The high amount of Electrical Conductivity (EC) in the groundwater is one of the major negative Geo-environmental problems which has a considerable effect on the quality of drinking water. To address this challenging problem we proposed an intelligent Machine Learning (ML) based approach to predict EC in urban areas. We applied the deep learning technique as one of the most applicable ML techniques with high capabilities for intelligent predictions. Five different deep neural networks (Net 1 to Net 5) were developed in this study and their reliability to predict EC with an emphasis on different settings of inputs, features, functions, and the number of hidden layers was evaluated. The achieved results showed that deep neural networks can predict EC parameters using minimum and economic input parameters. Results showed parameters Cl and SO4 with a high range of correlation and pH with a low range of Pearson correlation properties are influential parameters to be used as the input of neural networks. Activation function Relu, optimization function Adam with a learning rate of 0.0005 and loss function Mean Squared Error with the minimum of two hidden dense layers from Keras laboratory of Tensor Flow developed an efficient and fast network to predict the EC parameter in urban areas. Maximum epochs for developed networks were defined up to 2000 iterations while epochs are reducible up to 200 to drive minimum loss function outcome. The maximum training and testing R2 for developed networks was 0.99 in both the training and testing parts.

peer reviewed | The impact of climate change on water resource availability and soil quality is more and more emphasized under the Mediterranean basin, mostly characterized by drought and extreme weather conditions. The present study aims to investigate how electromagnetic induction technique and soil mapping combined with crop yield data can be used to optimize phosphorus (P) use efficiency by chickpea crop under drip fertigation system. The study was carried out on a 2.5-ha agricultural plot and the agronomic experiments in two growing cycles of chickpea crop. Soil spatial variability was first assessed by the measurement of soil apparent electrical conductivity (ECa) using the CMD Mini-Explorer sensor, and then, soil physicochemical properties were evaluated based on an oriented soil sampling scheme to explore other soil spatial variabilities influencing chickpea yield and quality. Data from the first agronomic experiment were used in geostatistical, multiple linear regression (MLR), and fuzzy c-means unsupervised classification algorithms to properly identify P drip fertigation management zones (MZs). Results from the Person's correlation analysis revealed that chickpea grain yield was more influenced by soil ECa (r = - 0.56), pH (r = - 0.84), ECe (r = - 0.6), P content (r = 0.72), and calcium (Ca) content (r = - 0.83). The proposed MLR-based model to predict chickpea grain yield showed good performances with a normalized root mean square error (NRMSE) of 0.11% and a coefficient of determination (R2) equal to 0.69. The identified MZs were verified by the one-way variance analysis for the studied soil and plant attributes, revealing that the first MZ1 presents a high grain yield, high soil P content, and low ECa. The low fertility MZ2 located in the south part of the studied site presented a low chickpea grain yield due to the low P content and the high ECa. Moreover, the application of P-variable rate fertigation regimes in the second field experiment significantly improved P use efficiency, chickpea grain yield, seed quality, and farmer income by 18%, 12%, 9%, and 136 $/ha, respectively, as compared to the conventional drip fertigation practices. The approach proposed in this study can greatly contribute to optimizing agro-input use efficiency under drip fertigation system, thereby improving farmers' incomes, preserving the ecosystem, and ensuring sustainable cropping systems in the Mediterranean climate.