Machine learning predictions of trophic status indicators and plankton dynamic in coastal lagoons

Béjaoui, Béchir; Ottaviani, Ennio; Barelli, Enrico; Ziadi, Boutheina; Dhib, Amel; Lavoie, Michel; Coidessa, Gianluca; Turki, Souâd; Solidoro, Cosimo; Aleya, Lotfi

doi:10.1016/j.ecolind.2018.08.041

Cited by 31 publications

(14 citation statements)

References 68 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…and similar to the one reported by Béjaoui et al (2018) for Ghar el Melh lagoon (R 2 =0.64). Both lagoons are located in the Mediterranean coast of north Tunisia.…”

Section: Figsupporting

confidence: 90%

See 1 more Smart Citation

Machine Learning modeling techniques for forecasting the trophic level in a restored South Mediterranean lagoon using Chlorophyll-α

Hadid

Goyet

Chaar

et al. 2021

Preprint

View full text Add to dashboard Cite

An Artificial Neural Network (ANN), a Machine Learning (ML) modeling approach is proposed to predict the ecological state of the North Lagoon of Tunis, a shallow restored Mediterranean coastal ecosystem. A Nonlinear Auto Regressive with exogenous input (NARX) neural network model was fitted to predict Chlorophyll- a (Chl- a ) concentrations in the North Lagoon of Tunis as an eutrophication indicator. The modeling is based on approximately three decades of monitoring water quality data (from January 1989 to April 2018) to train, validate and test the NARX model. The most relevant predictor variables used in this model were those having a high permutation importance ranking with Random Forest (RF) model, which simplified the structure of the network, resulting in a more accurate and efficient procedure. Those predictor variables are secchi depth, and dissolved oxygen. Various model scenarios with different NARX architectures were tested for Chl- a prediction. To verify the model performances, the trained models were applied to field monitoring data. Results indicated that the developed NARX model can predict Chl- a concentrations in the North Lagoon of Tunis with high accuracy (R= 0.79; MSE= 0.31). In addition, results showed that the NARX model generally performed better than multivariate linear regression (MVLR). This approach could provide a quick assessment of Chl- a variations for lagoon management and eco-restoration.

show abstract

“…and similar to the one reported by Béjaoui et al (2018) for Ghar el Melh lagoon (R 2 =0.64). Both lagoons are located in the Mediterranean coast of north Tunisia.…”

Section: Figsupporting

confidence: 90%

“…In another more recent research study, Béjaoui et al (2018) have used the RF model to study the dynamic of the plankton in Ghar Melh lagoon, located in the north of the Tunisian Mediterranean coast.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning modeling techniques for forecasting the trophic level in a restored South Mediterranean lagoon using Chlorophyll-α

Hadid

Goyet

Chaar

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Machine learning models also have the ability to be further refined and trained with more input data, presumably improving their predictive potential. These capabilities have been demonstrated for predicting a range of biogeochemical parameters in various coastal and ocean environments, including chlorophyll concentrations in coastal lagoons [19,20], hypoxic risk using geomorphological and bathymetric features in the coastal regions of the Baltic Sea [21], and nutrients and carbonate system variables in the Mediterranean Sea [18], among others ( [22][23][24][25] and the references therein). In many cases, even with the most advanced numerical models, accurately modeling these biogeochemical parameters is challenging and computationally expensive and, therefore, not as potentially useful as data-driven models and machine learning techniques for near real-time estimates.…”

Section: Introductionmentioning

confidence: 99%

“…The problem of predicting a possible hypoxic event from data can be solved while using two main approaches. The first approach would involve transforming the response variable to labeled categories by empirically estimating the thresholds of low, medium, or high hypoxia risk [19,21]. This approach simplifies the machine learning model to a classification problem, at the cost of introducing uncertainty to the prediction in the form of the empirically decided thresholds for each label.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning Based Predictions of Dissolved Oxygen in a Small Coastal Embayment

Valera

Walter

Bailey

et al. 2020

JMSE

View full text Add to dashboard Cite

Coastal dissolved oxygen (DO) concentrations have a profound impact on nearshore ecosystems and, in recent years, there has been an increased prevalance of low DO hypoxic events that negatively impact nearshore organisms. Even with advanced numerical models, accurate prediction of coastal DO variability is challenging and computationally expensive. Here, we apply machine learning techniques in order to reconstruct and predict nearshore DO concentrations in a small coastal embayment while using a comprehensive set of nearshore and offshore measurements and easily measured input (training) parameters. We show that both random forest regression (RFR) and support vector regression (SVR) models accurately reproduce both the offshore DO and nearshore DO with extremely high accuracy. In general, RFR consistently peformed slightly better than SVR, the latter of which was more difficult to tune and took longer to train. Although each of the nearshore datasets were able to accurately predict DO values using training data from the same site, the model only had moderate success when using training data from one site to predict DO at another site, which was likely due to the the complexities in the underlying dynamics across the sites. We also show that high accuracy can be achieved with relatively little training data, highlighting a potential application for correcting time series with missing DO data due to quality control or sensor issues. This work establishes the ability of machine learning models to accurately reproduce DO concentrations in both offshore and nearshore coastal waters, with important implications for the ability to detect and indirectly measure coastal hypoxic events in near real-time. Future work should explore the ability of machine learning models in order to accurately forecast hypoxic events.

show abstract

“…Machine learning methods enable the development and implementation of such algorithms. Machine learning techniques have already been successfully used in multiple environments to detect fish species automatically from imagery collected with underwater cameras [12] and to predict trophic status indicators in coastal lagoons [13]. Autonomous Systems for the Environmental Characterization of Lagoons DOI: http://dx.doi.org/10.5772/intechopen.90405…”

Section: Introductionmentioning

confidence: 99%

Autonomous Systems for the Environmental Characterization of Lagoons

Casado¹,

Palma²,

Leinster³

2020

Lagoon Environments Around the World - A Scientific Perspective

View full text Add to dashboard Cite

This chapter reviews the state of the art in robotics and autonomous systems (RAS) for monitoring the environmental characteristics of lagoons, as well as potential future uses of such technologies that could contribute to enhancing current monitoring programmes. Particular emphasis will be given to unmanned aerial vehicles (UAVs), autonomous under water vehicles (AUVs), remotely operated underwater vehicles (ROVs) and (semi-)autonomous boats. Recent technological advances in UAVs, AUVs and ROVs have demonstrated that high-resolution data (e.g. 0.4 cm imagery resolution) can be gathered when bespoke sensors are incorporated within these platforms. This in turn enables the accurate quantification of key metrics within lagoon environments, such as coral morphometries. For example, coral height and width can now be estimated remotely with errors below 12.6 and 14.7 cm, respectively. The chapter will explore how the use of such technologies in combination could improve the understanding of lagoon environments through increased knowledge of the spatial and temporal variations of parameters of interest. Within this context, both advantages and limitations of the proposed approaches will be highlighted and described from operational, logistical, and regulatory considerations. The chapter will be based on recent peer-reviewed research outputs obtained by the authors.

show abstract

Machine learning predictions of trophic status indicators and plankton dynamic in coastal lagoons

Cited by 31 publications

References 68 publications

Machine Learning modeling techniques for forecasting the trophic level in a restored South Mediterranean lagoon using Chlorophyll-α

Machine Learning modeling techniques for forecasting the trophic level in a restored South Mediterranean lagoon using Chlorophyll-α

Machine Learning Based Predictions of Dissolved Oxygen in a Small Coastal Embayment

Autonomous Systems for the Environmental Characterization of Lagoons

Contact Info

Product

Resources

About