Abstract

For Authors Guide


Application of Artificial Intelligence in addressing Land degradation: A case study of Jumar watershed, Ranchi, Jharkhand

Abstract


Land degradation is a globally recognized problem, causing a whooping damage of billions for both developed and developing countries. The purpose of the study is to understand the role and potential of AI-based models in analysing water quality which is one of the crucial indicators of land degradation. Through the literature analysis, it has been found that models like Artificial neural network (ANN), Support Vector Mechanism (SVM), Classification and Regression trees (CART), random forest (RF), XGBoost (XGB) etc., are widely used and exhibit higher accuracy in regression, modelling, and prediction of water quality data. For the understanding of the application of a few of these models, a case study of the Jumar watershed of Ranchi is chosen to predict the Water Quality Index (WQI) using the seasonal physiochemical data of water quality parameters for the years 2021 and 2022. 29 surface water sampling sites were selected for laboratory analysis based on APHA and IS 10500:2012. The results have been analysed using standards provided by WHO, BIS, and ICMR, and crucial parameters were extracted using Principal component analysis (PCA). The data indicated poor water quality throughout the year for both 2021 and 2022. The WQI has been determined using the Weighted water quality index (WWQI) and predicted using SVM, ANN and CART models. The results show that ANN (R2 =0.29) has comparatively poor performance while the SVM (R2 = 0.82) and CART (R2 = 0.90) models were found to be best in water quality prediction for the Jumar watershed. The data was validated using RSME which was found to be less for SVM and CART compared to ANN. It could be concluded that the CART model was best for predicting the surface water quality index of the Jumar watershed among the applied machine learning algorithms.



Keywords


Artificial intelligence; Artificial neural network; CART; SDG; soil quality; Support Vector Mechanism; water quality; Weighted water quality index