Multi-sensor airborne lidar requires intercalibration for consistent estimation of light attenuation and plant area density

Vincent Grégoire, Verley Philippe, Brede Benjamin, Delaitre Guillaume, Maurent Eliott, Ball James, Clocher Ilona, Barbier Nicolas. 2023. Multi-sensor airborne lidar requires intercalibration for consistent estimation of light attenuation and plant area density. Remote Sensing of Environment, 286:113442, 14 p.

Liste HCERES des revues (en SHS) : oui

Thème(s) HCERES des revues (en SHS) : Géographie-Aménagement-Urbanisme-Architecture

Résumé : Leaf area is a key structural characteristic of forest canopies because of the role of leaves in controlling many biological and physical processes occurring at the biosphere-atmosphere transition. High pulse density Airborne Laser Scanning (ALS) holds promise to provide spatially resolved and accurate estimates of plant area density (PAD) in forested landscapes, a key step in understanding forest functioning: phenology, carbon uptake, transpiration, radiative balance etc. Inconsistencies between different ALS sensors is a barrier to generating globally harmonised PAD estimates. The basic assumption on which PAD estimation is based is that light attenuation is proportional to vegetation area density. This study shows that the recorded extinction strongly depends on target detectability which is influenced by laser characteristics (power, sensitivity, wavelength). Three different airborne laser scanners were flown over a wet tropical forest at the Paracou research station in French Guiana. Different sensors, flight heights and transmitted power levels were compared. Light attenuation was retrieved with an open source ray-tracing code (http://amapvox.org). Direct comparison revealed marked differences (up-to 25% difference in profile-averaged light attenuation rate and 50% difference at particular heights) that could only be explained by differences in scanner characteristics. We show how bias which may occur under various acquisition conditions can generally be mitigated by a sensor intercalibration. Alignment of light weight lidar attenuation profiles to ALS reference attenuation profiles is not always satisfactory and we discuss what are the likely sources of discrepancies. Neglecting the dependency of apparent light attenuation on scanner properties may lead to biases in estimated vegetation density commensurate to those affecting light attenuation estimates. Applying intercalibration procedures supports estimation of plant area density independent of acquisition characteristics.

Mots-clés Agrovoc : couvert forestier, surface foliaire, peuplement forestier, écologie forestière, télédétection, calibrage, lumière du jour, lidar

Mots-clés géographiques Agrovoc : Guyane française, France

Mots-clés libres : LiDAR, ALS, UAV-LS, PAD, LAD, Forest

Classification Agris : U30 - Méthodes de recherche
K01 - Foresterie - Considérations générales
F40 - Écologie végétale

Champ stratégique Cirad : CTS 5 (2019-) - Territoires

Agences de financement hors UE : Agence Nationale de la Recherche, Centre National d'Etudes Spatiales, European Space Agency

Projets sur financement : (FRA) CEnter of the study of Biodiversity in Amazonia

Auteurs et affiliations

• Vincent Grégoire, IRD (FRA) - auteur correspondant
• Verley Philippe, IRD (FRA)
• Brede Benjamin, Wageningen University (NLD)
• Delaitre Guillaume, Université de Montpellier (FRA)
• Maurent Eliott, CIRAD-ES-UPR Forêts et sociétés (FRA) ORCID: 0000-0002-1939-5677
• Ball James, Université de Montpellier (FRA)
• Clocher Ilona, Université de Montpellier (FRA)
• Barbier Nicolas, IRD (FRA)

Source : Cirad-Agritrop (https://agritrop.cirad.fr/603292/)

Métriques

[ Page générée et mise en cache le 2024-11-26 ]