Short-term forecasting of monthly water consumption in hyper-arid climate using recurrent neural networks

Freshwater supply is a major challenge in regions with limited water resources and extremely arid climatic conditions. The objective of this study to model the monthly demand Kuwait using nonlinear autoregressive exogenous input (NARX) neural network approach. country lacks conventional surface characterized by climate. In addition, it has one fastest growing populations. study, linear detrending performed on consumption time series for period from January 1993 December 2018 eliminate influence population growth before application NARX model. Monthly temperature data are selected as model, because they strongly associated data. Correlation analyses determine feedback delays results demonstrate that recurrent efficient robust forecasting short-term demand, Nash-Sutcliffe (NS) coefficient 0.837 validation period. Seasonal assessment shows performs best during critical summer season. NARX-based established powerful promising tool predicting urban demand. Thus, can efficiently aid development resilient plans.