Forest-biomass contributions are essential for climate regulation because they act as sinks of atmospheric carbon dioxide and determine the water cycle. Climate change may lead to an increase in the frequency of ecosystem fires in regions such as southern Europe. To better understand and prevent the risks associated with forest-fire propagation and intensity, we require detailed information on the structure of the forest canopy. This is also essential for proper management and sustainable use of forest resources and to characterize the evolution of biodiversity.
These considerations are driving development of a new generation of active remote-sensing instruments and methodology. Such tools will contribute significantly to the planning of future satellite missions with payloads including canopy light-detection and ranging systems (lidars) for global surveys of the forest cover. In this framework, a French project (initiated jointly by the Pierre Simon Laplace Institute (IPSL) and the Agricultural and Environmental Engineering Research Institute, CEMAGREF) has led to deployment of a new airborne lidar prototype, LAUVA (lidar aérosol ultraviolet aéroporté: see Figure 1), to study the vertical characteristics of the forest canopy in the Landes region in France. It was originally developed by the French atomic-energy commission (CEA) and the CNRS for atmospheric applications. The system was first deployed for African-monsoon multidisciplinary analysis1 and subsequently adapted for canopy measurements. It is a compact, polyvalent lidar capable of measuring forest-canopy characteristics with unequalled flexibility, both in terms of adaptability of instrumental parameters and flight plan. Like typical spaceborne systems, LAUVA has a large footprint (~2.4m diameter from an altitude of 300m). By recording the full waveform of every laser return, the lidar simultaneously samples both the entire tree structure and the ground echo. Using a UV wavelength of 355nm enables eye-safe emission of energetic laser pulses (16mJ power at a frequency of 20Hz). In addition to the lidar and georeferencing instruments, the ultralight-airplace payload contains two cameras operating in three bands (UV, visible, and near-IR) to stereoscopically map the 3D canopy structure.
Figure 1. (Top) Ultralight airplane preparing for take-off. (a) LAUVA (lidar aérosol ultraviolet aéroporté) canopy lidar (black device on the pilot's left, shown enlarged in the inset). (b) Two three-band cameras (yellow box on the pilot's right). (c) Mapping the Landes forest (France).
We statistically analyzed the vertical structure of the vegetation (including tree height and crowns, bushes, and undergrowth) of forest regions. We sampled several areas of approximately 500×1000m2 composed primarily of maritime pines spanning a range of ages: see Figure 2(a). We performed multiple passes over the regions of interest—see Figure 2(b)—at a horizontal speed of ~70m/s (providing one image every ~1.5m) and with an initial vertical resolution of 1.5m (using a laser-pulse duration of 5ns). We analyzed every lidar return by assigning ‘ground level’ to the last lidar echo exceeding a given threshold. We subsequently georeferenced the measurements to reconstruct the horizontal canopy structure (see Figure 3). The retrieved distribution shows a degree of heterogeneity associated with the origin and age of the trees. Planted trees are tallest, at ~25m: see Figure 3(a) at ~1200 and ~900m along the abscissa and ordinate, respectively. Natural trees are shorter (~18m, to the south) and the shortest trees (~12m) correspond to a young plot (>2000m along the abscissa). The undergrowth is lower for young trees and located close to their bases.
Figure 2. (a) Satellite view of a forest region composed of maritime pines. (b) Sampling of the plot with the nadir-pointing LAUVA flying at an altitude of 300m.
Figure 3. Vertical vegetation distribution of an area in the Landes forest retrieved from the LAUVA measurements: (a) tree tops, (b) maximum lidar return from the tree crowns, (c) bases of the tree crowns, and (d) undergrowth.
We are planning new experiments in 2009–2010, funded by the National Center for Spatial Studies (CNES), to sample different types of forest cover (such as leafs and conifers) and optimize the instrument and the associated methodology to develop a multifunctional tool for measuring the 3D distributions of both the forest canopy and atmospheric components.
Laboratory of Dynamical Meteorology (LMD)
Laboratory of Atmospheric, Environmental, and Spatial Observations Research (LATMOS)
Patrick Chazette, Joseph Sanak
Laboratory of Climate and Environmental Science (LSCE)
CEA/Directorate of Materials Science (DSM)
Tristan Allouis, Sylvie Durrieu
Pascal Genau, Cyrille Flamant
Pierre H. Flamant