Groundwater quality assessment using data clustering based on hybrid Bayesian networks

Aguilera, Pedro A.; Fernández, Antonio; Ropero, Rosa F.; Molina, Luís

doi:10.1007/s00477-012-0676-8

Cited by 66 publications

(33 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…BN models based on the methodology proposed here have already been successfully applied in environmental modelling (Aguilera et al, 2010(Aguilera et al, , 2013, as well as in other fields (Fern andez et al, 2007;Cobb et al, 2013).…”

Section: Model Descriptionmentioning

confidence: 95%

Modelling uncertainty in social–natural interactions

Ropero

Rumí

Aguilera

2016

Environmental Modelling & Software

View full text Add to dashboard Cite

Section: Model Descriptionmentioning

confidence: 95%

Modelling uncertainty in social–natural interactions

Ropero

Rumí

Aguilera

2016

Environmental Modelling & Software

View full text Add to dashboard Cite

“…BNs are being increasingly used to model ecological systems (Borsuk et al 2003;McCann et al 2006;Ticehurst et al 2007;Allan et al 2012) as well as being used to assist decision making within water resource management (Castelletti and Soncini-Sessa 2007;Molina et al 2010;Aguilera et al 2011). It has been proposed that BNs can be used for surface water quality assessment and prediction (Reckhow 1999) and there are some emerging applications for groundwater quality studies (Aguilera et al 2013). Most applications are directed at eutrophication processes and biological water quality (e.g.…”

Section: Water Quality Predictionsmentioning

confidence: 99%

The effects of climate change on ecologically-relevant flow regime and water quality attributes

Dyer

Elsawah

Croke

et al. 2013

Stoch Environ Res Risk Assess

View full text Add to dashboard Cite

The management of freshwater ecosystems is usually targeted through the regulation of water quantity (limiting diversions and providing environmental flows) and regulation of water quality (setting limits or targets for constituent concentrations). Climate change is likely to affect water quantity and quality in multiple ways and the future management of freshwater ecosystems requires predictions of plausible future conditions. We use a suite of ecologically-relevant hydrological indicators to determine the significance of projected climate-driven hydrological changes in the Upper Murrumbidgee River Catchment in south eastern Australia in relation to river regulation. We also determine the possible water quality changes (in relation to guidelines for aquatic ecosystem protection) associated with the climate change projections to identify the combined effects of hydrological and water quality changes. The results of this study suggest that river regulation has resulted in greater changes to ecologically-relevant streamflow characteristics than climate change scenarios that involve a 1 and 2°C temperature rise in the Upper Murrumbidgee River Catchment. In contrast to the projected hydrological changes, Bayesian Network modelling suggests very small changes to violations of water quality thresholds designed to protect aquatic ecosystems as a result of climate change. By identifying key components of the flow and water quality regimes that may be affected by climate change, we are able to provide managers with information that assists in developing adaptation initiatives.

show abstract

“…A classification problem in which no information about the class variable is available (called an unsupervised classification or clustering problem) can be solved by a BN classifier (Aguilera et al, 2013;Anderberg, 1973;Fernández et al, 2014;Gieder et al, 2014). This soft-clustering methodology implies the partition of the data into groups in such a way that the observations belonging to one group are similar to each other but differ from the observations in the other groups.…”

Section: Introductionmentioning

confidence: 99%