Support of Forest Inventory Data Collection by Citizen Scientists

Thiel, Christian GND; Klan, Friederike GND; Pathe, Carsten; Schmullius, Christiane GND; Baade, Jussi GND

Precise forest inventory data are requested by a wide range of users such as scientists, politicians, administrators, forest owners, or the forest industry. One forest inventory parameter of great importance is the forest stem volume (or growing stock volume, GSV). On the one hand, GSV is related to the monetary value of a forest. On the other hand, the amount of bound carbon can be estimated based on GSV. For the determination of the GSV the stem diameter (usually diameter at breast height, DBH), the tree height, the number of trees per unit area, and a species and forest stand specific form factor are required. In forestry, sample based approaches are used to gather these parameters. For minimizing effort and expense, the number and dimensions of these samples are small compared to the total forest area. Also, the repeat time between two inventories is rather large (in the order of ten years). Accordingly, relative GSV errors of approximately 20% have to be accepted. There exists a great interest to minimize both, effort and inventory errors. Precise inventory data are of particular interest in the research domain. For instance, satellite based methods aiming at GSV estimation suffer from inaccurate reference measurements, as the inventory errors propagate to the final satellite based estimates. Airborne light detection and ranging data (LiDAR) can be utilized to detect single trees and to measure the corresponding tree heights with sufficient accuracy for forestry applications. In some Scandinavian countries forest inventories are supported by LiDAR campaigns by default. Moreover, most European countries execute regular and country-wide LiDAR acquisitions, thus LiDAR based tree height measurements could be achieved. For instance, the LiDAR campaign repetition rate in Germany is five years. However, the stem diameter cannot be measured using airborne LiDAR data. Although some technical ground- and low altitude airborne solutions have been proposed, currently the most efficient approach is manual DBH measurement. The simplicity of DBH measurements makes this task an excellent citizen science exercise. To assess the achievable DBH measurement precision, an experiment involving students of a secondary school was carried out in late 2017. The test site “Roda Forest” is located 20 km in the Southeast of Jena. The selected stand is dominated by pine with an age of 60 years. The reference data for the experiment was generated by means of a terrestrial laser scanner (TLS). Based on the TLS data the precise location and the GSV of approximately 200 trees were delineated. The students were equipped with a smartphone application to localize the single trees. During the campaign the circumference of approximately 100 trees was determined using simple measuring tape. These measurements were converted to DBH after the field campaign. The measured DBH varied between 7 cm and 38 cm. In overall, TLS-based and student campaign based measurements were in great agreement (R² = 0.98). Nevertheless, the identification of the correct trees by the students during the campaign was challenging, which was related to general orientation difficulties and a weak GPS signal underneath the forest canopy. This resulted in a remarkable offset between GPS-based and real coordinates. Forthcoming campaigns have to deal with this issue. One option we will explore in the future is the absolute calibration of the GPS signal using checkpoints with precise coordinates.


Citation style:

Thiel, Christian / Klan, Friederike / Pathe, Carsten / et al: Support of Forest Inventory Data Collection by Citizen Scientists. Jena 2018.

Access Statistic

Last 12 Month:

open graphic


Use and reproduction: