Artificial Intelligence-Based Decision Support System for Groundwater Management Under Climate Change: Application to Mornag Plain in Tunisia

This research aims to investigate the influence of climate change on the groundwater level (GWL) in Mornag plain in Tunisia. Due to the spatiotemporal variability of rainfall (RF) and temperature, aquifers all over the world have seen significant water level decline in recent decades. Therefore, it is crucial to analyze and estimate the GWL variability for reliable groundwater (GW) management in the context of climate change. In this study, we focus on the plain of Mornag, located in the southeast of Tunisia, since it contributes 33% of the national agricultural production. From this plain, we have collected historical piezometric and RF data covering the period 2005–2015. Knowing the RF data, our goal is to forecast the GWL. This issue has already been investigated using numerical GW modeling tools such as Modflow and Feflow. Unfortunately, these techniques are data and time-consuming. To overcome all these drawbacks, we propose to use an Artificial Intelligence (AI) approach that has shown great performance in the literature for recurrent data modeling and forecasting. This approach corresponds to the Long Short-Term Memory (LSTM) Neural Network. Compared to Modflow, LSTM showed a notable improvement in terms of minimizing the mean square error, confirming its suitability for GWL forecasts. Using the proposed AI prediction model, the impact of climate change on Mornag GWL has been studied under two Representative Concentration Pathway (RCP) scenarios; RCP 4.5 and RCP 8.5 for three future periods: 2015–2040 (short term), 2041–2065 (medium term), and 2066–2100 (long term). As expected, the results reveal a future decline for Mornag GWL. The performed study of future Mornag GWL behavior using LSTM could classify this AI approach as a good decision support system that could be used to optimize the management of our limited water resources to satisfy the population needs for drinking water and agricultural production, as well as to avert upcoming drought.