Vol.:(0123456789) 1 3 Applied Water Science (2022) 12:272 https://doi.org/10.1007/s13201-022-01798-x ORIGINAL ARTICLE Utilizing deep learning machine for infow forecasting in two diferent environment regions: a case study of a tropical and semi‑arid region Saad Mawlood Saab 1  · Faridah Othman 1  · Chee Ghuan Tan 1  · Mohammed Falah Allawi 2  · Mohsen Sherif 3  · Ahmed El‑Shafe 1,3 Received: 2 July 2022 / Accepted: 17 October 2022 © The Author(s) 2022 Abstract Reservoir infow (Q fow ) forecasting is one of the crucial processes in achieving the best water resources management in a particular catchment area. Although physical models have taken place in solving this problem, those models showed a notice- able limitation due to their requirements for huge eforts, hydrology and climate data, and time-consuming learning process. Hence, the recent alternative technology is the development of the machine learning models and deep learning neural network (DLNN) is the recent promising methodology explored in the feld of water resources. The current research was adopted to forecast Q fow at two diferent catchment areas characterized with diferent type of infow stochasticity, (semi-arid and topi- cal). Validation against two classical algorithms of neural network including multilayer perceptron neural network (MLPNN) and radial basis function neural network (RBFNN) was elaborated and discussed. The research was further investigated the potential of the feature selection algorithm “genetic algorithm (GA)”, for identifying the appropriate predictors. The research fnding confrmed the feasibility of the developed DLNN model for the investigated two case studies. In addition, the DLNN model confrmed its capability in solving daily scale Q more accurately in comparison with the monthly scale. The applied GA as feature selection algorithm was reduced the dimension and complexity of the learning process of the applied predictive model. Further, the research fnding approved the adequacy of the data span used in the current investigation development of computerized ML algorithm. Keywords Reservoir infow · Deep learning · Lead time · Tropical region · Semi-arid region Introduction One of the useful and most direct ways of guiding reservoir operation and management is reservoir infow (Q fow ) predic- tion; it is also useful for food control, reservoir operation, drought management, irrigation water management, and reservoir operation (Rezaeianzadeh et al. 2016; Xu et al. 2021). Using the forecasted Q fow as an input information, the delicacy management of water resources at a reservoir is strongly reliant on precise Q fow predictions (Herbert et al. 2021). In most parts of the world, accurate and real-time daily or monthly prediction of Q fow remains a difcult chal- lenge due to the nonlinearity and non-stationarity of the associated real hydrological data (Kim et al. 2019; Lee et al. 2020). Hence, this research topic has received much atten- tion by the water engineers and decision makers. Reservoir infow prediction has become a major topic in hydrologic time series over the last few decades (Esmaeilza- deh et al. 2017; Bashir et al. 2019; Allawi et al. 2019a). * Ahmed El-Shafe elshafe@um.edu.my Saad Mawlood Saab s2003992@siswa.um.edu.my Faridah Othman faridahothman@um.edu.my Chee Ghuan Tan tancg@um.edu.my Mohammed Falah Allawi mohammed.falah@uoanbar.edu.iq Mohsen Sherif MSherif@uaeu.ac.ae 1 Civil Engineering Department, Faculty of Engineering, Universiti Malaya, 50603 Kuala Lumpur, Malaysia 2 Dams and Water Resources Engineering Department, College of Engineering, University of Anbar, Ramadi 31001, Iraq 3 National Water and Energy Center, United Arab Emirate University, P.O. Box, 15551 Al Ain, United Arab Emirates