Go to top
European Commission logo
RAMI website

Open forest canopies


This set of 3D experiments was suggested to compare the repartitioning of incident solar energy into an Absorbed (A) and a Transmitted (T) flux component given that the surface Reflectance (R) and a complete description of the structurally heterogeneous canopy architectures are known. This being said, we nevertheless encourage models (that are capable to deal explicitly with heterogeneous canopy structures) to be run in forward mode such as to simulate all three radiative fluxes (A, R and T) on the basis of the provided structural and spectral surface properties only. The overall canopy structure for these test cases is reminiscent of open forest canopies (where the randomly oriented foliage is confined to relatively large spherical volumes located at varying heights above the ground). A total of 72 test cases are proposed. Various canopy density, soil brightness and illumination conditions are given both for the visible (VIS) (400-700nm) and near infra-red (NIR) (700-3000nm) spectral ranges. Whether you are using Look-Up-Tables, analytical radiative transfer formulations, or some other form of parameterisation to prescribe/simulate the radiative flux values for these test cases, please stick as closely as possible to the structural, spectral and illumination conditions provided below (in some look-up-table based approaches this may mean that the same estimates will be submitted for multiple test cases). Since all required RAMI4PILPS measurements consist of flux ratios, the actual magnitude of the incident irradiance is irrelevant. The canopy reflectance value (R) that is indicated for each of the various test cases below can be assumed to lie within 1 % of the true value.

RAMI4PILPS represents open forest canopies with heterogeneous leaf canopies composed of a large number of identical, non-overlapping spherical objects - representing the individual plant crowns - located over and only partially covering a horizontal plane standing for the underlying background surface. These spherical objects have a radius of 5m and their centers are located between ~7 and ~11 meters above the background (random height distribution) to yield a maximum canopy height of ~16m. Three different fractional coverages (i.e., number of spheres × maximum cross section of sphere / overall area of scene) are proposed, corresponding to a domain-averaged leaf area index, LAI (i.e., total number of leaves in scene × area of single leaf ⁄ area of scene) value of 0.5, 1.5 and 2.5. Each individual sphere contains a 'cloud' of oriented but dimension-less particles representing the foliage. The leaf area index of a single sphere (area of leaves in sphere ⁄ cross section of sphere) is 5 [m² ⁄ m²]. The foliage elements (scattering particles) are characterized by specified radiative properties (reflectance, transmittance) defined separately for both the visible and near-infrared spectral domains. The orientation of the normals of the foliage elements (scatterers) follows a uniform (or what is sometimes called a spherical) distribution function, i.e., the probability to be intercepted by a leaf is independent of the direction of travel of the radiation (see definitions). The woody area index (WAI) or stem area index (SAI) is equal to zero for all spheres (and scenes). The radiative properties of the underlying background are specified for the visible and near-infrared spectral domains separately. The scattering law of the underlying background is Lambertian. The background albedo is specified for three different brightness levels of which the black soil (perfect absorption) and the snow (very high reflection) cases are perhaps only of scientific interest since they may not actually occur very often in what ecologist would term an open forest environment.

The following figures exhibit a graphical representation of the proposed open forest canopies scenes:

OPEN FOREST scene with LAI=0.50, canopy height=15m, sphere radius=5m
OPEN FOREST scene with LAI=0.50, canopy height=15m, sphere radius=5m
OPEN FOREST scene with LAI=1.50, canopy height=15m, sphere radius=5m
OPEN FOREST scene with LAI=1.50, canopy height=15m, sphere radius=5m
OPEN FOREST scene with LAI=2.50, canopy height=15m, sphere radius=5m
OPEN FOREST scene with LAI=2.50, canopy height=15m, sphere radius=5m

Participants that use Look-Up-Table or other parameterisation schemes should make sure that they represent as close as possible the structural, spectral and illumination conditions for the various canopies presented below. If, on the other hand, RAMI4PILPS participants require a detailed description of the canopy characteristics (in addition to the statistical values provided in the various tables below) they will be able to find a file with the detailed x,y,z locations of the sphere centers for each of the proposed test cases (see links after each table below). These sphere center locations will allow to populate a scene of dimensions 250×250 m2 which is sufficient in size to represent the radiative fluxes of these canopy types in an appropriate manner. When simulating the radiation transfer within a canopy, reconstructed using the sphere centers provided below, participants are encouraged to use the following angular scheme: The zero azimuth line is defined along the northern direction and coincides with the positive x axis as indicated in the diagram below:

The zero azimuth line
The zero azimuth line

Overall three different foliage densities are proposed: sparse (LAI=0.5), medium (LAI=1.5) and dense (LAI=2.5). For each of these structural scenarios, three different soil brightnesses (black, medium ,snow) and four illumination conditions (direct illumination only at 27, 60 and 83 degree zenith, as well as isotropic diffuse illumination only) are specified both for the visible and near-infrared spectral domain. The leaf spectral properties, the foliage orientation, size and scattering properties remain identical for each one of the spectral domains (visible and near-infrared). This should facilitate the batch processing of all the proposed test cases with a shell script using a series of for loops, for example, as follows:

for LAI in LAI_LIST do
for BAND in BAND_LIST do
for SOIL in SOIL_LIST do
endfor (SOIL)
endfor (BAND)
endfor (LAI)

where LAI_LIST stands for the canopy structural properties (LAI=0.5, 1.5 and 2.5), BAND_LIST stands for the visible (VIS) or near-infrared (NIR) spectral domain (which in turn will determine the leaf spectral properties in these domains), SOIL_LIST determines the soil brightness values of the background (black, medium, snow) in each one of the spectral domains, and ILLUMINATION_LIST prescribes the illumination conditions (SZA=27, 60 and 83 degrees, and diffuse isotropic). Last but not least STRUCTURE relates to the structural properties of the canopy (e.g., canopy height, leaf orientation, leaf scattering law, etc.) and remains the same for all 72 test cases.

The following table provides an overview of the variables that change for the various test cases proposed in the open forest canopy category. Individual experiment identifier tags are used to label these test cases e.g., OFC150_MED_UNI_VIS_27. They are needed in the naming of the various measurement results files (see file naming and formatting conventions).