Comparison of Double and Triple Exponential Smoothing Methods for Rainfall Prediction

Rainfall is water that falls to the ground surface over a certain period and is measured in millimeters (mm). Rainfall is essential for the life of living things. Forecasting plays a significant role in decision-making in modern times, with two main methods: causal models and time series. Time series models have five types of data patterns: random, constant, seasonal, cyclical, and trend. For rainfall forecasting, the Double Exponential Smoothing and Triple Exponential Smoothing methods are used for trend pattern data. This research compares the two approaches based on error values using average rainfall data in Bojonegoro. The results show that Double Exponential Smoothing has a Mean Absolute Percentage Error (MAPE) of 0.6996%, while Triple Exponential Smoothing has a MAPE of 119.1497%. So, Double Exponential Smoothing is more accurate.