Mihai A Tanase, Juan Pablo Martini, Daniel García García, M. Zavala, P. Ruiz‐Benito
2026.2.25Annals of Forest Research
Abstract
Monitoring forest structure across large and heterogeneous landscapes is essential to understanding ecosystem dynamics, carbon stocks, and the impacts of global change. This study leverages lidar surveys from Spain's national airborne laser scanning (ALS) program to produce harmonized, high-resolution maps of canopy height, canopy cover, and aboveground biomass (AGB) across peninsular Spain over two periods: 2008-2016 and 2015-2021. We developed a consistent processing workflow to overcome challenges posed by heterogeneous lidar acquisitions, and integrate the resulting metrics with National Forest Inventory (NFI) data to model AGB using machine learning algorithms. Validation against field data demonstrated high accuracy for canopy height (R² > 0.8; RMSE < 3 m in some regions) and acceptable performance for AGB estimation (RMSE: 23–44 t ha⁻¹). Between the two lidar surveys, forests exhibited structural development in forest canopies, with average annual increases of 0.2 m in height and 0.9% in canopy cover, particularly pronounced in the Atlantic regions. Species-level analysis highlighted the structural and functional contrast between biomes: high-biomass Atlantic species such as Fagus sylvatica and Pinus radiata reached up to 188 t ha⁻¹ and 126 t ha⁻¹ AGB, respectively, while Mediterranean species like Quercus ilex and P. halepensis remained under 35 t ha⁻¹. Comparisons with global and regional satellite-derived products revealed that airborne lidar offers superior spatial detail and accuracy, especially in structurally complex forests. ALS-based height estimates showed significantly lower mean absolute error (MAE: 1.6–3.7 m) than global products (MAE: 2.9–7.8 m), and the GEDI dataset consistently overestimated canopy height, 5.9 m on average, across the peninsular Spain. This work highlights the critical role of harmonized lidar datasets for robust forest monitoring, and provides a valuable baseline for future ecological assessments, carbon accounting, and validation of satellite products across the peninsular Spain.
Citation format
TANASE, Mihai A, et al. Forest structure and its changes from multi-temporallidar data: A homogeneously derived database forpeninsular spain. Annals of Forest Research, 2026, 69(1): 3–20.