Optimal Design of Multitype Groundwater Monitoring Networks Using Easily Accessible Tools

Abstract

Monitoring networks are expensive to establish and to maintain. In this paper, we extend an existing data-worth estimation method from the suite of PEST utilities with a global optimization method for optimal sensor placement (called optimal design) in groundwater monitoring networks. Design optimization can include multiple simultaneous sensor locations and multiple sensor types. Both location and sensor type are treated simultaneously as decision variables. Our method combines linear uncertainty quantification and a modified genetic algorithm for discrete multilocation, multitype search. The efficiency of the global optimization is enhanced by an archive of past samples and parallel computing. We demonstrate our methodology for a groundwater monitoring network at the Steinlach experimental site, south-western Germany, which has been established to monitor river-groundwater exchange processes. The target of optimization is the best possible exploration for minimum variance in predicting the mean travel time of the hyporheic exchange. Our results demonstrate that the information gain of monitoring network designs can be explored efficiently and with easily accessible tools prior to taking new field measurements or installing additional measurement points. The proposed methods proved to be efficient and can be applied for model-based optimal design of any type of monitoring network in approximately linear systems. Our key contributions are (1) the use of easy-to-implement tools for an otherwise complex task and (2) yet to consider data-worth interdependencies in simultaneous optimization of multiple sensor locations and sensor types.

Data‐Worth Assessment for a Three‐Dimensional Optimal Design in Nonlinear Groundwater Systems

“…The GA is then used to select multiple rows corresponding to the combinations in the X new matrix (for details about GA see Wöhling et al 2016). Combinations of n rows are selected from the X new matrix.…”

“…Combinations of n rows are selected from the X new matrix. In the next step, a new population of design (size N ) is generated by applying the standard GA selection schemes, that is, selection, mutation, and crossover (Wöhling et al 2016). A p × 1 vector of randomly sampled N rows is then generated, where p is a user-defined population size (p = 50 in this work).…”

“…The GA is then used to select multiple rows corresponding to the combinations in the X new matrix (for details about GA see Wöhling et al 2016). In the present study, we retained 40% of the population for the next generation (i = i + 1), muted 5% of population (which was allowed to increase if the population was too uniform), and allowed a 15% chance of selecting outside of the admissible location (similar to Vilhelmsen andFerre 2017 andWöhling et al 2016). This vector forms the initial population of designs (i = 1).…”

“…The worth of such an addition (subsequently referred to as data worth [DW]) on reducing model prediction uncertainty is then evaluated. Wallis et al (2014) extended the DW-based OD for selecting multiple observations, and Wöhling et al (2016) extended it further by using a genetic algorithm (GA) to incorporate multiple new observations of head and/or hydraulic conductivity to decrease the predictive uncertainty. Wallis et al (2014) extended the DW-based OD for selecting multiple observations, and Wöhling et al (2016) extended it further by using a genetic algorithm (GA) to incorporate multiple new observations of head and/or hydraulic conductivity to decrease the predictive uncertainty.…”

Bayesian Monitoring Design for Streambed Heat Tracing: Numerical Simulation and Sandbox Experiments

“…In the past decade, temperature has been used independently or as an additional information source to improve the modeling of river-aquifer systems (Jiang and Woodbury 2006;Krause et al 2012;Kurtz et al 2014;Huang et al 2016). Thus, we can select the monitoring strategy providing the most informative temperature measurements by conducting experimental design (Man et al 2016;Pham and Tsai 2016;Woehling et al 2016), which is also referred to as data worth (DW) analysis Neuman et al 2012;Xue et al 2014). Thus, we can select the monitoring strategy providing the most informative temperature measurements by conducting experimental design (Man et al 2016;Pham and Tsai 2016;Woehling et al 2016), which is also referred to as data worth (DW) analysis Neuman et al 2012;Xue et al 2014).…”