A maximum entropy approach to defining geographic bounds on growth and yield model usage

Abstract

Growth and yield models are essential tools in modern forestry, especially for intensively managed loblolly pine plantations in the southeastern United States. While model developers often have a good idea of where these models should be used with respect to geographic location, determining geographic bounds for model usage can be daunting. Such bounds provide suitable areas where model predictions are likely to behave as expected or identify areas where models may do a poor job of characterizing the growth of a resource. In this research, we adapted a niche model methodology, commonly used to identify suitable spots for species occurrence (maximum entropy), to identify areas for using growth and yield models built from plots established in the Lower Coastal Plain and Piedmont/Upper Coastal Plain in the southeastern United States. The results from this analysis identify areas with similar climatic envelopes and soil properties to the areas where data was collected to fit these growth and yield models. These areas show notable overlap with the areas prescribed for use by the evaluated growth and yield models and support practitioners use of these models throughout these regions. Furthermore, this methodology can be applied to different forest models built using large regional extents as long as climatic and soil values are available for each site.