Process-based models for forest ecosystem management: current state of the art and challenges for practical implementation

Recent progress toward the application of process-based models in forestmanagement includes the development of evaluation and parameter estimation methods suitable for models with causal structure, and the accumulation of data that can be used in model evaluation. The current state of the art of process modeling is discussed in the context of forest ecosystem management. We argue that the carbon balance approach is readily applicable for projecting forest yield and productivity, and review several carbon balance models for estimating stand productivity and individual tree growth and competition. We propose that to develop operational models, it is necessary to accept that all models may have both empirical and causal components at the system level. We present examples of hybrid carbon balance models and consider issues that currently require incorporation of empirical information at the system level. We review model calibration and validation methods that take account of the hybrid character of models. The operational implementation of process-based models to practical forest management is discussed. Methods of decision-making in forest management are gradually moving toward a more general, analytical approach, and it seems likely that models that include some process-oriented components will soon be used in forestry enterprises. This development is likely to run parallel with the further development of ecophysiologically based models.