COMBINING HARVEST SAMPLE DATA WITH INVENTORY DATA TO ESTIMATE FOREST BIOMASS

Forest biomass databases which go beyond stem volume are needed for ca rbon balance calculations and also for forest monitoring, forest damag e inventories and the solution of other related problems. In this arti cle, it is shown how recursive multiple regression analysis can combin e forest inventory data with biomass harvest data. The stand-level con version factor derived for Pinus sylvestris L. branch biomass for the ''Severka'' Forest Farm shows a reasonable goodness of fit, with a coe fficient of determination of R-2 = 0.87. Statistically significant coe fficients were achieved by including stand density and the quadratic m ean diameter in the regression. Equations developed for Switzerland, u sing Burger's Picea abies (L.) Karst. branch and needle conversion fac tors for individual trees, show R-2 = 0.87 and R-2 = 0.91, achieved by taking h(dom,50), elevation, diameter at breast height and age into a ccount.