Šumarski list, Vol. 144 No. 5-6, 2020.
Izvorni znanstveni članak
https://doi.org/10.31298/sl.144.5-6.1
Influence of a measurer and measurement errors on measurement of a tree diameter
Mario Božić
; Šumarski fakultet Sveučilišta u Zagrebu
Filip Đureta
; Đelekovac
Ernest Goršić
; Šumarski fakultet Sveučilišta u Zagrebu
Mislav Vedriš
; Šumarski fakultet Sveučilišta u Zagrebu
Sažetak
A tree diameter at breast height (dbh) is a basic variable measured in forest inventory. Generally, it is measured with calipers whose arms have to be perpendicular to the beam. Therefore, caliper has to be rectified before measurement. Measurement should be performed in the way that caliper itself is perpendicular to the tree stem with arms and beam touching the tree without applying too much pressure to caliper arms. Measurer mistakes and failing to follow standard measurement procedures lead to errors that reflect in all inventory results (basal area, stand volume). Experience gathered while working with students and colleagues in practice, showed that measurement procedures sometimes significantly depart from required. Therefore, the aim of this research was to quantify and analyze known and observed dbh measurement errors.
For that purpose two sites in Zagreb Training and Forest research center were selected: one site in area of lowland oak and hornbeam forest stand and another in mountain area with mixed fir and beech forest. Total 282 trees (74 Pedunculate oak, 56 Common hornbeam, 76 Silver fir and 76 Common beech) in wide diameter range were measured. (Table 1). All selected trees were marked with a number before measurement.
On both research plots three measurers performed following diameter measurements using Haglof Mantax caliper: crosswise diameters at self-estimated breast height with (1a, 1b), crosswise diameters at marked breast height (2a, 2b), 10 cm above (3a) and 10 cm under the marked breast height (3b), at marked breast height point with caliper beam held apart from tree and normal pressure on the caliper arms (4a) and higher pressure on the caliper arms (4b) and caliper beam on marked point with caliper tips facing up (5a). Measurement group consisted of two measurers. One measurer preformed measurement and other was writing the data. Since measurers height can affect the placement of the calipers on the tree, so position of breast height of 1,3m was determined using measuring tape before beginning of measurement for each measurer. Marking of breast height point on the tree was made after measurement with self-estimated breast height. Collected data were analyzed using t-test and ANOVA with 0,05 significance level.
Results show that average differences between two crosswise diameter measurements at self estimated breast height are statistically significant for hornbeam (average difference 1 cm) and beech (average 1,7 cm). Fir and oak had less differences (up to 0,7cm) that were not proven statistically significant. (Table 2).
At marked breast height differences of crosswise measurements are as well statistically significant for hornbeam and beech. (Table 3). Comparison between average diameter difference at self-estimated
and marked breast height show statistically significant difference for oak and beech. Negative average values for all measured species suggest that measurers usually slightly overestimated breast height (Table 4). According to Figure 1 it is visible that crosswise diameter differences are mostly within ±2-3 cm (50% difference). Certain trees show difference up to 15 cm in crosswise diameters. At the same time, diameter differences between self-estimated and marked breast height are smaller than crosswise diameter difference. Figure 2 shows that there is almost no trees which are not elliptical with fir having the smallest values.
Diameter measured 10 cm under marked breast height expectedly resulted in statistically significant higher values for all tree species with 0,49 cm on average (Figure 4), and diameters measured 10 cm above marked breast height resulted in lower values being statistically significant for oak, beech and fir (Figure 5). The biggest difference is for oak trees.
When caliper beam was held apart from tree and normal pressure on caliper arms applied, results were -0,12 to -0,26 cm lower and statistically significant compared to readings at correct position (caliper beam on the tree). The similar results were achieved with stronger pressure on the arms but with higher negative values, as expected.
Measurement with arm tips facing up resulted with statistically significant lower average diameter reading for all measured tree species (Table 9). Moreover, Figure 4 shows that in comparison to diameter measured on marked breast height readings on the caliper scale were in range between -1,5 to +0,5cm with extremes ranging from -4 to +2cm respectively.
Differences between measurers were observed on self-estimated breast height and for crosswise diameters on marked breast height. Table 10 and Figure 5 show that diameter differences between measurer 1 and other two measurers have proven to be statistically significant for self-estimated breast height of fir and beech trees (forest stand on sloping ground) where measurer 1 in average measured smaller diameters. Further, with diameters measured on marked breast height, differences were significant between measurer 3 and other two measurers in the way that measurer 3 on average measured larger diameters (Table 11 and Figure 6)
Diameter measurement is the basis for volume estimation of individual trees and stand as well. Therefore, measurement errors have impact on volume estimation, so measurement is subjected to control and sometimes allowable measurement departures are prescribed. Omule (1980) mentions that in Columbia Forest Service allowable errors of 1% for dbh is allowed and Melson et al. (2002) states that allowable dbh errors are ±0,25 cm for trees from 12,5 to 50 cm, ±0,5 cm for trees from 50-100 cm and 0,75 cm for treesfrom 100-150 cm dbh. Measurer responsibility is to perform the measurement according to the rules and with highest possible precision.
As the results show, differences in diameters can be a result of eliptical trees (Table 2, 3 and Figure 1). On tree level, differences sometimes reach 30% of tree diameter (Figure 1). With increase of tree diameter the difference between two crosswise diameters increases as well (Figure 3). Also, measurement under or above actual beast height results in statistically significant positive or negative diameter values.
Results based on measured sample trees show that hornbeam and beech breast height measurement gives a significant difference in crosswise diameters as a result of sloping ground and geographic orientation. Turning the caliper beam towards the plot center can cancel out these measurement errors assuming an adequate sample size. Furthermore, on trees that are to be cut, the breast height should be marked (on the upper side of the slope). In this way measurement error for a single tree level would be reduced. Finally, this research points out the importance of following correct procedures while tree diameter measurement in order to minimize measurement errors which can affect accuracy of measured data.
Ključne riječi
diameter at breast height; measurement errors; bias; forest inventory
Hrčak ID:
240113
URI
Datum izdavanja:
30.6.2020.
Posjeta: 2.038 *