Presenter/Author Information

A. Castelletti
F. Pianosi
L. Alberti
G. Oberto

Keywords

response surface, groundwater, optimization, salt intrusion

Start Date

1-7-2012 12:00 AM

Abstract

Infiltration galleries and scavenger wells are usually constructed to controlpumping based saltwater intrusion in coastal aquifers. The optimal allocation of theseinfrastructures can be decided by solving a multi-objective optimization problem balancingavailability of fresh water supply and installation/operation costs, where the effects ofdifferent design options on the planning objectives are simulated through a high fidelitymodel of the flow and transport processes. The incorporation of these simulation modelswithin an optimization-based planning framework is not always straightforward becauseof the computational requirements of the model itself and the computational limitations ofthe optimization algorithms. In this paper we explore the potential for the Global InteractiveResponse Surface (GIRS) methodology to overcome these technical limitations. TheGIRS methodology is used to recursively build a non-dynamic emulator of the processbasedmodel that maps the design options into the objectives values and can be usedin place of the original model to more quickly explore the design option space. The approachis used to plan infrastructural interventions for controlling saltwater intrusion andensuring sustainable groundwater supply for Nauru, a Pacific island republic in Micronesia.GIRS is used to emulate a SEAWAT density driven groundwater flow-and-transportsimulation model. The evaluation results show the potential applicability of the proposedapproach for optimal planning of coastal aquifers.

Share

COinS
 
Jul 1st, 12:00 AM

Planning infrastructural measures for controlling saltwater intrusion in a coastal aquifer by Global Interactive Response Surfaces: the Nauru island case study

Infiltration galleries and scavenger wells are usually constructed to controlpumping based saltwater intrusion in coastal aquifers. The optimal allocation of theseinfrastructures can be decided by solving a multi-objective optimization problem balancingavailability of fresh water supply and installation/operation costs, where the effects ofdifferent design options on the planning objectives are simulated through a high fidelitymodel of the flow and transport processes. The incorporation of these simulation modelswithin an optimization-based planning framework is not always straightforward becauseof the computational requirements of the model itself and the computational limitations ofthe optimization algorithms. In this paper we explore the potential for the Global InteractiveResponse Surface (GIRS) methodology to overcome these technical limitations. TheGIRS methodology is used to recursively build a non-dynamic emulator of the processbasedmodel that maps the design options into the objectives values and can be usedin place of the original model to more quickly explore the design option space. The approachis used to plan infrastructural interventions for controlling saltwater intrusion andensuring sustainable groundwater supply for Nauru, a Pacific island republic in Micronesia.GIRS is used to emulate a SEAWAT density driven groundwater flow-and-transportsimulation model. The evaluation results show the potential applicability of the proposedapproach for optimal planning of coastal aquifers.