Forward and inverse modelling of canopy directional reflectance using a neural network

This article explores the use of artificial neural networks for both forward and inverse canopy modelling. The forward neural modelling paradigm involved training a network for predicting the bidirectional reflectance distribution function (BRDF) of a canopy given the density of the trees, their height, crown shape, viewing, and illumination geometry. The neural network model was able to predict the BRDF of unseen canopy sites with 90% accuracy. Analysis of the signal captured by the model indicates that the canopy structural parameters, and illumination and viewing geometry, are essential for predicting the BRDF of vegetated surfaces. The inverse neural network model involved learning the underlying relationship between canopy structural parameters and their corresponding bidirectional reflectance. The inversion results show that the R2 between the network predicted canopy parameters and the actual canopy parameters was 0.85 for density and 0.75 for both the crown shape and the height parameters. The res...

[1]  J. Hammersley,et al.  Monte Carlo Methods , 1965 .

[2]  D. Kimes.  Dynamics of directional reflectance factor distributions for vegetation canopies. , 1983, Applied optics.

[3]  A. Strahler,et al.  Geometric-Optical Modeling of a Conifer Forest Canopy , 1985, IEEE Transactions on Geoscience and Remote Sensing.

[4]  N. Goel,et al.  Evaluation of a Canopy Reflectance Model for LAI Estimation through Its Inversion , 1985, IEEE Transactions on Geoscience and Remote Sensing.

[5]  Richard L. Thompson,et al.  Optimal solar/viewing geometry for an accurate estimation of leaf area index and leaf angle distribution from bidirectional canopy reflectance data , 1985 .

[6]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[7]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[8]  Alan H. Strahler,et al.  On the nature of models in remote sensing , 1986 .

[9]  A. Kuusk,et al.  A reflectance model for the homogeneous plant canopy and its inversion , 1989 .

[10]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[11]  R. Myneni,et al.  Radiative transfer in three dimensional leaf canopies , 1990 .

[12]  Tiit Nilson,et al.  A forest canopy reflectance model and a test case , 1991 .

[13]  Darrel L. Williams,et al.  An off-nadir-pointing imaging spectroradiometer for terrestrial ecosystem studies , 1991, IEEE Trans. Geosci. Remote. Sens..

[14]  Ranga B. Myneni,et al.  Photon-Vegetation Interactions , 1991, Springer Berlin Heidelberg.

[15]  Alan H. Strahler,et al.  Geometric-optical bidirectional reflectance modeling of the discrete crown vegetation canopy: effect of crown shape and mutual shadowing , 1992, IEEE Trans. Geosci. Remote. Sens..

[16]  Kamal Sarabandi,et al.  Application of an artificial neural network in canopy scattering inversion , 1994 .

[17]  Richard H. Waring,et al.  Environmental Limits on Net Primary Production and Light‐Use Efficiency Across the Oregon Transect , 1994 .