Marc Leblanc’s research while affiliated with Université d´Avignon et des Pays du Vaucluse and other places

Urban Population growth coupled with human activities are the main drivers inducing land use and land cover changes (LU/LCC), which impact earth’s landscapes dynamics. In the era of global challenges, a comprehensive modelling of past and future prediction of LU/LCC is therefore essential. In this regard, the current study aims to model LU/LCC and predict its changes by 2030, 2050 and 2100. Moreover, determine the main driver of these changes, and assess the impact of urbanization on natural ecosystems. For doing so, a three-decade times series (1992–2020) data have been used, including LU/LC, environmental, societal, and auxiliary data to conduct a statistical analysis and modelling using hybrid Multi-Layer-Perceptron Markov-Chain model (MLP-MC). The analyses results showed that urbanization has significantly increased between 1992 and 2020, mainly in the outskirts of the Casablanca metropolis, where economic dynamic, population growth, and territorial infrastructure takes place. This increase was at the expense of barren and agricultural land. Concerning the future projection, the MLP-MC model showed satisfactory results with an overall accuracy (OA) and Cohen’s Kappa greater than 0.98 Revealing that the model has batter reliability to predict LU/LC maps that are identical to observed ones. Consequently, the Future projection indicates that urban areas will persistently sprawl, especially in satellite cities of the economic capital which adversely affect biodiversity, local climate, water, and public health. These findings remain so promising that can guide decision-making and stakeholders to improve sustainable territorial and urban development.

In order to facilitate the monitoring of groundwater quality in France, the groundwater bodies (GWB) in the Provence-Alpes-Côte d’Azur region have been grouped into 11 homogeneous clusters on the basis of their physico-chemical and bacteriological characteristics. This study aims to test the legitimacy of this grouping by predicting whether water samples belong to a given sampling point, GWB or group of GWBs. To this end, 8673 observations and 18 parameters were extracted from the Size-Eaux database, and this dataset was processed using discriminant analysis and various machine learning algorithms. The results indicate an accuracy of 67% using linear discriminant analysis and 69 to 83% using ML algorithms, while quadratic discriminant analysis underperforms in comparison, yielding a less accurate prediction of 59%. The importance of each parameter in the prediction was assessed using an approach combining recursive feature elimination (RFE) techniques and random forest feature importance (RFFI). Major ions show high spatial range and play the main role in discrimination, while trace elements and bacteriological parameters of high local and/or temporal variability only play a minor role. The disparity of the results according to the characteristics of the GWB groups (geography, altitude, lithology, etc.) is discussed. Validating the grouping of GWBs will enable monitoring and surveillance strategies to be redirected on the basis of fewer, homogeneous hydrogeological units, in order to optimize sustainable management of the resource by the health agencies.

Rainfall-runoff modeling plays a crucial role in achieving efficient water resource management and flood forecast- ing, particularly in the context of increasing intensity and frequency of extreme meteorological events induced by climate change. Therefore, the aim of this research is to assess the accuracy of the Long-Short-Term Memory (LSTM) neural networks and the impact of its architecture in predicting runoff, with a particular focus on capturing extreme hydrological discharges in the Ouergha basin; a Moroccan Mediterranean basin with historical implica- tions in many cases of flooding; using solely daily rainfall and runoff data for training. For this purpose, three LSTM models of different depths were constructed, namely LSTM 1 single-layer, LSTM 2 bi-layer, and LSTM 3 tri-layer, their window size and hyperparameters were first tuned, and on seven years of daily data they were trained, then validated and tested on two separate years to ensure the generalization on unseen data. The performance of the three models was compared using hydrogram-plots, Scatter-plots, Taylor diagrams, and several statistical metrics. The results indicate that the single-layer LSTM 1 outperforms the other models, it consistently achieves higher overall performance on the training, validation, and testing periods with a coefficient of determination R-squared of 0.92, 0.97, and 0.95 respectively; and with Nash-Sutcliffe efficiency metric of 0.91, 0.94 and 0.94 respectively, challenging the conventional beliefs about the direct link between complexity and effectiveness. Furthermore, all the models are capable of capturing the extreme discharges, although, with a moderate underprediction trend for LSTM 1 and 2 as it does not exceed -25% during the test period. For LSTM 3, even if its underestimation is less pronounced, its increased error rate reduces the confidence in its performance. This study highlights the impor- tance of aligning model complexity with data specifications and suggests the necessity of considering unaccounted factors like upstream dam releases to enhance the efficiency in capturing the peaks of extreme events.

Water transfers through a multilayered aquifer system are difficult to characterize. This study explores whether the conceptual model of water mixing at depth can be extrapolated over a hydrosystem extended across several tens of kilometers and including multiple aquifer layers. The processes are investigated using a combination of isotope tracers and piezometric monitoring over 10 years. The goal of this approach is to better understand how water transfer occurs throughout a complex and poorly documented hydrosystem of the Mahafaly Plateau in southwestern Madagascar. The results show a clear smoothing of isotopic variability with depth, associated with a smoothing of the recharge peaks. Isotopic values are strongly variable in the near surface (from -6.8 to -2.5‰ 18O) and stabilize at a critical depth (near 20 m) at around -4.7‰ 18O. These results indicate high vertical flows through the aquifer system, where there is neither obvious dominant recharge via preferential pathways nor lateral mixing. Such a strong smoothing effect on groundwater isotopic variability with depth has been rarely observed so clearly over a large spatial scale. These results provide information on a remote groundwater flow system at a scale pertinent to groundwater resource assessment. The results also indicate that the Neogene aquifers of the Mahafaly Plateau are poorly connected with other water resources (rivers, old sedimentary formations) except for the percolation of water towards the deep Eocene karst. This means that groundwater resources in the Ankazomanga Basin are limited and that it is essential to understand and quantify recharge for sustainable groundwater management.

The planning and management of groundwater in the absence of in situ climate data is a delicate task, particularly in arid regions where this resource is crucial for drinking water supplies and irrigation. Here the motivation is to evaluate the role of remote sensing data and Input feature selection method in the Long Short Term Memory (LSTM) neural network for predicting groundwater levels of five wells located in different hydrogeological contexts across the Oum Er-Rbia Basin (OER) in Morocco: irrigated plain, floodplain and low plateau area. As input descriptive variable, four remote sensing variables were used: the Integrated Multi-satellite Retrievals (IMERGE) Global Precipitation Measurement (GPM) precipitation, Moderate resolution Imaging Spectroradiometer (MODIS) normalized difference vegetation index (NDVI), MODIS land surface temperature (LST), and MODIS evapotranspiration. Three LSTM models were developed, rigorously analyzed and compared. The LSTM-XGB-GS model, was optimized using the GridsearchCV method, and uses a single remote sensing variable identified by the input feature selection method XGBoost. Another optimized LSTM model was also constructed, but uses the four remote sensing variables as input (LSTM-GS). Additionally, a standalone LSTM model was established and also incorporating the four variables as inputs. Scatter plots, violin plots, Taylor diagram and three evaluation indices were used to verify the performance of the three models. The overall result showed that the LSTM-XGB-GS model was the most successful, consistently outperforming both the LSTM-GS model and the standalone LSTM model. Its remarkable accuracy is reflected in high R2 values (0.95 to 0.99 during training, 0.72 to 0.99 during testing) and the lowest RMSE values (0.03 to 0.68 m during training, 0.02 to 0.58 m during testing) and MAE values (0.02 to 0.66 m during training, 0.02 to 0.58 m during testing). The LSTM-XGB-GS model reveals how hydrodynamics, climate, and land-use influence groundwater predictions, emphasizing correlations like irrigated land-temperature link and floodplain-NDVI-evapotranspiration interaction for improved predictions. Finally, this study demonstrates the great support that remote sensing data can provide for groundwater prediction using ANN models in conditions where in situ data are lacking.

Study region Morocco, Northwestern Africa. Study focus Since April 2002, the Gravity Recovery and Climate Experiment (GRACE) mission have opened new pathways for hydrologists to monitor the changes in terrestrial total water storage (TWS). Here, the Center for Space Research (CSR), Goddard Space Flight Center (GSFC), Jet Propulsion Laboratory (JPL), and the average (AVG) GRACE mascon solutions were used to examine the changes in TWS and groundwater storages (GWS) in Morocco, with an emphasis on natural replenishment events and their link to snow cover area (SCA) and rainfall variability. New hydrological insights for the region The results showed that GRACE TWS from AVG (TWSAVG) and GSFC (TWSGSFC) can fairly describe the temporal patterns of the groundwater level (GWL). Moreover, during 2002–2020, the TWS underwent a strong depletion relatively masked by natural recharge events. This was revealed as we identified two intermittent depletion episodes with statistically significant rates (−1.03 ± 0.11 to −0.31 ± 0.1 cm yr⁻¹) higher than those obtained for the long-term trend lines (−0.28 ± 0.11 to −0.15 ± 0.07 cm yr⁻¹). The TWS appeared to be strongly linked with the SCA metrics and rainfall indices with 1–3 months of lag. Our findings suggest that the rainfall distribution can be more insightful about changes in groundwater levels compared to the rainfall monthly totals.

Understanding recharge in the Mekong Delta is critical for the delta's groundwater resources, and requires the investigation of recharge processes at the local scale. In this study of the north eastern area of the Mekong Delta, time-series of environmental tracer data (δ¹⁸O, δ²H, major ions and ³H) and markers of rural pollution (NH4 and NO3) were used to highlight localized recharge and impacts on groundwater quality. Results highlighted new hydrological insights into recharge processes, including that the Pleistocene aquifer receives recent recharge (< 60 years), predominantly during high rainfall months (> 100 mm/month). However, due to shallow clay layers there are significant spatial variations in these recharge processes, which were observed in the seasonal fluctuation of groundwater δ¹⁸O values in groundwater. Wet season δ¹⁸O changes ranged from below analytical uncertainty (≤ 0.10 ‰) to up to 0.56 ‰, and the calculated fraction of rainfall contribution to the aquifer is ≤5 % to 16 %. Rainfall recharge via the acrisol soils results in low groundwater EC (20–55 μS/cm), acidic groundwater (pH 3.6–5.6), and may also have resulted in the low groundwater NO3 concentrations (≤ 5.3 mg NO3/L) at many sites due to adsorption, therefore delaying not reducing NO3 contamination. Site specific variations in nitrogen processes includes increased NO3 (to 29.7 mg/L) from fertiliser transfers or nitrification, and increased NH4 (to 1.4 mg/L) likely due to the recharge of irrigation waters. Unlike other recharge areas across the northern Mekong Delta, this north-eastern region provides a groundwater resource unaffected by arsenic contamination. Therefore, these results should inform on priority areas for protection from further contamination by rural anthropogenic activities.