Perspectives for the application of computer models to forest dynamics forecasting in bieszczadzki national park (Poland)

This paper presents the perspectives for application of computer models in forecasting the dynamics of forest development on example of Moczarne area, in Bieszczadzki National Park, based on authors’ computer models. First, the possibilities for forecasting the dynamics of forest development in a local scale, i.e. within single rectangular or circular study plot, are presented. For this purpose, a computer prognostic model FORKOM E has been applied, using both general mathematical relationships functioning within a forest ecosystem and empirical ones, characteristic for tree stands at analysed plots. Additionally, a layer of 3D visualisation of a tree stand, which is an integral part of the mentioned model, is also presented. Presented also are the possibilities for forecasting the dynamics of forest development at landscape scale, applying the theory of cellular automata. For this purpose, a prognostic computer model CELLAUT was used in which the whole analysed tree stand is understood as a set of single cells, where stages of landscape development dominating within those cells are considered as also the influence of particular cells upon their neighbours. The paper also describes the perspectives for application of self-learning neural networks in the process of supplementation and verification of some parameters of a tree stand, calculated by the above-mentioned models.