Abstract: Objectives: Precipitable water vapor (PWV) plays an increasingly significant role in the quantitative study of the potential meteorological factors that cause rainfall. The PWV-based three-factor (PWV, PWV change and rate of PWV change) threshold method for the rain forecast has been established, empirically proving its effectiveness in some scenarios. However, an apparent issue is that not fully using real-time information restricts performance. Our study proposed an improved monthly threshold method to tackle this problem. Methods: The basic idea is to refine the predictors. In this paper, the three-factor monthly threshold method based on PWV, PWV increment and rate of PWV increment is used to realize the short-term rainfall forecast within 6 hours, in which PWV is used as the main predictor, and PWV increment and PWV increase rate are used as secondary factors to assist in the prediction. We offer a quantitative standard for picking the threshold based on maximizing Critical Success Index (CSI). The specific prediction steps are as follows: (1) Get the time series of the three predictors from the raw data. Set a sliding window length of 6 hours, after calculating the PWV increment and rate of PWV increment under the current window, slide the window backwards for 30 minutes. By repeating the operations, we obtain the PWV increment sequence and the rate of PWV increment sequence; (2) Select the monthly threshold based on the data from 2015 to 2018. Set the optimal threshold for PWV according to CSI every month, then fix the threshold of PWV, and use the CSI maximum principle to select the optimal thresholds for PWV increment and the rate of PWV increment jointly; (3) Use the optimal threshold set on the 2019 data to predict the rainfall in the next 6 hours. If the PWV exceeds the threshold at a certain moment, predict rainfall; if the PWV does not exceed the threshold at a certain moment, but both the PWV increment and the rate of PWV increment exceed the threshold, predict rainfall; otherwise, predict no rainfall. Finally, the forecasted rainfall and actual rainfall are counted to calculate the correct rate (CR), false alarm rate (FAR) and CSI. Meanwhile, we explain that compared with using Probability of Detection (POD), it is more reasonable to use CR to evaluate the proportion of predicted rainfall events. Results: We apply our method to 11 different stations of SuomiNet. The CR of the improved method is above 89%, while the FAR is controlled below 73%. Among them the CR of the three stations P031, P047 and CN00 is about 95%, and the FAR is not higher than 65%. In general, the proposed method can predict more than 92% of rainfall on average, and the average FAR of each forecast is about 63%. Compared with the traditional three-factor algorithm, the average CR of the improved three-factor algorithm is increased by nearly 6%, the average FAR is reduced by more than 4%, and the average POD is at the same level. Conclusion: Using PWV increment and the rate of PWV increment, the new predictor can better reflect the characteristics of the rising phase of PWV and keep updating synchronously with PWV. Compared with the traditional method, our algorithm can predict rainfall more effectively and has higher applicability.