Empirical forest growth models that are incorporated into software simulation systems are an analytical tool necessary for sustainable forest management at the stand and landscape scale. The complex interactions of model form, forest characteristics, sampling design and data processing techniques were examined through simulation using common California simulators. Results showed that the simulation techniques employed were capable of detecting introduced bias. Certain combinations of factors did cause biased projections over a 20-year period. In some cases simple techniques, such as sample plot averaging over a stand, could be employed to correct for the bias.

Incorporating climatic and topographic variables into a traditional empirical forest growth model structure was hypothesized to improve long-term model behavior in the context of climate change and variability. Tree growth data was assembled from four completed research projects across the interior of northern California. Combined with historic monthly climate data, the tree growth data was used in mixed effects linear regression models of tree diameter and height growth for six conifer species. Evaluations of the models appeared reasonable in both theoretical behavior and prediction using independent data.

Downscaled General Circulation Model (GCM) climate projections for California were used to examine forest growth to 2099. An elevational east-west transect north of Lake Tahoe was used. Existing mature stands and young plantations were simulated. Forest productivity increased in general, up to 15% for plantations. The use of existing tree growth and climate data combined with the latest biometrical knowledge was shown to provide a reasonable next step in improving forestry and ecosystem planning in the context of climate change and adaptation.