SPSéverin Pistre

Séverin Pistre

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Séverin Pistre has a total of 67 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Hydrochemical Assessment and Exploratory Machine Learning Analysis of Nitrate Contamination in Groundwater of the Garet Tarf Alluvial Plain, NE Algeria

    by Djoudi Salima, Pistre Séverin

    The pollution of the shallow groundwater with nitrate (NO₃⁻) is considered as one of the problems concerning the water quality of semi-arid agricultural zones of northeastern Algeria. In this study, the physicochemical analysis of 80 groundwater samples was performed in order to characterize the hydrochemistry of the alluvial aquifer of the Garet Tarf plain in the Oum El Bouaghi wilaya, Algeria, and to estimate the degree of NO₃⁻ contamination. Twelve parameters were determined: the major ions HCO₃⁻, SO₄²⁻, Cl⁻, Ca²⁺, Mg²⁺, Na⁺, K⁺, and NO₃⁻; electrical conductivity (EC); pH; total dissolved solids (TDS); and a Water Quality Index (WQI). Nitrate concentrations vary in the range of 0 to 134 mg/L (mean concentration is 32.6 ± 27.7 mg/L); 18.8% of samples exceeded the WHO (2017) standard of 50 mg/L. Based on the results of piper diagrams, four hydrochemical facies were identified, including the dominance of Ca²⁺–SO₄²⁻ (46.3%) and Ca²⁺–HCO₃⁻ (27.5%), probably reflecting the effect of gypsum/anhydrite dissolution and carbonate weathering. The model based on the Random Forests® algorithm was used as an explorative method to determine which co-measured physicochemical parameters are most likely to be correlated with NO₃⁻ variability. The model demonstrated limited out-of-bag (OOB) R² performance of 16.36% (RMSE = 25.21 mg/L; MAD = 19.12 mg/L). One should consider this result with caution taking into account the limited number of data and predictor variables included. Random Forests® analysis revealed HCO₃⁻, Cl⁻, Ca²⁺, and Mg²⁺ as physicochemical parameters most strongly associated with NO₃⁻ variability within the evaluated dataset (OOB R² of the model is 16.36%). HCO₃⁻ is the first predictor in terms of variable importance (normalised value is 100%). It should be noted that this figure reflects the relative importance of HCO₃⁻ normalised to the most influential predictor. It means that HCO₃⁻ explains 100% of the nitrate variability. WQI values (mean is 53.4) indicate that the predominant quality of the water is poor due to geogenic mineralisation and, sometimes, due to nitrate content. These are different aspects of the water quality issue and cannot be mixed.

  • Open Access

    Tanis (2025)

    by Frédéric Payraudeau, Ahmed Nekshara, Aline Banaszak, Cécile Harlaut, Emmanuel Laroze, Frédérique Marchand-Beaulieu, Lucie Antoine, Patrice Le Guilloux, Philippe Bromblet, Raphaële Meffre, et al.

  • Open Access

    Deep Learning at Two Timescales: Dual Neural Networks for Predicting Fast Urban and Slow Karst Floods

    by Julien Yise Peniel Adounkpe, Valentin Wendling, Alain Dezetter, Bruno Arfib, Kalil Traoré, Guillaume Artigue, Octavian Dobricean, Séverin Pistre, Anne Johannet

    Flash floods in urban and karst environments present major modeling challenges due to their complex hydrodynamics, characterized by a rapid urban runoff response and a delayed slower karst groundwater response. This study explores the use of artificial neural networks ANN (multilayer perceptron in particular) to predict flash floods at the downstream of the Las River in Toulon (France). The Las River is fed in a larger proportion by the nearby karst springs and in a smaller proportion by the urban drainage network. In this study, we propose an ensemble modeling strategy to address the system’s double hydrological regime. The initial step was to identify rainfall events in the six-year hydrometeorological database and classify them according to their karst contributions. Two specialized models: urban runoff model (UM) and karst model (KM) were trained solely on each event type (urban runoff and karst events). These models were combined by two methods in an attempt model all events, disregarding of their event type: the first approach was to combine the outputs of the specialized models in an ANN called output combination model (OM), the second approach was to combine the structures of the specialized models and retraining the model parameters called structure combination model (SM). A third more “basic” approach, called bulk model (BM), was to optimize the ANN by selecting the inputs with the best performance improvements. As expected, the specialized models (UM and KM) performed the best on the cross-validation sets and on the test sets on their respective event types but failed to generalize across regimes. The OM was the most robust ensemble strategy across all event types with consistent accuracy on predicting both urban runoff and karst flood events. The BM was better on the karst events while having worst performance on karst events and the SM was the least accurate model. These findings confirm the added value of combining specialized ANNs to model complex hydrological systems. In addition, selecting the right inputs to the models has a bigger impact on the model’s performance than choosing its structure by changing its hyperparameters.

  • Open Access

    A Comparison Between Adam and Levenberg–Marquardt Optimizers for the Prediction of Extremes: Case Study for Flood Prediction with Artificial Neural Networks

    by Julien Yise Peniel Adounkpe, Valentin Wendling, Alain Dezetter, Bruno Arfib, Guillaume Artigue, Séverin Pistre, Anne Johannet

    Artificial neural networks (ANNs) adjust to the underlying behavior in the dataset using a training rule or optimizer. The most popular first-and second-order optimizers, Adam (AD) and Levenberg–Marquardt (LM), were compared with the aim of predicting extreme flash floods of a runoff-dominated hydrological system. A fully connected multilayer perceptron with a shallow structure was used to reduce complexity and limit overfitting. The inputs of the ANN were determined by rainfall–water level cross-correlation analysis. For each optimizer, the hyperparameters of the ANN were selected using a grid search and the cross-validation score on a novel criterion (PERS PEAK) mixing the persistency (PERS) and the quality of flood-peak restitution (PEAK). For an extreme and unseen event used as a test set, LM outperformed AD by 25% on all performance criteria. The peak water level of this event, 66% greater than that of the training set, was predicted by 92% after more training iterations were done by the LM optimizer. This shows that the ANN can predict beyond the ranges of the training set, given the right optimizer. Nevertheless, the LM training time was up to five times longer than that of AD during grid search.


  • Top Cited Papers
    Open Access

    Decline in Iran’s groundwater recharge

    by Roohollah Noori, Mohsen Maghrebi, Søren Jessen, Sayed M. Bateni, Essam Heggy, Saman Javadi, Mojtaba Noury, Severin Pistre, Soroush Abolfathi, Amir AghaKouchak

    Groundwater recharge feeds aquifers supplying fresh-water to a population over 80 million in Iran—a global hotspot for groundwater depletion. Using an extended database comprising abstractions from over one million groundwater wells, springs, and qanats, from 2002 to 2017, here we show a significant decline of around −3.8 mm/yr in the nationwide groundwater recharge. This decline is primarily attributed to unsustainable water and environmental resources management, exacerbated by decadal changes in climatic conditions. However, it is important to note that the former’s contribution outweighs the latter. Our results show the average annual amount of nationwide groundwater recharge (i.e., ~40 mm/yr) is more than the reported average annual runoff in Iran (i.e., ~32 mm/yr), suggesting the surface water is the main contributor to groundwater recharge. Such a decline in groundwater recharge could further exacerbate the already dire aquifer depletion situation in Iran, with devastating consequences for the country’s natural environment and socio-economic development.


  • Coastal karst aquifers and submarine springs: what future for their water resources?

    by Perrine Fleury, Séverin Pistre, Michel Bakalowicz

    This paper is intended to be a reminder of a series of scientific works whose origin is largely attributable to G. de Marsily, without attempting to make an exhaustive overview of submarine springs and coastal karst aquifers. The main results show that coastal karst aquifers with their submarine outlets and the possibility of natural seawater intrusion are particularly fragile and must be studied and managed with appropriate tools. For the coming decades, the evolution of the climate as results from the IPCC scenarios leads us to question the effects of the increase of sea level on this coastal resource, and the effects of frequent and severe droughts, added to an increase in withdrawals from aquifers caused by the coastalisation of the population.


  • Open Access

    Prediction of water level variations in aquifers using recurrent and convolutional neural networks

    by Lea Duran, Rida Mouni, Valentin Duperron, Loïc Maisonnasse, Guillaume Artigue, Séverin Pistre, Anne Johannet

    The use of artificial intelligence models to simulate and forecast groundwater has become more and more popular since the early 1990ies. These models able to forecast groundwater levels are increasingly being used to support decision making for water resources management and climate change impact local assessments. This study presents a comparison of different algorithms to simulate and forecast the piezometry of the Astian sands aquifer in the Occitanie region (France) at sub seasonal to seasonal scales. The aim is to set up a transferable methodology allowing quick implementation on different aquifers, to lay the foundations for a tool for water resources management focusing on droughts.Different models were tested in this study with prediction horizons from a few weeks and up to the hydrological year. Models used include (1) Multi-layer Perceptron (MLP), (2) recurrent models such as Long short-term memory models, Gated Recurrent Unit (GRU), Non-Linear Regressive Neural Networks (NARX), Dual Stage Attention Based Recurrent Networks (DA-RNN); (3) a Temporal Convolutional Network (TCN); (4) a Reservoir Computing Network (Echo State Network). Pre-processing methods were tested to increase performance and attempt to better reproduce higher frequencies, especially the Ensemble Empirical Modal Decomposition (EEMD). Input variables were the daily rainfall data, temperature data and piezometry. To validate the model and assess the performance of the simulations, different metrics were used (Nash Sutcliffe Efficiency, Kling Gupta Efficiency, Coefficient of persistency). The loss function was a combination of RMSE and NSE, the optimizer used was Adam, the training was done on two thirds of the time series and validation and testing on the remaining data.Several models achieved high performances: DA-RNN, the GRU, MLP and NARX (NSE>0.91 and Cp>0.6 at 20 days). Predictions were carried out iteratively with a 20-days window, reinjecting the predicted data and re-training the model (adaptative model) until a prediction of more than 400 days was reached (using observed rainfall and temperature). The stability and robustness of the model is further investigated through additional cross validation, estimation of confidence intervals, and simulations in different hydrogeological contexts, to aim toward a transferable tool for various aquifers.

  • Open Access

    Groundwater floods: linking the hydrodynamic characterisation of the karst aquifer at the Cadarache CEA centre (South of France) to the karst structure using statistical analysis of temperature.

    by Manon Erguy, Sébastien Morilhat, Guillaume Artigue, Julien Trincal, Anne Johannet, Séverin Pistre

    Karst environments can be subject to rising water tables leading to rapid kinetic floods (a few hours) of great amplitude (about tens of meters) due to heavy rainfall. The Cadarache centre of the French Alternative Energies & Atomic Energy Commission (CEA), located in the French Mediterranean region (Bouches-du-Rhône), is subject to this type of underground flooding. For example, in November 2011, a cumulative rainfall of 225 mm in 72 hours was recorded at the centre, leading the Cretaceous limestone water table to a rise of nearly 40m, locally. The Cadarache centre of 1,600 hectares, has an exceptional hydrogeological monitoring network of 8 rain gauges and more than 400 piezometers, including 230 equipped with pressure/temperature sensors (115 measuring the Cretaceous aquifer). This dense monitoring network has enabled the acquisition of a large database (rainfall, water-level and temperature) over more than 15 years.The goal of this study is to improve the understanding of the Cretaceous karst aquifer hydrodynamic behaviour by analysing temperature data. This would allow an improvement of groundwater floods prediction on the site. This variable is little studied but it can provide interesting information on the aquifer hydrodynamic response and specifically the role of karst structures on pressure and mass transfers at kilometer scale.For this purpose, statistical analyses were conducted on measured temperature data in various piezometers defined as representative of the aquifer. Sorted temperature data and correlations analyses were made for different types of rainy events. This study aims to: i. Identify potential flow dynamics (deep or superficial flows), ii. Compare the temperature signal with those of rainfall and water levels.


  • Open Access

    Évaluation de la qualité des eaux de surface à l'aide de techniques statistiques multivariées et de méthodes hydrochimiques : étude de cas de l'oued El Hatab (Centre de la Tunisie).

    by Wafa Hayouni, Najiba Chkir, Severin Pistre, Kamel Zouari

    Multivariate statistical techniques, and hydrochemical tracers of surface water of wadi Hatab watershed, central Tunisia, are used to assess the surface water quality for drinking and irrigation purposes. Then, in this context, it is fundamental to identify the sources of water pollution in this basin which represents a good example of anthropized and exploited basin in semi-arid region.Stream chemistry was monitored in 28 sampling sitesThe projection of major ions on the Piper diagram reveals a variability of water chemical facies. It indicates that water mineralization is primarily governed by sodium, magnesium and calcium for cations and sulphates and chloride for anions. Surface water salinity ranged between 0.364 g/l to 29.173 g/l, and exceeds the WHO standard of salinity for 82% of stations. This variation is explained by domestic and industrial discharges in watershed and by the proximity of a salt-water source draining evaporitic formations. With naturally occurring mineralization, an anthropogenic mineralization is observed which is linked to domestic discharges, untreated industrial activities including paper mill discharges, and poorly controlled agricultural practices.Several water quality parameters were applied for drinking and irrigation purposes. Water quality index (WQI) values indicate that 75% of surface water quality are not suitable for drinking usage. According to Percent Sodium (%Na), Residual Sodium Carbonate (RSC), the electrical conductivity (EC) and Wilcox classification, 85% of surface water samples are suitable for agricultural irrigation activities. Based on Sodium Adsorption Ratio (SAR) and USSL classification, it appears that only 47 % of water samples, mostly locared in the western part of the watershed, are suitable for irrigation use and almost 53 % of these samples are “good” or “permissible”.This study demonstrated the impact of several processes on surface water resource. The dataset revealed that water physiochemical properties are mainly controlled by regional lithological features, but the nature of the discharges that are dumped there significantly changes it. In this respect, this research provides insights for sustainable water management. Generally, this study suggests the need and feasibility of multivariate statistics and hydrochemical techniques to evaluate chemical water quality and to determine the origin of water pollution.