Kumar et al. (2025) Forecasting the Level of Groundwater in India Using a Machine Learning

Identification

• Journal: Lecture notes in networks and systems
• Year: 2025
• Date: 2025-11-24
• Authors: Anuj Kumar, Tarun Kumar Sharma
• DOI: 10.1007/978-981-96-7753-5_13

Research Groups

• Uttaranchal University, Dehradun, India
• Shobhit University, Gangoh, Saharanpur, India

Short Summary

This study develops and compares machine learning models to forecast yearly groundwater availability and extraction usage in India for 2025, finding that the Random Forest algorithm achieves the highest accuracy of 82%.

Objective

• To forecast the yearly groundwater availability and anticipate average groundwater availability and extraction usage for the year 2025 in India using various machine learning algorithms.

Study Configuration

• Spatial Scale: India
• Temporal Scale: Yearly forecasts for future use, specifically anticipating for the year 2025.

Methodology and Data

• Models used: Logistic Regression, Random Forest, Support Vector Machine (SVM), K-Nearest Neighbor (KNN).
• Data sources: Input variables for modeling depend on water utilization and recharge over different seasons.

Main Results

• All evaluated machine learning models were capable of producing accurate forecasts for groundwater levels.
• The Random Forest algorithm demonstrated the highest performance among the tested models, achieving an 82% accuracy rate.

Contributions

• This research contributes by applying and comparing multiple machine learning algorithms for forecasting yearly groundwater availability in India, a critical resource facing depletion.
• It identifies Random Forest as the most accurate model for this specific forecasting task, providing a robust tool for water preservation and drought mitigation efforts.

Funding

• Not specified in the provided paper text.

Citation

@article{Kumar2025Forecasting,
  author = {Kumar, Anuj and Sharma, Tarun Kumar},
  title = {Forecasting the Level of Groundwater in India Using a Machine Learning},
  journal = {Lecture notes in networks and systems},
  year = {2025},
  doi = {10.1007/978-981-96-7753-5_13},
  url = {https://doi.org/10.1007/978-981-96-7753-5_13}
}