Evaluation of forest growth simulators with NFI permanent sample plot data from Finland

In general the examined simulators seem to work well in circumstances where natural mortality is low or does not exist, as happens to be the case in intensively managed commercial forests. Instead, in unmanaged, unhealthy or non-homogeneous forests the estimates are less reliable. Usually the models tend to underestimate the natural mortality, but with the combined simulator and the birch strata the mortality was also often highly overestimated. It seems necessary to make both the growth and mortality models more adaptable to varying conditions in the future. Also the models for the deciduous trees require improvement.