Accurate rainfall prediction plays a crucial role in agriculture, disaster management, and climate analysis. Traditional statistical methods often struggle to model complex relationships among meteorological variables. In this research, a rainfall prediction model using synthetic weather data and deep neural networks is proposed. A synthetic dataset resembling weather data collected by the National Oceanic and Atmospheric Administration was generated using statistical distributions for temperature, humidity, wind speed, pressure, and rainfall. The dataset was preprocessed using normalization techniques and divided into training and testing sets. A deep neural network implemented using TensorFlow and Keras was trained to predict rainfall. Experimental results demonstrate that neural networks can effectively learn nonlinear relationships among atmospheric parameters. The model achieved satisfactory prediction accuracy measured using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and R-squared metrics. The proposed approach can assist in weather forecasting systems and climate monitoring applications.
