Title: Conjunction of cluster ensemble – model ensemble techniques for groundwater drought assessment in semi-arid regions
Abstract:
Due to the importance of groundwater protection for drinking and agricultural purposes, selecting practical approaches to evaluate the drought condition of these vital resources and recognizing the spatiotemporal pattern of groundwater level (GWL) is among the essential requirements for making informed decisions on water resources protection. To enhance such GWL modeling and its drought condition, in this study first 3 clustering methods including linear K-means, fuzzy based FCM and neural network based methods and their ensemble results were used to cluster observatory wells into several clusters and then, for the centroid well of each cluster 3 machine learning methods were applied for GWL prediction and then neural based ensemble was used to combine the obtained results to enhance the overall performance of the modeling. Hybrid validity index and coefficient of determination metrics were used for clustering and prediction evaluation results. The results showed up to 25% of improvement in clustering and up to 23% enhancement on prediction performance of the proposed ensemble techniques, respectively. Finally using the GWL predictions, groundwater drought conditions assessed in each cluster, highlighting different drought conditions at different clusters.
Short Biography:
Dr. Elnaz Sharghi, received her PhD in Civil Engineering- Hydraulic Structures at 2014 with honor. While studding her PhD, she received Iran MoST full scholarship and after graduation, Dr Sharghi joined Department of Water Resources Engineering- Faculty of Civil Engineering at University of Tabriz, IRAN as Assistant Professor and then promoted to Associate Professorship in 2022. Her research interests include Hydrogeology, AI and ML application in hydrology and geology and groundwater modeling. Her research outcomes published as more than 45 journal papers, several conference presentations as well as some national projects. She supervised more than 40 M.S. and 4 PhD students. Dr Sharghi spent one-year sabbatical in CAS-Beijing as visiting associate professor in 2023.