Predictive modelling of eucalyptus tree metrics using Stump Measurements

Predictive modelling of eucalyptus tree metrics using Stump Measurements

Estimating diameter at breast height (DBH) from stump measurements can be used for calculating tree diameter distributions and predicting tree height (H) and other tree metrics.
Stump data in terms of stump diameter (SD) and stump height (SH) often provides a foundation for reconstructing pre-harvest stand conditions, particularly in cases of unplanned or unauthorised tree removals.
To enable these estimations, two regression equations were developed, the first to predict DBH from outside-bark stump measurements. These DBH estimates were subsequently used in a second equation tailored to determine tree H. This approach could support applications in carbon stock assessments, habitat quality evaluations, and biodiversity monitoring where rapid yet accurate tree dimension data is needed.
The equations were derived from data collected from 35 mature Eucalyptus tereticornis (Forest Red Gum) urban trees, with heights ranging between 23 to 42 meters. This eucalyptus species, widespread in Southeast Queensland, Australia, offers sustainable, resilient timber that is both cost-effective and ecologically advantageous due to its fast growth and carbon-sequestering properties.
Trees H was measured using a surveying total station. Stump diameters were recorded at heights above ground level ranging from 0.20 m to 1.30 m, capturing likely SH below the DBH, which was conventionally measured at 1.30 m. Data collection included capturing scaled images of the tree trunks at close range, delineating sections, and processing them with open-source image software. The resulting equations, based on SH and SD, provided DBH estimates with an RMSE of ±0.027 m, corresponding to an average discrepancy of 2.14% relative to measured DBH values. Furthermore, the equation estimating tree H, derived from these DBH values, exhibited an RMSE of ±1.85 m, equating to a 3.62% discrepancy from the average tree heights observed in this study.