Prediction of Groundwater level at Varanasi using Neural Network

Abstract

Groundwater (GW) is a crucial constituent of human life. Freshwater is deposited in various landscape areas, including rivers, creeks, lakes, reservoirs, and streams. Due to the rapid increase in population, urbanization, and industrialization demand for water resources has increased. Also, Changes in the land pattern, topography, and meandering of rivers affect the water beneath the surface. Hence there is a need to forecast the GW level to maintain the effective development in groundwater resources. An excellent modeling technique, an Artificial neural network (ANN) has been introduced in this paper to predict the groundwater level of Varanasi, India. ANN is a mathematical model which works on the concept of the nerve cell. An ANN model was introduced in this investigation to predict the pre-monsoon and post-monsoon groundwater fluctuation at Varanasi, Uttar Pradesh India.  This study explains the groundwater level involved within the boundaries of river Ganga, further investigated by using artificial neural networks with the help of various metrological parameters like rainfall (R), relative humidity (RH), most extreme temperature (T_max), least temperature (T_min.) etc. The best productive models were examined by using various types of architectural networks, such as the hidden layers, number of neurons, and activation function with an additional percent of data in training, validation, and testing. The architectural network of 4-9-1 produces the best outcomes. Hence, it can be applied to forecast GW fluctuation in urban areas, as it will effectively manage water resources.