shows the minimum number of larger-flow springs that are expected to dry up under various recharge and pumping scenarios. This figure refers to a ‘minimum’ number of springs, because multiple springs located in the same model cell were combined into a single representative drain. Again, the 10-year drought has the most negative effect followed by the 5-year drought scenario, except at higher pumping rates, above the P–SY ratio of 75 %, for which climate change has more negative effects. At the BPR, it is expected that between 8 and 22 % of springs would dry up under various scenarios. The respective range for the admittance level is between 15 and 27 %.
Water-level results
Table 2 lists the values of the mean and respective standard deviation for changes in hydraulic head for various scenarios, while Fig. 13 displays the mean values against the P–SY ratios. At the BPR, and depending on the scenario, the average head is expected to decline between about 1 and 5 m (the dotted line in Fig. 13). For the admittance case, the decline is between about 3 and 7 m. The worst-case scenario would be related to the 5 or 10-year droughts at a P–SY ratio less than about 35 %. Under higher P–SY ratios, the climate-change scenario has the largest negative effect (up to 8-m decline).
An example of spatial patterns of results is shown in Fig. 14, which depicts contour maps of the head-distributions for the climate-change case for a number of P–SY ratios. As should be expected, the decline in water-table elevations increases as well-pumping increases. Various scenarios predicted more pronounced head declines in the western parts of the island, based on the values of decline and the respective size of the area bounded by the shoreline and the 2-m contour, similar to that shown in Fig. 14. Such declines are most likely related to the relatively lower hydraulic conductivity, lower recharge, and higher pumping in the area, when compared to the other regions; thus, this study suggested using the location of the 2-m contour line as a criterion in assessing the sustainability of resources regarding water elevations.
Salinity results
Figure 15 illustrates the expected changes in salinity for various scenarios after 10 years for the case of 100 % P–SY ratio. The largest increases in salinity values occur under a drought condition, followed by the climate and the baseline cases. The respective maximum increases are about 8.7, 3.6, and 1.0 g/L. After a 5-year drought condition, the maximum increase is still significant at about 5 g/L. The largest increases are confined to the salt-intrusion toe area where the head value is near 2 m. Such changes emphasize the need for developing strong monitoring strategies and restrictive management practices in such areas.
Sustainable management of groundwater resources
Results of this study provide the following guidelines towards the sustainable management of groundwater resources on Jeju Island.
-
1.
Future management decisions should take into consideration the fact that long sustained drought and climate change has serious adverse effects on sustainability indicators.
-
2.
Without even drought or climate change, the current estimated P–SY ratio of 64 % can cause long-term strains on groundwater resources.
-
3.
A strong linear relationship exists between values of spring flow and pumping as shown in Fig. 10. Such a representation can provide appropriate total pumping rates based on a certain acceptable total spring flow. Similarly, Fig. 11 can provide suitable total pumping rates based on acceptable number of dried-up springs.
-
4.
New water-use policies should be implemented in the western part of the island where head declines can be excessive under various scenarios of recharge and water use.
-
5.
The location of the 2-m head contour should be adopted as an indicator of sustainability, regarding both its relative location from the shoreline and salinity level near such a line. Collection of lithologic data and monitoring information should be emphasized in the relevant areas for future model refinement. In addition, restrictive management practices should be developed in such areas to avoid further head decline and seawater intrusion.
Summary and conclusions
The models MODFLOW and SEAWAT were used to assess the sustainability of Jeju Island’s groundwater resources. Due to complex challenges in modeling the central part of the island, the analysis was simplified by excluding the area above 600-m ground elevation from modeling. Flow to the model was estimated as precipitation-induced recharge in addition to that routed from the excluded center area of the island. The models were calibrated against hydraulic head data, average spring-flow estimates, and groundwater salinity measurements. Reasonable matches were obtained despite uncertainties in the model data and the absence of detailed characterization of parameter variability. Other factors contributing to the uncertainty are the lack of information concerning the elevation of the top of the low-permeability Seogwipo Formation, the complex migration behavior of groundwater in crystalline-fractured aquifers, data related to spring flows, coastal stratigraphy, vertical migration of perched groundwater in the interior of the island, and the large possible range of hydraulic conductivities of the island’s volcanic lithologies.
MODFLOW was calibrated to match the average spring-flow values by treating larger-flow springs as drains, with a calibration error of about 5 %. Reasonable calibration of hydraulic head was evident by the fact that spatial distribution of hydraulic conductivity values were correlated with various geological features of the island such as areas with lower conductivities correlating to the locations of tuff cones around the island.
SEAWAT was used to address potential salinity effects on sustainability in the eastern part of the island where saltwater intrusion is significant. Hydraulic conductivity estimates based on MODFLOW’s calibration seems reasonable for use in SEAWAT; however, additional data collection is needed for a closer match of both hydraulic head and salinity.
A number of scenarios were developed to test the sustainability of groundwater resources, and the results were evaluated based on comparison to a baseline simulation. For MODFLOW, a range of pumping rates from 18 to 100 % of the sustainable yield was used with each of the three recharge scenarios. The aquifer sustainability was assessed based on the decrease in head and spring flows.
At the baseline pumping, drought has the most pronounced negative effect on spring flows, with a decrease up to 21 % relative to average condition at the BPR. Up to 22 % of large-flow springs dried up under these conditions. The deterioration is worse with the admittance-value rates, with decline of spring flows up to 28 % and a dried-up percentage of 27 %. Climate change has lower impact on spring flows considering that the change was mostly related to recharge redistribution from the baseline recharge case; however, climate change has the most significant influence on head decline, probably due to the same reason by causing significant local effects. At the baseline pumping, the average head declines between about 1 and 5 m, while for the admittance case, the decline is between about 3 and 7 m.
The west side of Jeju Island is the most sensitive to recharge decline, which is most likely related to the relatively lower hydraulic conductivity and higher pumping rates, when compared to the other regions. The location of the 2-m contour can be used as a criterion for assessing the decline in head under changing recharges.
Drought simulations showed the most increase in salinity when compared to other scenarios. Largest increases are mainly confined around the 2-m head-contour line. As with any modeling studies, it should be stressed that the results are subjected to uncertainty due to model and data limitations. In addition, calibration was done using a snapshot of hydraulic head; therefore, and considering the non-uniqueness inherent in modeling, such a calibration should be considered incomplete.
References
Alley WM, Leake SA (2004) The journey from safe yield to sustainability. Ground Water 42(1):12–16
Aquaveo (2011) Groundwater Modeling System 8.0. Aquaveo, Provo, UT
Bakker M, Schaars F, Hughes JD, Langevin CD, Dausman AM (2013) Documentation of the seawater intrusion (SWI2) package for MODFLOW. US Geol Surv Techniques and Methods, book 6, chap. A46, US Geological Survey, Reston, VA, 47 pp
Bredehoeft JD (2002) The water budget myth revisited: why hydrogeologists model. Ground Water 40(4):340–345
Bredehoeft JD (2007) It is the discharge. Ground Water 45(5):523
Dausman AM, Langevin CD (2005) Movement of the saltwater interface in the Surficial Aquifer System in response to hydrologic stresses and water-management practices, Broward County, Florida: USGS Scientific Investigations Report: SIR 2004–5256
Dausman AM, Langevin CD, Sukop MC (2007) Simulation of submarine groundwater discharge salinity and temperature variations: implications for remote detection. In: Sanford W, Langevin CD, Polemio M, and Povinec P (eds) A new focus on groundwater–seawater interactions. IAHS Publ. 312, IAHS, Oxfordshire, UK, pp 272–280
Devlin JF, Sophocleous M (2005) The persistence of the water budget myth and its relationship to sustainability. Hydrogeol J 13:549–554
Fetter CW (2001) Applied Hydrogeology, 4th ed., Prentice Hall, Englewood Cliffs, NJ
Freeze RA, Cherry, JA (1979) Groundwater. Prentice Hall, Englewood Cliffs, NJ
Gingerich SB (2008) Ground-water availability in the Wailuku area, Maui, Hawai’i. US Geol Surv Sci Invest Rep 2008–5236
Hagedorn B, El-Kadi AI, Mair A, Whittier RB, Ha K (2011) Estimating recharge in fractured aquifers of a temperate humid to semiarid volcanic island (Jeju, Korea) from water table fluctuations, and Cl, CFC-12 and 3H chemistry. J Hydrol 409(3–4):650–662
Hahn J, Lee Y, Kim N, Hahn C, Lee S (1997) The groundwater resources and sustainable yield of Cheju volcanic island, Korea. Environ Geol 33(1):43–53
Hamm S-Y, Cheong J-Y, Jang S, Jung C-Y, Kim B-S (2005) Relationship between transmissivity and specific capacity in the volcanic aquifers of Jeju Island, Korea. J Hydrol 310:111–121
Harbaugh AW, Banta ER, Hill MC, McDonald MG (2000) MODFLOW-2000, the U.S. Geological Survey modular ground-water model: user guide to modularization concepts and the ground-water flow process. US Geol Surv Open-File Rep 00-92
Hawaii State Water Commission (2013) Chapter 174C, State Water Code. http://www.state.hi.us/dlnr/cwrm/regulations/Code174C.pdf. Accessed 9 Dec 2013
Heywood CE, Pope JP (2009) Simulation of ground-water flow in the Coastal Plain aquifer system of Virginia. US Geol Surv Sci Invest Rep 2009-5039, 115 pp
Hughes JD, Langevin CD, Brakefield-Goswami L (2009) Effect of hypersaline cooling canals on aquifer salinization. Hydrogeol J 18(1):25–38
Izuka SK, Gingerich SB (1998) Estimation of the depth to the fresh-water/salt-water interface from vertical head gradients in wells in coastal and island aquifers. Hydrogeol J 6:365–373
Kendy E (2003) The false promise of sustainable pumping rates. Ground Water 41(1):2–4
Kim G, Lee K-K, Park K-S, Hwang D-W, Yang H-S (2003a) Large submarine groundwater discharge (SGD) from a volcanic island. Geophys Res Lett 30:2098. doi:10.1029/2003GL018378, 21
Kim Y, Lee KS, Koh DC, Lee DH, Lee SG, Park WB, Koh GW, Woo NC (2003b) Hydrogeochemical and isotopic evidence of groundwater salinization in a coastal aquifer: a case study in Jeju volcanic island, Korea. J Hydrol 270:282–294
Kim K-Y, Seong H, Kim T, Park KH, Woo N-C, Park Y-S, Koh G-W, Park W-B (2006) Tidal effects on variations of fresh-saltwater interface and groundwater flow in a multilayered coastal aquifer on a volcanic island (Jeju Island, Korea). J Hydrol 330:525–542
Kim T, Moon D-C, Park W-B, Park K-H, Koh G-W (2007) Classification of springs of Jeju Island using cluster analysis of annual fluctuations in discharge variables: investigation of the regional groundwater system. Geosci J 11(4):397–413
Kim N-W, Chung I-M, Yoo S, Lee J, Yang S-K (2009) Integrated surface–groundwater analysis in Jeju Island (in Korean). J Environ Sci 18(9):1017–1026
Koh D-C, Plummer LN, Solomon DK, Busenberg E, Kim Y-J, Chang H-W (2006a) Application of environmental tracers to mixing, evolution, and nitrate contamination of ground water in Jeju Island, Korea. J Hydrol 327:258–275
Koh G-W, Kang B-R, Moon D-C (2006b) Hydrogeological features and groundwater management systems of Jeju Island. Proceedings of the Jeju Hawaii Water Forum, Jeju, Korea, July 2006, pp 325–362
Koh G-W, Park Y-S, Kim G-P (2006c) Fluctuation factors and changes of groundwater levels on Jeju Island. Proceedings of the Jeju Hawaii Water Forum, Jeju, Korea, July 2006, pp 378–409
Koh D-C, Ko K-S, Kim Y, Lee S-G, Chang H-W (2007) Effect of agricultural land use on the chemistry of groundwater from basaltic aquifers, Jeju Island, South Korea. Hydrogeol J, 15:727–743
KOWACO (2003a) Comprehensive investigation of hydrogeology and groundwater resources in Jeju Island (in Korean). Korea Water Resources Corporation, Jeju Province, Korea
KOWACO (2003b) Meteorology, hydrology, and water budget analysis. Technical report, Korea Water Resources Corporation, Jeju Province, Korea
Langevin CD (2001) Simulation of ground-water discharge to Biscayne Bay, southeastern Florida. US Geol Surv Water Resour Invest Rep 00-4251, 127 pp
Langevin CD (2003) Simulation of submarine ground water discharge to a marine estuary, Biscayne Bay, Florida. Ground Water J 41(6):758–771
Langevin CD, Thorne DT Jr., Dausman AM, Sukop MC, Guo W (2007) SEAWAT version 4: a computer program for simulation of multi-species solute and heat transport. US Geol Surv Techniques and Methods, book 6, chap. A22, US Geological Survey, Reston, VA
Liu CCK (2007) RAM2 modeling and the determination of sustainable yield of Hawaii basal aquifers. Project Report PR-2008-06, Water Resources Research Center, University of Hawaii, Honolulu, HI
Liu C-W, Lin C-N, Jang C-S, Chen C-P, Chang J-F, Fan C-C, Lou K-H (2006) Sustainable groundwater management in Kinmen Island. Hydrol Process 20:4363–4372
Maimone M (2004) Defining and managing sustainable yield. Ground Water 42(6):809–814
Mair A, El-Kadi AI, Ha K, Koh G-W (2013a) Temporal and spatial variability of rainfall and climate trend on Jeju Island. Geosci J 17(1):75–85
Mair A, Hagedorn B, Tillery S, El-Kadi AI, Westenbroek S, Ha K, Koh G-W (2013b) Temporal and spatial variability of groundwater recharge on Jeju Island, Korea. J Hydrol 501:213–226
Masterson JP (2004) Simulated interaction between freshwater and saltwater and effects of ground-water pumping and sea-level change, Lower Cape Cod aquifer system, Massachusetts. US Geol Surv Sci Invest Rep 2004-5014, 78 pp
Masterson JP, Garabedian SP (2007) Effects of sea-level rise on ground-water flow in a coastal aquifer system. Ground Water 45(2):209–217
Min S-K, Park E-H, Kwon W-T (2004) Future projections of east Asian climate change from multi-AOGCM ensembles of IPCC SRES scenario simulations. J of the Meteorological Society of Japan, 82(4):1187–1211
Mink JF (1980) State of the groundwater resources of southern Oahu, Board of Water Supply, City and County of Honolulu, Honolulu, HI
Mink JF (1981) Determination of sustainable yields in basal aquifer. In: Groundwater in Hawaii: a century of progress. Water Resources Research Center, University of Hawaii at Manoa, Honolulu, HI, pp 101–116
Mink JF, Lau LS (1992) Aquifer identification and classification for Kauai: groundwater protection strategy for Hawaii. Technical report no. 186, Water Resources Research Center, University of Hawaii at Manoa, Honolulu, HI
Mink JF, Lau LS (1990) Aquifer identification and classification for Oahu: groundwater protection strategy for Hawaii. Technical report no. 179, Water Resources Research Center, University of Hawaii at Manoa, Honolulu, HI
Minnesota Department of Natural Resources (2010) Evaluation of models and tools for assessing groundwater availability and sustainability, priorities for investment. Division of Ecological and Water Resources, St. Paul, MN
Nam M, Kim H, Song Y, Lee T, Suh J (2009) Three-dimensional topographic and bathymetric effects on magnetotelluric responses in Jeju Island. Geophys J Int 176:457–466
Oki DS (2005) Numerical simulation of the effects of low-permeability valley-fill barriers and the redistribution of ground-water withdrawals in the Pearl Harbor area, Oahu, Hawaii. US Geol Surv Sci Invest Rep 2005-5253, 111 pp
Park S (2010) Integration of satellite-measured LST data into cokriging for temperature estimation on tropical and temperate islands. Int J Climatol 31:1653–1664
Pollock DW (1994) User’s Guide for MODPATH/MODPATH-PLOT, version 3: a particle tracking post-processing package for MODFLOW, the U.S. Geological Survey finite difference ground-water flow model. US Geol Surv Open-File Rep 94-464
Sanford WE, Pope JP, Nelms DL (2009) Simulation of groundwater-level and salinity changes in the Eastern Shore, Virginia. US Geol Surv Sci Invest Rep 2009-5066, 126 pp
Shoemaker WB, Edwards KM (2003) Potential for saltwater intrusion into the lower Tamiami aquifer near Bonita Springs, southwestern Florida. US Geol Surv Sci Invest Rep 03-4262, 74 pp
Sohn Y (1996) Hydrovolcanic processes forming basaltic tuff rings and cones on Cheju Island, Korea. Geol Soc Am Bull 108:1199–1211
Sophocleous M (2000) From safe yield to sustainable development of water resources: the Kansas experience. J Hydrol 235:27–43
Takasaki KJ, Mink JF (1985) Evaluation of major dike-impounded ground-water reservoirs, Island of Oahu. US Geol Surv Water Suppl Pap 2217
Westenbroek SM, Kelson VA, Dripps WR, Hunt RJ, Bradbury KR (2010) SWB: a modified Thornthwaithe-Mather soil-water-balance code for estimating groundwater recharge. US Geol Surv Techniques and Methods G-A31, book 6, chap. 31. US Geoloogical Survey, Reston, VA, 60 pp
Whittier R, Rotzoll K, Dhal S, El-Kadi AI, Ray C, Chang D (2010) Groundwater source assessment program for the state of Hawaii, USA: methodology example application. Hydrogeol J 18:711–723
Wilson Okamoto Corporation (2008) Hawaii water plan, water resource protection plan. Dept. of Land and Natural Resources, Commission on Water Resource Management, Honolulu
Won J-H, Kim J-W, Koh G-W, Lee J-Y (2005) Evaluation of hydrogeological characteristics in Jeju Island, Korea. Geosci J 9(1):33–46
Won J-H, Lee J-Y, Kim J-W, Koh G-W (2006) Groundwater occurrence on Jeju Island, Korea. Hydrogeol J 14:532–547
Zheng C (2006) MT3DMS v5.2 supplemental user’s guide: technical report to the U.S. Army Engineer Research and Development Center. Dept. of Geological Sciences, University of Alabama, Birmingham, AB
Zheng C, Wang PP (1999) MT3DMS: a modular three-dimensional multispecies transport model for simulation of advection, dispersion and chemical reactions of contaminants in ground-water systems. Documentation and user’s guide. US Army Corps of Eng., Vicksburg, MI
Acknowledgements
This study was funded by a grant from the Korea Institute for Geoscience and Mineral Resources (KIGAM). This is contributed paper CP-2014-02 of the Water Resources Research Center, University of Hawaii at Manoa, Honolulu.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
El-Kadi, A.I., Tillery, S., Whittier, R.B. et al. Assessing sustainability of groundwater resources on Jeju Island, South Korea, under climate change, drought, and increased usage. Hydrogeol J 22, 625–642 (2014). https://doi.org/10.1007/s10040-013-1084-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10040-013-1084-y