Estimating of aqueduct water withdrawal via a wavelet-hybrid soft-computing approach under uniform and non-uniform climatic conditions.

Due to climate change and the decrease of surface water resources recently, groundwater resources, especially aqueducts, have special importance to meet various human requirements in arid and semi-arid regions. With the aim of aqueduct water withdrawal (AWW) estimating for agricultural uses, the present research was implemented, in Golpayegan and Kashan regions of Iran; classified in non-uniform and uniform climate zones with water scarcity situation. The AWW variables were estimated based on four scenarios including (1) aqueduct local features, (2) hydrological, (3) land-use, and (4) combined scenarios. The [(Mother-well Depth (MWD), Aqueduct Channel Length (ACL)), (minimum flow rate (Q_Min), maximum flow rate (Q_Max)), and (Cultivated Area (CA), Orchard Area (OA))] variables reagent the first to third scenarios, respectively. Estimation of AWW was operated via single and Wavelet-hybrid (W-hybrid with de-noising) Soft-computing (SC) approaches, including artificial neural networks (ANNs), Wavelet-ANN (WANNs), adaptive neuro-fuzzy inference system (ANFIS), Wavelet-ANFIS (WANFIS), gene expression programming (GEP), and Wavelet-GEP (WGEP). The WGEP model's efficiency with the hybrid characteristics of MWD, ACL, Q_Min, Q_Max, CA, and OA variables was recommended as the best model to estimate AWW variables without climate conditions' effects. With increasing levels of decomposition in wavelet approach and noise reduction, the performance of the models for estimating AWW increased. Also, the findings revealed that the implementation of the proposed method in uniform climates can have a higher performance than non-uniform climates. The achieved values of RMSE for the combined factor of WGEP models were 23.249 and 17.227 (×10³ m³), for estimating AWW in Golpayegan and Kashan, respectively. The performance of WGEP was excellent (R > 0.920) in the estimation of AWW in both climatic types for maximum extreme amounts. Abstracting mathematical formulation of GEP and WGEP models is part of the research finding profound effects implementing policies related to Integrated Water Resources Management to protect the aqueduct's destruction by excessive consumption. [ABSTRACT FROM AUTHOR]