Water Supply (May 2022)

Potential impacts of climate change on groundwater levels in Golpayegan Plain, Iran

  • Mahnoosh Moghaddasi,
  • Mehdi Mohammadi Ghaleni,
  • Mahmood Akbari,
  • Sedigheh Salari

DOI
https://doi.org/10.2166/ws.2022.179
Journal volume & issue
Vol. 22, no. 5
pp. 5295 – 5308

Abstract

Read online

Groundwater level forecasting is an essential priority for planning and managing groundwater resources. This study aims to investigate the effect of climate change on the monthly groundwater level in the Golpayegan aquifer in the future (2017–2032). After a spatio-temporal analysis, the Least Squares Support Vector Regression (LSSVR) model was used to simulate the monthly groundwater level in the historical period (2002–2017). The input data included precipitation, temperature, pan evaporation, soil moisture (from the ESA CCA SM product), and groundwater level in observation wells on a monthly time-scale. Future climatic data were downloaded from the CanEsm5 model of CMIP6 for the SSP1-2.6 and SSP5-8.5 climate scenarios and then downscaled using the Change Factor Approach (CFA). The spatial analysis of groundwater levels indicated four different behaviors in the observation wells in the Golpayegan aquifer, resulting in four different clusters using the AGNES clustering method. Historical and future period modeling were performed separately for each of the four observation wells from each cluster. The modeling in the historical period demonstrated an average of NRMSE (0.09), MBE (0.030), and R2 (0.94) for the four clusters. The groundwater level in all clusters showed a decreasing trend in the future period, with SSP5-8.5 (average: 3.9 cm/month) showing a greater decrease than the SSP1-2.6 (average: 0.5 cm/month) scenario. The decline in groundwater level under SSP5-8.5 compared with SSP1-2.6 was more, respectively, 4.8, 5.8, 9.9 and 3.7 metres for clusters 1–4. The results indicate the acceptable efficiency and accuracy of the LSSVR model results in evaluating the effects of climate change on groundwater levels. HIGHLIGHTS Use of historical and future period data such as variables including rainfall, temperature, soil moisture, free water evaporation and groundwater level.; Using new clustering and multivariate regression methods in the data-mining process.; Using the LSSVR learning machine as a groundwater-level prediction tool.;

Keywords