Modelling diameter distribution of Tetraclinis articulata in Tunisia using normal and Weibull distributions with parameters depending on stand variables

Abstract

The objective of this study was to evaluate the effectiveness of both Normal and two-parameter Weibull distributions in describing diameter distribution of Tetraclinis articulata stands in north-east Tunisia. The parameters of the Weibull function were estimated using the moments method and maximum likelihood approaches. The data used in this study came from temporary plots. The three diameter distribution models were compared firstly by estimating the parameters of the distribution directly from individual tree measurements taken in each plot (parameter estimation method), and secondly by predicting the same parameters from stand variables (parameter prediction method). The comparison was based on bias, mean absolute error, mean square error and the Reynolds’ index error (as a percentage). On the basis of the parameter estimation method, the Normal distribution gave slightly better results, whereas the Weibull distribution with the maximum likelihood approach gave the best results for the parameter prediction method. Hence, in the latter case, the Weibull distribution with the maximum likelihood approach appears to be the most suitable to estimate the parameters for reducing the different comparison criteria for the distribution of trees by diameter class in Tetraclinis articulata forests in Tunisia.

Keywords

• Diameter Class Model
• Normal Distribution
• Weibull Distribution
• Maximum Likelihood Approach
• Moments Method
• Tetraclinis articulata