Review of surrogate modeling in water resources

Abstract

[1] Surrogate modeling, also called metamodeling, has evolved and been extensively used over the past decades. A wide variety of methods and tools have been introduced for surrogate modeling aiming to develop and utilize computationally more efficient surrogates of high-fidelity models mostly in optimization frameworks. This paper reviews, analyzes, and categorizes research efforts on surrogate modeling and applications with an emphasis on the research accomplished in the water resources field. The review analyzes 48 references on surrogate modeling arising from water resources and also screens out more than 100 references from the broader research community. Two broad families of surrogates namely response surface surrogates, which are statistical or empirical data-driven models emulating the high-fidelity model responses, and lower-fidelity physically based surrogates, which are simplified models of the original system, are detailed in this paper. Taxonomies on surrogate modeling frameworks, practical details, advances, challenges, and limitations are outlined. Important observations and some guidance for surrogate modeling decisions are provided along with a list of important future research directions that would benefit the common sampling and search (optimization) analyses found in water resources.

Multi-objective parameter optimization of common land model using adaptive surrogate modeling

“…However, optimization based on surrogate models can be a challenging task because the response surface might be very bumpy and has many local optima. Razavi et al (2012) gave a comprehensive review of the surrogate modeling methods and applications in water resources, and discussed the pitfalls of surrogate modeling as well.…”

“…According to statistical learning theory, such a build-prune strategy can extract information from training data and meanwhile avoid the influence of noise (Hastie et al, 2009). Because of its pruning and fitting ability, MARS method can be used as parameter screening method (Gan et al, 2014;Li et al, 2013;Shahsavani et al, 2010), and also surrogate modeling method (Razavi et al, 2012;Song et al, 2012;Zhan et al, 2013).…”

“…Pilát and Neruda (2013) developed a surrogate model selector for multi-objective surrogate-assisted optimization. In hydrology and water resources, Razavi et al (2012) has summarized recent applications, advantages and existing problems. Wang et al (2014) evaluated the influence of initial sampling and adaptive sampling methods for surrogateassisted optimization of a simple hydrological model, the Sacramento Soil Moisture Accounting (SAC-SMA) model.…”

Evolutionary algorithms and other metaheuristics in water resources: Current status, research challenges and future directions

“…This is what is called "metamodel-enabled optimizers" by Razavi et al (2012), and is referred to as SM-based optimisation (SMBO) algorithms (SMBOA) in this paper.…”

“…There are a number of reviews of the use of these techniques in optimisation, including Jin (2005) and Knowles and Nakayama (2008), as well as a dedicated edited volume on "Computational Intelligence in Expensive Optimization Problems" (Tenne and Goh, 2010) (in particular, reviews by Shi and Rasheed (2010) and Santana-Quintero et al (2010) therein). In addition, Razavi et al (2012) presented an extensive review of surrogate modelling in water resources and recent developments and applications to environmental systems are also presented in a special issue on "Emulation techniques for the reduction and sensitivity analysis of complex environmental models" (see Ratto et al, 2012).…”