USING A GEOGRAPHIC INFORMATION SYSTEM AND GEOSTATISTICS TO ESTIMATE SITE INDEX OF PINUS RADIATA

Site index is used as a measure of productivity for large plantation forests. Although site index had been calculated in less than half of the compartments in Kaingaroa Forest, data were fairly evenly spread. A Geographic Information System (GIS) was used to produce a contour map of site indices associated with compartment centre points. The limitations of estimation techniques within the GIS were highlighted by the difficulty of predicting values between contour lines. Instead, geostatistics, a statistical interpolation method, was adopted as it can estimate local values from data that varies spatially. The variogram for site index in Kaingaroa Forest was fitted by a linear model up to 25 km. The parameters of this model were used in estimation (kriging) procedure. Values for 757 compartments were predicted, ranging from 18.8 to 34.3 m. The standard error ranged from 1.6 to 3.6 m, with a mean of 1.7 m. A jack-knifing procedure showed estimates to agree well with actual values. It was concluded that linking a GIS with geostatisties allowed more effective use to be made of the GIS