Simulated net primary productivity (NPP) anomalies (percent deviation) in 1961-2018 years relative to the 1961–1990 reference period for Picea abies and Fagus sylvatica. NPP was simulated for the species' potential distribution range in Switzerland on a 1 × 1 km grid using 3-PG model. We first assimilated nearly 800 observation years from 271 permanent long-term forest monitoring plots across Switzerland, obtained between 1980 and 2017, into the 3-PG forest ecosystem model using Bayesian inference, reducing the bias of model predictions from. We then estimated the NPP anomalies by first simulating the growth of P. abies and F. sylvatica monocultures with the average climate observed during the 1961–1990 period, until the age of 40 years (spin-up). The stands were simulated starting as 2-year-old plantations with an initial density of 10,000 trees/ha. Thinning was performed at age 20 and 35 to reach a final density of ca. 1,000 trees/ha at age 40. We then simulated 30 years forced by monthly resolved climatic data from either the 1961–1990 (reference, according to MeteoSwiss) or the 1991–2018 period. We neglected the first 40 years of simulations due to high variation in productivity caused by early stage stand development. To study the impact of climate extremes on NPP, we focused on the deviation in NPP (expressed in percentage difference from the reference period) during the 30 year period (age 41–70).