What is the difference between additive and multiplicative time series decomposition?

Additive decomposition assumes constant seasonal effect, while multiplicative assumes seasonality scales with trend.

Time series decomposition separates data into trend, seasonality, and residuals. In additive models, components are summed: Y = T + S + R. In multiplicative models, components are multiplied: Y = T × S × R. Multiplicative is used when seasonal fluctuations increase with trend magnitude, while additive is used when seasonality is constant.