Future groundwater drought analysis under data scarcity using MedCORDEX regional climatic models and machine learning: The case of the Haouz Aquifer

Study region: The Haouz aquifer, situated in central Morocco, a data-scarce region. Study focus: Groundwater resources in semi-arid regions face increasing threats from climate change, particularly due to warming and overexploitation. However, data scarcity limits the ability to monitor and predict...

Full description

Saved in:
Bibliographic Details
Main Authors: El Bouazzaoui Imane, Ait Elbaz Aicha, Ait Brahim Yassine, Machay Hicham, Bougadir Blaid
Format: Article
Language:English
Published: Elsevier 2025-04-01
Series:Journal of Hydrology: Regional Studies
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2214581825000734
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849774831734620160
author El Bouazzaoui Imane
Ait Elbaz Aicha
Ait Brahim Yassine
Machay Hicham
Bougadir Blaid
author_facet El Bouazzaoui Imane
Ait Elbaz Aicha
Ait Brahim Yassine
Machay Hicham
Bougadir Blaid
author_sort El Bouazzaoui Imane
collection DOAJ
description Study region: The Haouz aquifer, situated in central Morocco, a data-scarce region. Study focus: Groundwater resources in semi-arid regions face increasing threats from climate change, particularly due to warming and overexploitation. However, data scarcity limits the ability to monitor and predict groundwater changes accurately. This study addresses this challenge by predicting future drought conditions in the Haouz aquifer using SPI and SPEI climatic drought indices, Machine Learning models, and Med-CORDEX regional climate models under RCP 4.5 and 8.5 scenarios. New Hydrological Insights for the Region: This study is the first in the region to predict groundwater drought based on precipitation and temperature data, relying on the principle of drought propagation. The comparative analysis of the machine learning models shows that Random Forest stands out for its superior predictive performance, influenced by annual trends and long-term climatic indices, with significant contributions from geographical variables. The results indicate a combined influence of land use and natural characteristics on the drought of the Haouz aquifer, following a longitudinal variation and showing a trend towards decreasing variability from the mid- to long-term. Additionally, extreme drought conditions are expected to intensify in most study points particularly under RCP 8.5. The eastern area of the aquifer remains the least impacted by this future trend, continuing to reflect normal drought conditions even in the long term under RCP 8.5.
format Article
id doaj-art-5f5b70f97e5a460fa86d5b323e6b98b4
institution DOAJ
issn 2214-5818
language English
publishDate 2025-04-01
publisher Elsevier
record_format Article
series Journal of Hydrology: Regional Studies
spelling doaj-art-5f5b70f97e5a460fa86d5b323e6b98b42025-08-20T03:01:35ZengElsevierJournal of Hydrology: Regional Studies2214-58182025-04-015810224910.1016/j.ejrh.2025.102249Future groundwater drought analysis under data scarcity using MedCORDEX regional climatic models and machine learning: The case of the Haouz AquiferEl Bouazzaoui Imane0Ait Elbaz Aicha1Ait Brahim Yassine2Machay Hicham3Bougadir Blaid4International Water Research Institute, Mohammed VI Polytechnic University, Ben Guerir, Morocco; Corresponding author.Islamic Financial Engineering Laboratory, Mohammadia School of Engineering, Mohammed V University, Rabat, MoroccoInternational Water Research Institute, Mohammed VI Polytechnic University, Ben Guerir, MoroccoDepartment of Computer Science, Faculty of Science and Technology, Cadi Ayyad University, Marrakech, MoroccoLaboratory of Sciences Applied to the Environment and Sustainable Development, Cadi Ayyad University, Essaouira, MoroccoStudy region: The Haouz aquifer, situated in central Morocco, a data-scarce region. Study focus: Groundwater resources in semi-arid regions face increasing threats from climate change, particularly due to warming and overexploitation. However, data scarcity limits the ability to monitor and predict groundwater changes accurately. This study addresses this challenge by predicting future drought conditions in the Haouz aquifer using SPI and SPEI climatic drought indices, Machine Learning models, and Med-CORDEX regional climate models under RCP 4.5 and 8.5 scenarios. New Hydrological Insights for the Region: This study is the first in the region to predict groundwater drought based on precipitation and temperature data, relying on the principle of drought propagation. The comparative analysis of the machine learning models shows that Random Forest stands out for its superior predictive performance, influenced by annual trends and long-term climatic indices, with significant contributions from geographical variables. The results indicate a combined influence of land use and natural characteristics on the drought of the Haouz aquifer, following a longitudinal variation and showing a trend towards decreasing variability from the mid- to long-term. Additionally, extreme drought conditions are expected to intensify in most study points particularly under RCP 8.5. The eastern area of the aquifer remains the least impacted by this future trend, continuing to reflect normal drought conditions even in the long term under RCP 8.5.http://www.sciencedirect.com/science/article/pii/S2214581825000734GroundwaterClimate changeDroughtMachine LearningRCP scenarioMed-CORDEX
spellingShingle El Bouazzaoui Imane
Ait Elbaz Aicha
Ait Brahim Yassine
Machay Hicham
Bougadir Blaid
Future groundwater drought analysis under data scarcity using MedCORDEX regional climatic models and machine learning: The case of the Haouz Aquifer
Journal of Hydrology: Regional Studies
Groundwater
Climate change
Drought
Machine Learning
RCP scenario
Med-CORDEX
title Future groundwater drought analysis under data scarcity using MedCORDEX regional climatic models and machine learning: The case of the Haouz Aquifer
title_full Future groundwater drought analysis under data scarcity using MedCORDEX regional climatic models and machine learning: The case of the Haouz Aquifer
title_fullStr Future groundwater drought analysis under data scarcity using MedCORDEX regional climatic models and machine learning: The case of the Haouz Aquifer
title_full_unstemmed Future groundwater drought analysis under data scarcity using MedCORDEX regional climatic models and machine learning: The case of the Haouz Aquifer
title_short Future groundwater drought analysis under data scarcity using MedCORDEX regional climatic models and machine learning: The case of the Haouz Aquifer
title_sort future groundwater drought analysis under data scarcity using medcordex regional climatic models and machine learning the case of the haouz aquifer
topic Groundwater
Climate change
Drought
Machine Learning
RCP scenario
Med-CORDEX
url http://www.sciencedirect.com/science/article/pii/S2214581825000734
work_keys_str_mv AT elbouazzaouiimane futuregroundwaterdroughtanalysisunderdatascarcityusingmedcordexregionalclimaticmodelsandmachinelearningthecaseofthehaouzaquifer
AT aitelbazaicha futuregroundwaterdroughtanalysisunderdatascarcityusingmedcordexregionalclimaticmodelsandmachinelearningthecaseofthehaouzaquifer
AT aitbrahimyassine futuregroundwaterdroughtanalysisunderdatascarcityusingmedcordexregionalclimaticmodelsandmachinelearningthecaseofthehaouzaquifer
AT machayhicham futuregroundwaterdroughtanalysisunderdatascarcityusingmedcordexregionalclimaticmodelsandmachinelearningthecaseofthehaouzaquifer
AT bougadirblaid futuregroundwaterdroughtanalysisunderdatascarcityusingmedcordexregionalclimaticmodelsandmachinelearningthecaseofthehaouzaquifer