Modelling geosmin concentrations in three sources of raw water in Quebec, Canada

Abstract

The presence of off-flavour compounds such as geosmin, often found in raw water, significantly reduces the organoleptic quality of distributed water and diverts the consumer from its use. To adapt water treatment processes to eliminate these compounds, it is necessary to be able to identify them quickly. Routine analysis could be considered a solution, but it is expensive and delays associated with obtaining the results of analysis are often important, thereby constituting a serious disadvantage. The development of decision-making tools such as predictive models seems to be an economic and feasible solution to counterbalance the limitations of analytical methods. Among these tools, multi-linear regression and principal component regression are easy to implement. However, due to certain disadvantages inherent in these methods (multicollinearity or non-linearity of the processes), the use of emergent models involving artificial neurons networks such as multi-layer perceptron could prove to be an interesting alternative. In a previous paper (Parinet et al., Water Res 44: 5847-5856, 2010), the possible parameters that affect the variability of taste and odour compounds were investigated using principal component analysis. In the present study, we expand the research by comparing the performance of three tools using different modelling scenarios (multi-linear regression, principal component regression and multi-layer perceptron) to model geosmin in drinking water sources using 38 microbiological and physicochemical parameters. Three very different sources of water, in terms of quality, were selected for the study. These sources supply drinking water to the Québec City area (Canada) and its vicinity, and were monitored three times per month over a 1-year period. Seven different modelling methods were tested for predicting geosmin in these sources. The comparison of the seven different models showed that simple models based on multi-linear regression provide sufficient predictive capacity with performance levels comparable to those obtained with artificial neural networks. The multi-linear regression model (R(2) = 0.657, <0.001) used only four variables (phaeophytin, sum of green algae, chlorophyll-a and potential Redox) in comparison with ten variables (potassium, heterotrophic bacteria, organic nitrogen, total nitrogen, phaeophytin, total organic carbon, sum of green algae, potential Redox, UV absorbance at 254 nm and atypical bacteria) for the best model obtained with artificial neural networks (R(2) = 0.843).

Biochemistry and genetics of taste- and odor-producing cyanobacteria

“…Modeling has also been used to predict T&O and relate outbreaks to environmental parameters and in some cases, cyanobacteria (Smith et al, 2002;Davies et al, 2004;Dzialowski et al, 2009;Graham et al, 2008;Parinet et al, 2013;Su et al, 2015;). Yet despite the substantial costs that cyanobacterialderived and other T&O events impart on drinking water and other industries, there are few guidelines or standards for acceptable odor levels in drinking water and this issue is most often managed in a reactive mode.…”

Trace analysis of off-flavor/odor compounds in water using liquid-liquid microextraction coupled with gas chromatography—positive chemical ionization-tandem mass spectrometry

“…Extraction of a single sample was completed within five minutes; as opposed to requiring more than 20 min for SPME procedures [17,18]. Furthermore, sample throughput using SPME methods is limited by the number of SPME devices available.…”

Evaluating the impact of percolated reclaimed water from river-channel reservoir on groundwater using tracers in Beijing, Northern China

“…The physical and chemical parameters were standardized using the mean and standard deviation (z-scores) to avoid the effect of dimensions of values prior to the multivariate analysis (Lanzante 1996). Kaiser-Meyer-Olkin (KMO) index was proposed for a measure of sampling adequacy (Cerny and Kaiser 1977) and Bartlett test of sphericity (Jackson 1991) were conducted to ensure the relevance of the results and usefulness of the FA (Parinet et al 2013). After data screening, they met the conditions required in FA with KMO index of 0.71 and the Bartlett's test significance probability of 0.003 in this study, indicating that FA could be considered appropriate and useful tool to provide significant reduction in the dimensionality of the data.…”