OPTIMAL CROSS-CUTTING AND SENSITIVITY ANALYSIS FOR VARIOUS LOG DIMENSION CONSTRAINTS BY USING DYNAMIC-PROGRAMMING APPROACH

An application of a forward reaching algorithm of dynamic programming is presented for cross-cutting trees. Sensitivity analysis was applied for evaluating the effects of various log length and dimension constr aints on the produced yield and the relationships between the timber v olumes obtained and the dendrometric properties of the data were analy sed. The data used consisted of 3283 Scots pine (Pinus sylvestris L.) stems collected from 52 forest stands located in seven regions in Finl and, The timber grades used were branchless logs, logs with living bra nches, logs with dead branches, pulpwood and waste. It was shown that the volumes of different grades vary considerably when different dimen sional criteria are applied and that log length constraints seem to af fect the yield more than the limit for the minimum top diameter. By re ducing the minimum log length from 37 dm to 22 dm the volume of the br anchless log grade was tripled in some stands. The volume gain obtaine d by loosening the dimension constraints was generally largest iu stan ds of poor quality. The average height of the lower part of the dead c rown was found to be the best indicator for the amount of the branchle ss timber obtained in a stand. In general. the timber volumes obtained varied a lot between the stands of a region as well as within a stand . Especially between-tree variation within a stand makes the estimatio n of the timber volume unreliable without some measurement data.