A review of the applicability of existing tree and forest characteristics prediction models to forest inventory in Vietnam and Nepal

Abstract

Forest inventories provide vital and up-to-date information for use in basic decision making on the ma¬nagement and conservation of forest resources. Data collected in forest inventories are stored and pro¬cessed in databases which can be updated by conducting additional measurements or by applying pre¬dictive models for imputing missing values of tree and forest stand-level variables. The inventory results can thereafter be calculated based on sample units, i.e. sample plots or forest stands within them, after which the forest inventory variables can be aggregated using different stratification units. For strategic decision-making, however, the future development of forest resources needs to be predicted. For this purpose, growth and yield simulators comprising tree and stand-level growth models are utilised to ob¬tain prediction results for alternative scenarios based on inventory information, i.e. sample-based field data. In large-scale forest inventories, only easily assessable characteristics are measured for all tallied trees, whereas height characteristics and other variables, which are difficult to measure accurately, are collected from a sub-sample only. In order to generalise the variables measured from sample trees to also cover tally trees, generalization techniques need to be applied. The ongoing national-level forest as¬sessments conducted in Nepal and Vietnam require efficient calculation procedures for reporting inven¬tory results and quantifying the availability and location of forest resources. The aim of this review was to assess the availability of the existing models for the prediction of tree and forest characteristics and their applicability to large-scale forest inventory in Nepal and Vietnam. Through comparisons made bet¬ween country- and species-specific models and prediction systems and through an assessment based on modelling literature, recommendations are also given for further developing the model-based prediction systems used in the ongoing national forest inventories of Nepal and Vietnam. The existing model sets can be used to estimate conventional stand volume characteristics for the inventoried areas. However, according to the new reporting requirements set for the current National Forest Inventory (NFI) of Viet¬nam and the Forest Resource Assessment (FRA) of Nepal, it is recommended that their model bases, which are currently under upgrading, be updated and improved in the future.