Mapping and imputing potential productivity of Pacific Northwest forests using climate variables

Regional estimation of potential forest productivity is important to diverse applications, including biofuels supply, carbon sequestration, and projections of forest growth. Using PRISM (Parameter-elevation Regressions on Independent Slopes Model) climate and productivity data measured on a grid of 3356 Forest Inventory and Analysis plots in Oregon and Washington, we evaluated four possible imputation methods to estimate potential forest productivity: nearest neighbour, multiple linear regression, thin plate spline functions, and a spatial autoregressive model. Productivity, measured by potential mean annual increment at culmination, is explained by the interaction of annual temperature, precipitation, and climate moisture index. The data were randomly divided into 2237 reference plots and 1119 target plots 30 times. Each imputation method was evaluated by calculating the coefficient of determination, bias, and root mean square error of both the target and reference data set and was also tested for evidence of spatial autocorrelation. Potential forest productivity maps of culmination potential mean annual increment were produced for all Oregon and Washington timberland.