Post-wildfire vegetation and wildfire prevention: drone and LiDAR surveys to investigate structure, biomass, and vulnerability

The passage of a wildfire modifies the landscape in both the short and long term. The types of tree species, their three-dimensional structure, the quantity of biomass, and post-fire vegetation regrowth directly influence flammability, fire regime and, consequently, territorial vulnerability.

At the beginning of February, a group of researchers and technical staff from CIMA Research Foundation, experts in wildfires and forest biodiversity conservation and specialists in monitoring systems and drones, in collaboration with Gabriele Arcangelo Macellaro from the Geo-Environmental Data Laboratory of the University of Basilicata, carried out field activities in a hilly inland area near Albenga that has been severely affected by wildfires in recent years.

20260204 120507
Fig. 1. GPS acquisition of the planimetric position of plant species

Specifically, the objective of these surveys was to study vegetation structure and condition (biomass, geometric characteristics, and vegetative status) in order to investigate the evolution of the tree and shrub layers following past fires and to highlight their possible future development in relation to potential new fire events. These data are fundamental to improving prevention, risk management and post-fire intervention planning, as well as enabling researchers to define a scalable data acquisition protocol capable of transferring high-definition analyses from local contexts to much larger areas.

Different types of instrumentation were employed: drones equipped with various sensors for aerial data acquisition, ground-based instruments such as TLS (Terrestrial Laser Scanner), in addition to floristic characterization and geolocation of tree species.

Drones and TLS, equipped with LiDAR (Light Detection and Ranging) sensors, are essential for detailed reconstruction of the three-dimensional structure of vegetation. Drone flights also provided spectral reflectance data to analyze the different light bands reflected by vegetation; these data provide information on plant condition and physiological characteristics. Finally, vegetation sampling activities were essential to link sensor-derived data with the actual conditions of the vegetation cover.

drone surveys for post-wildfire vegetation
Fig. 2. Sampling scheme for point cloud acquisition using a terrestrial LiDAR sensor

The study area was selected not only because it had been affected in the past by fires of varying intensity, but also because of its vegetation composition, ranging from Mediterranean shrubland undergoing regrowth to more closed forested areas such as pine stands or mixed forests. The analysis of these environments helps to understand how the landscape changes after a wildfire and which conditions may render an area more or less vulnerable in the future, enabling the quantification of hazard and predisposition to new events, as well as the identification of medium- to long-term management actions.

Picture1
Fig. 3. Digital Terrain Model (DTM) generated from UAV (Unmanned Aerial Vehicle) acquisitions using a photogrammetric sensor. Within the area, plots (small-sized areas) are present and were analyzed using drone-based LiDAR and terrestrial platforms for reconstructing the three-dimensional structure of vegetation.
Picture2
Fig. 4. Fusion of point clouds derived from TLS and UAV-LiDAR. On the left, information from UAV (red) and TLS (white) is visible, while on the right the fused and colorized point cloud is shown. Note that the fusion procedure ensures increased data density and spatial resolution.
Esempio acquisizione
Fig. 5. Example of multispectral UAV acquisition focusing on a single plot. On the left, the RGB (Red, Green, Blue) orthophoto is displayed, while on the right the Normalized Difference Vegetation Index (NDVI) product is shown, a multispectral index used to assess vegetation condition.

Share