Modelling water flowing series of the Mosul Dam Lake for the year 2023-2024(730 days) using support vector mechanism for forecasting in order to train the SVM and determine its most accurate parameters, the first 657 days (0.90 of the Time series data) were utilized for both validation and training. The test set, which was used to determine the prediction potential of SVM, consisted of the remaining 73 days (0.10 of the series). There are numerous varieties of SVM models., These models depending on the different parameters (, c, ) which are experimentally selected for a limited number of values, the appropriate range for C=1,10,100,1000, In order to obtain a more comprehensive analysis, we have expanded the maximum range to 10,000. the appropriate range for γ =0.0001 TO 100, The best model is C=100, γ = 100, which contains 92 support vectors that model has the fewest support vectors and the lowest training error.