Hydrodynamic Analysis and Clustering of Piezometers of Baghmalek Plain Aquifer Using the K-Means Algorithm

Majedi, Hamid Reza

doi:10.22048/wss.2026.588846.1032

Hydrodynamic Analysis and Clustering of Piezometers of Baghmalek Plain Aquifer Using the K-Means Algorithm

Document Type : Research Article

Author

Hamid Reza Majedi

Khuzestan Water and Power Authority

10.22048/wss.2026.588846.1032

Abstract

Groundwater is essential for drinking, agriculture, and industry, especially in arid regions like Iran. The Baghmalek plain aquifer, covering 49 km² in northeastern Khuzestan, faces serious challenges from over-extraction, climate change, and falling water levels. This study applies the K-Means algorithm to analyze hydrodynamic behavior and cluster piezometers in this aquifer based on water table fluctuations. Water level data from five active piezometers were collected over 144 months (October 2011–September 2023), normalized, and processed in Orange software. The K-Means analysis grouped the piezometers into three distinct clusters. Silhouette index values (0.5–0.685) confirmed good clustering quality, indicating strong hydrodynamic similarity within each group. Long-term hydrographs and groundwater flow direction maps validated these results. These findings support redesigning the groundwater monitoring network and optimizing new piezometer placements. Additionally, identifying zones with similar hydrodynamic behavior can guide the strategic siting of production wells and targeted abstraction planning, promoting more sustainable management of the Baghmalek aquifer.

Keywords

Groundwater

Baghmalek aquifer

Hydrodynamic clustering

K-means algorithm

Orange software

Subjects