A framework for parameter estimation using sharp-interface seawater intrusion models

Pumping Optimization Under Uncertainty in an Island Freshwater Lens Using a Sharp‐Interface Seawater Intrusion Model

“…As the water resources are under pressure due to climate change, hydrocarbon exploration and intense tourism, a groundwater knowledge acquisition project was launched (a) to assess groundwater resources at the island scale and (b) to provide sustainable groundwater pumping scenarios. On Grande Entrée Island, which is the most vulnerable island to seawater intrusion because of its shallow saline groundwater, the development of a SI model, parameter estimation and linear uncertainty analysis allowed for an assessment of the extent of the freshwater lens (Coulon et al., 2021). In this paper, pumping optimization under uncertainty is conducted to provide decision‐support for the management of municipal groundwater pumping.…”

“…The reader is referred to Coulon et al. (2021) for more details on the study area. Nine municipal wells have been pumping water from the freshwater lens since 2013, with pumping rates and water levels being recorded since mid‐2014.…”

“…SI models are most useful for simulating regional‐scale seawater intrusion, where dispersion can be neglected. In island aquifers, SI models are commonly used to assess the extent of freshwater lenses and establish water budgets (Izuka et al., 2021; Lotti et al., 2021; Shuler & Mariner, 2020), while their fast simulation times also facilitate parameter estimation and quantification of predictive uncertainty (Coulon et al., 2021). Because of their rapid simulation times, SI models are widely used to optimize pumping rates in continental coastal aquifers, which are faced with the threat of lateral seawater intrusion (Kopsiaftis et al., 2019).…”

A novel objective function DYNO for automatic multivariable calibration of 3D lake models

“…Also, there are numerous other model calibration paradigms in general hydrology and water resources (besides the hydrodynamic model calibration) where simultaneous multivariable and multisite calibrations are required. Some examples of such multivariable and multisite calibration problems include watershed model calibration (Franco et al, 2020;Odusanya et al, 2019), seawater intrusion model calibration (Coulon et al, 2021), and water quality model calibration Xia and Shoemaker, 2022b) among others. In these problems, there are usually multiple constituents (e.g., substances) to be calibrated and the observations are usually available at multiple locations.…”

“…This is especially true for multivariable calibration problems, where the multiple variables being calibrated are sensitive to the parameters being calibrated. Other examples of such multivariable calibration problems include watershed model calibration (Franco et al, 2020) and seawater intrusion model calibration (Coulon et al, 2021), among others. These multivariable problems require calibration frameworks that enable the simultaneous calibration of all variables rather than calibrating one variable and then the second.…”