A Comprehensive Survey on CNN Models on Assessment of Nitrate Contamination in Groundwater

Abstract

In many places in the world, groundwater nitrate pollution is a major issue. Close to the livestock waste disposal site (LWDS), coprostanol and nitrate concentrations in the soil were altered by livestock manure. There was a considerable correlation between the nitrate contents in the groundwater and soil. There was evidence that nitrates were carried downstream in both soil and groundwater. It is, however, difficult to identify the main nitrate sources because of the diffuse and widespread spatial overlap of multiple non-point pollution sources. This research study presents a comprehensive survey and evaluation of various convolutional neural network (CNN) models for the assessment of groundwater nitrate contamination. The survey provides the accuracy of various models of CNN method that records the prediction accuracy of groundwater nitrate contamination. The model provides an accuracy evaluation with the proposed method on nitrate concentration and shows how well the proposed method archives better accuracy than other CNN models.