| Abstrakt: |
Using a clustering and tree method combination, this research looked at the prediction of changes in irrigation network groundwater depth in the Abyek plain. Groundwater depth variations in various plain regions were examined initially, utilizing the K-means technique for geographic grouping and aquifer depth changes. It was then applied to a tree algorithm using K-means findings. A tree method was then used to forecast changes in aquifer depth across all clusters. There were five clusters of groundwater alterations based on the K-means algorithm findings, and aquifer decline increased from cluster 1 to 5. Clusters 1 and 5 showed the greatest increases in aquifer depth and the greatest decreases. K-means and classification and regression tree findings show that in locations where the most aquifer decline was recorded, human causes were successful, while in regions where the highest groundwater depth rise was found, natural factors were effective. Factors of precipitation, agricultural water demand (million cubic meters), and water delivered to irrigation network (million cubic meters) in regions with high aquifer drawdown and factors of volume of precipitation, water delivered to irrigation network, and air humidity percentage in regions with increased groundwater depth had the greatest impact. For varied variations in the Abyek plain groundwater depth, rainfall volume was evident in most tree diagrams. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink