For the past 20 years, forest modellers have largely used the Weibull distribution to summarize diameter distributions and to predict timber volume in a forest at clearfell time. The wider family of Burr distributions is investigated here as an alternative to the Weibull distribution. Using data from 20 permanent sample plots of Pinus radiata, we compare Burr and Weibull fits, showing that the Burr distribution enhances precision by roughly 13%. The effectiveness of the Burr distribution for prediction is found to be similar to that of the Weibull distribution.
|Number of pages||11|
|Journal||Journal of Applied Statistics|
|Publication status||Published - 1996|