Introduction: Due to the fact that forecasting and predicting the amount of precipitation and surface water has a significant role in human life and in management plans, it is necessary to study it in different places. Groundwater is an important and vital source of fresh water. which are considered to meet drinking and agricultural needs in urban and rural areas. Management of underground water resources plays a key role in the sustainability of water resources in arid and semi-arid areas. The total amount of water available at any given time is important. Some water users (humans) need water intermittently. The purpose of this research is neural wavelet and multi-layer perceptron models in simulating the monthly flow of water.

Research method: Research method: The research method used in the analysis of the meteorological parameters of Jubala and the land surface is using the factor analysis method. Factor analysis is one of the statistical methods used to reduce the number of variables. The proposed algorithm is implemented in the Matlab 2019 software environment. All the experiments of this research were done on a computer with an Intel® corei5 processor and 12 GB of RAM. To evaluate the proposed method with other methods, three criteria of correlation coefficient and two other criteria of MAE and RMSE have been used for the amount of work error. It is provided by Lar Regional Water Joint Stock Company. Our study area is Lar city. The time domain is 2016-2021.
Results: The results of this research showed that the wavelet neural network was investigated to predict the inflow to the Kamir dam reservoir in Turkey and to predict the inflow of the dam reservoir from the monthly time scale with the combination of discrete wavelet and the optimization of Lunberg-Marquardt based algorithms.