2026-10-14T15:14:16Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33839655 2026-09-16T10:16:30Z category_272 category_873 category_39 category_734 category_931 category_133 portal_87 item_type_3 month_year_09_2026

A hidden variables approach: assessing the influence of environmental, flight and processing parameters on LiDAR accuracy in wet heathlands in the northwest of Spain Patricia Rodriguez Dorribo (24973338) Ramón Alberto Díaz-Varela (24973341) Environmental Sciences not elsewhere classified Chemical Sciences not elsewhere classified Ecology Biological Sciences not elsewhere classified Information Systems not elsewhere classified Plant Biology LiDAR 3D canopy parameters digital models accuracy random forest, noninvasive monitoring <p>Vegetation height is an important parameter in botany, ecology, and forest management, serving as a key indicator in wet heathland conservation status. To estimate structural attributes, light detection and ranging (LiDAR) is widely used; however, limited attention has been given to how environmental, processing, and flight-related factors influence these estimations in dense shrublands, or how LiDAR-derived digital models compare close-range field measurements. To address this gap, we conducted a study in the priority habitat 4020* (Annex I, Council Directive 92/43/ECC) in Northwest of Spain, employing the random forest (RF) algorithm to model the error between methods and assess the influence of these parameters. Results show that model performance improves when incorporating variables such as maximum vegetation height, CHM spatial resolution, mode, flight altitude and slope (R2 = 0.94 and an RMSE = 0.04 cm). Notably, maximum height alone explains 75% of the variability, with increases of 7 cm inducing changes of ±40 cm. Moreover, although interactions among parameters had limited impact (between 2 cm and 2 mm, except for slope ±15 cm), they provide insights into the strengths and limitations of LiDAR-derived canopy height models (CHMs) in challenging environments (dense cover and steep areas). Furthermore, flying at a lower altitude, non-repetitive mode, and lower spatial resolutions (DTM 1 m, CHM 0.5 m) were associated with higher vegetation model contribution on average. Errors also tend to increase in areas where taller vegetation is associated with higher canopy cover, while reduced point density in flight further exacerbated inaccuracies. This approach provides a practical framework for improving precision and managing uncertainties in structurally complex ecosystems.</p> 2026-09-16T10:16:30Z Dataset Dataset 10.6084/m9.figshare.33839655.v1 https://figshare.com/articles/dataset/A_hidden_variables_approach_assessing_the_influence_of_environmental_flight_and_processing_parameters_on_LiDAR_accuracy_in_wet_heathlands_in_the_northwest_of_Spain/33839655 CC BY 4.0