Numerical simulation of the effects of groundwater withdrawal and injection of high-salinity water on salinity and groundwater discharge, Kaloko-Honokōhau National Historical Park, Hawaiʻi

Kaloko-Honokōhau National Historical Park (KAHO) is located on the west coast of the island of Hawaiʻi and contains water resources exposed in fishponds, anchialine pools, and marine waters that are cultural resources and that provide habitat for threatened, endangered, and other culturally important native species. KAHO’s water resources are sustained by and dependent on groundwater discharge. In 1978, the year of KAHO authorization, the lands immediately surrounding KAHO were undeveloped and zoned for conservation purposes; at present, most surrounding lands are either developed or zoned for industrial, commercial, or residential use. Urbanization of the North Kona District has increased the need for additional drinking and nonpotable (irrigation) water. Because KAHO’s water resources may be affected by existing and proposed groundwater withdrawals and injections of high-salinity water in the surrounding area, the U.S. Geological Survey, in cooperation with the National Park Service, undertook this study to refine the understanding of how groundwater withdrawals and injection of high-salinity water may affect KAHO’s water resources.

Changes in KAHO water resources, in terms of changes in salinity and groundwater discharge, were modeled for selected scenarios of groundwater withdrawal and high-salinity water injection in the aquifer. The numerical model was developed using the model code SUTRA, which accounts for density-dependent flow and salinity transport, and included the coastal-confined groundwater system beneath the coastal freshwater-lens system. Model results indicate that withdrawal of additional groundwater from the coastal freshwater-lens system will affect the salinity of KAHO’s anchialine pools, which provide habitat for the endangered orangeblack Hawaiian damselfly (Megalagrion xanthomelas). The magnitude of the effect is dependent on the amount and location of the withdrawal. Greater withdrawal rates cause greater increases in salinity in KAHO, other factors being equal. For a given withdrawal rate, the greatest increase in salinity in KAHO is associated with wells withdrawing groundwater in an area inland of KAHO and the least increase in salinity is associated with wells near the coast. Model results also indicate that withdrawal of additional groundwater from the coastal freshwater-lens system will affect the groundwater discharge, in terms of the freshwater component (water with zero salinity) of the discharge, through KAHO. Greater withdrawal rates cause greater reductions in freshwater discharge through KAHO. For a given withdrawal rate, the greatest reduction in freshwater discharge through KAHO is associated with wells near the north boundary of KAHO and the least reduction is associated with wells near the coast to the north and south of KAHO.

Injection of high-salinity water that is denser than ocean water can affect the salinity of damselfly habitat in KAHO, with the magnitude of the effect dependent on the location of the injection. Model results indicate that salinity may either increase or decrease in the anchialine pools that provide damselfly habitat in KAHO, depending on the site of injection. Injection inland of KAHO and at sites immediately north and south of KAHO causes a simulated decrease in salinity at the damselfly habitat, whereas injection farther north and south of KAHO causes an increase in salinity. Injection of high-salinity water also causes a reduction in freshwater discharge through KAHO, with the greatest reduction associated with distant injection wells to the north and south of KAHO and the least reduction associated with wells located near and immediately inland from KAHO.

The numerical groundwater models developed for this study have a number of limitations. Lack of understanding of the subsurface geology constrains the ability to accurately model the groundwater-flow system. The models developed for this study are nonunique, cannot account for local-scale heterogeneities in the aquifer, and contain uncertainties related to recharge, boundary conditions, assigned parameter values in the model, and representations of the different hydrogeological features. Confidence in model results can be improved by addressing these and other limitations. In spite of these limitations, the three-dimensional numerical model developed for this study provides a useful conceptual understanding of the potential effects of additional withdrawals and injections on groundwater resources in KAHO. Further evaluation of the ecologic effects of the simulated changes in groundwater quality and quantity in KAHO is needed but is beyond the scope of this study.