Field-Acquired DBH Measurements of Trees in Durango, Mexico, using Terrestrial Laser Scanning (TLS) and Tree Caliper

Introduction
The data generated in this study provide a valuable resource for developing and testing algorithms focused on circumference fitting and diameter estimation of tree stems. By combining field-measured diameters and high-resolution terrestrial laser scanning (TLS) data, the dataset includes precise point clouds representing individual trees, with diameter at breast height (DBH) measured at 1.30 meters above ground level. This standardized height ensures consistency in diameter measurements, a crucial metric in forestry and ecological studies. The dataset supports advancements in remote sensing applications, offering researchers an opportunity to refine methods for accurately determining DBH and enhancing the reliability of biomass and carbon estimation models.

Study Area
The research was conducted in Mexico within a permanent forest research plot, established based on the methodology outlined by Corral-Rivas et al. The plot covers a quadrangular area of 625 m² and is situated in the Sierra Madre Occidental region, specifically within the "La Victoria" management unit in the municipality of Pueblo Nuevo, Durango, Mexico. This area experiences a temperate climate, with average annual temperatures between 20 and 22 °C and rainfall ranging from 800 to 1200 mm per year. The dominant vegetation type is a coniferous forest, primarily consisting of Pinus cooperi, with an estimated tree density of 960 trees per hectare.

Field Data
In this study, we focused on 50 trees with diameters exceeding 10 cm, a commonly used threshold in forest inventory protocols for estimating biomass and carbon, chosen to enhance accuracy and consistency. According to Hoover and Smith, excluding smaller trees generally has a negligible effect on biomass estimates across most forest types, which supports the practicality of this 10 cm cutoff. This threshold also helps eliminate saplings, which can add variability due to their inconsistent growth patterns, and aligns with standard inventory practices. We conducted a conventional inventory to measure DBH for each tree, using a Häglof caliper and taking two measurements at perpendicular angles to achieve a reliable average (data_field.xlsx).

Data Capture and Acquisition via Terrestrial Laser Scanning
The terrestrial laser scanning data were collected using a FARO Focus M70 scanner, capable of measuring distances up to 70 meters with an accuracy of ±3 mm. The scanner was set to an "exterior" profile with a resolution of 1/4 and quality level of 4x, achieving an average point density of 234,679 points per square meter and a resolution of 10,310 x 4,268 points. Before starting the scans, ten targets were strategically positioned on-site to assist with alignment during post-processing. Four scans were then performed, with a 180-degree vertical angle and a 360-degree horizontal angle, capturing a complete view of the surrounding environment. The scans were merged using FARO Scene software to ensure accurate alignment. All details regarding the scanner and tree positions can be found in Appendix A.

Extraction of 2D Planes from Individual Tree Stem Point Clouds
The point cloud was initially loaded into CloudCompare, where selected trees were manually segmented, and ground points were removed to minimize slope effects. Each tree was represented as a series of n points in three-dimensional space:

(x_1, y_1, z_1), (x_2, y_2, z_2), ..., (x_i, y_i, z_i), ..., (x_n, y_n, z_n)

For each tree stem, a specific cylindrical section was isolated by selecting points with z-coordinates between 1.25 m and 1.35 m, filtering out points outside this range. This method enabled the extraction of a segment corresponding to the diameter at breast height (DBH).

Following this step, the z-coordinates within the selected cylindrical sections were excluded, and duplicate points were removed, resulting in a two-dimensional dataset. This reduction in dimensionality produced a new collection of n points:

(x_1,y_1),(x_2,y_2),…,(x_i,y_i),…,(x_n,y_n).

The information about the planes and their representation can be found in tls_tree_discs.zip and tls_tree_imgs.zip.