Scholarlink

<p>Groundwater drought refers to a period of decreased groundwater level. Systematic information about the likely occurrence and distribution of drought may assist in the preparedness and mitigation of drought disasters. A study was carried out in the Kalpathypuzha watershed, which is one of the most drought-prone regions of the Bharathapuzha river basin in Kerala. This research aimed to develop an Artificial Neural Network (ANN) model for groundwater level (GWL) prediction and to assess groundwater drought using the Standardized Groundwater Index (SGI). Twelve observation wells evenly distributed in the blocks of Kuzhalmannam, Palakkad, Malampuzha, and Chittur were selected for the study. The groundwater level data for 15&#xa0;years, from 2007 to 2021, were used as the target data for ANN model development. The Standardized Groundwater Index (SGI) values were calculated to evaluate the groundwater drought conditions in the study area. The results showed that the years 2013, 2016, and 2017 were the most severely affected by drought in the region. According to the spatial distribution of SGI values for 2013, 2016, and 2017, Chittur and Malampuzha blocks were the most drought-affected, followed by the Kuzhalmannam and Palakkad blocks. The machine learning Feed Forward ANN models were developed using MATLAB R 2016a software to predict the groundwater level of each well. The input parameters for model development were precipitation and maximum and minimum temperature. The predicted groundwater level was in close agreement with the observed groundwater level in the study area with performance indicators, Correlation Coefficient R (0.93 to 0.74), Root Mean Square Error RMSE (0.11 to 0.45&#xa0;m), and Coefficient of Determination R<sup>2</sup> (0.87 to 0.69) in the acceptable range. The developed ANN model predicted the monthly groundwater levels and drought conditions for 2023. The analysis showed that the input variables of the model significantly impacted groundwater levels, demonstrating that the ANN model is a reliable predictive tool for modeling groundwater levels and assessing groundwater drought.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Characterization of spatio-temporal groundwater drought in Kalpathypuzha watershed, India

  • 期刊论文

  • 发表日期：

    2025年7月14日

查看全文
  • Rabeea Assainar K K,
  • Asha Joseph,
  • Thendiyath Roshni,
  • Josephina Paul,
  • Sheeja P S

摘要

Groundwater drought refers to a period of decreased groundwater level. Systematic information about the likely occurrence and distribution of drought may assist in the preparedness and mitigation of drought disasters. A study was carried out in the Kalpathypuzha watershed, which is one of the most drought-prone regions of the Bharathapuzha river basin in Kerala. This research aimed to develop an Artificial Neural Network (ANN) model for groundwater level (GWL) prediction and to assess groundwater drought using the Standardized Groundwater Index (SGI). Twelve observation wells evenly distributed in the blocks of Kuzhalmannam, Palakkad, Malampuzha, and Chittur were selected for the study. The groundwater level data for 15 years, from 2007 to 2021, were used as the target data for ANN model development. The Standardized Groundwater Index (SGI) values were calculated to evaluate the groundwater drought conditions in the study area. The results showed that the years 2013, 2016, and 2017 were the most severely affected by drought in the region. According to the spatial distribution of SGI values for 2013, 2016, and 2017, Chittur and Malampuzha blocks were the most drought-affected, followed by the Kuzhalmannam and Palakkad blocks. The machine learning Feed Forward ANN models were developed using MATLAB R 2016a software to predict the groundwater level of each well. The input parameters for model development were precipitation and maximum and minimum temperature. The predicted groundwater level was in close agreement with the observed groundwater level in the study area with performance indicators, Correlation Coefficient R (0.93 to 0.74), Root Mean Square Error RMSE (0.11 to 0.45 m), and Coefficient of Determination R2 (0.87 to 0.69) in the acceptable range. The developed ANN model predicted the monthly groundwater levels and drought conditions for 2023. The analysis showed that the input variables of the model significantly impacted groundwater levels, demonstrating that the ANN model is a reliable predictive tool for modeling groundwater levels and assessing groundwater drought.