SILVAH, FIBER, NE-TWIGS, and OAKSIM simulators, commonly used in the northeastern United States, were evaluated by comparing predicted stand development with actual stand development records for periods ranging from 15 to 50 years. Results varied with stand parameter, forest type, projection length, and geographic area. Except in the spruce-fir forest type where FIBER stands out as the best simulator, no single simulator is clearly superior to the others for all locations within a forest type. In general, FIBER, NE-TWIGS, and SILVAH performed best in the northern hardwood (beech- birch-maple) forest type; NE-TWIGS and SILVAH performed best in the Allegheny hardwood (cherry-maple) forest type; SILVAH and OAKSIM performed best in the oak- hickory forest type; and SILVAH was most suitable in the transition hardwood (mixture of northern hardwoods and oaks) forest type. The results give growth and yield model users more information for selecting the simulator most suitable for their particular needs. The results also can be used as a diagnostic tool for growth and yield model development.

S2Describes a species-specific, distance-independent individual-tree diameter growth model for the Northeastern United States. Diameter growth is predicted in two steps using a two parameter, sigmoidal growth function modified by a one parameter exponential decay function with species-specific coefficients. Coefficients are presented for 28 species groups. The model accounts for variability in annual diameter growth due to species, tree size, site quality, and the tree's competitive position within the stand. Model performance is evaluated using the mean predicted error and the root mean square error. Results are presented for the calibration data and an independent validation data set. The model has been, incorporated into NE-TWIGS, a computerized forest growth model for the Northeastern United States.S3.

S2In West Virginia, yellow-poplar (Liriodendron tulipifera) is abundant and is a prime candidate for increased utilization in a variety of manufacturing industries. Computer simulations are a cost-effective tool for estimating potential cutting yields from lumber. They can be used to promote increased use of yellow-poplar in the furniture, cabinet, and architectural woodworking industries and may also lead to increased utilization of the lower grades of lumber. This paper describes the data collection methods and the format of the new West Virginia yellow-poplar lumber defect database that was developed for use with computer simulation programs. The database contains descriptions of 627 boards, totaling approximately 3,800 board feet, collected in West Virginia for grades FAS, FASIF, No. 1 Common, No. 2A Common, and No. 28 Common.S3.