Optical–Biophysical Relationships of Vegetation Spectra without Background Contamination (original) (raw)
Related papers
Http Dx Doi Org 10 5589 M09 040, 2014
The reflectance signal from the forest tree canopy is influenced by the optical properties of the background formed by understory vegetation. Our study shows that the herb-moss layer in the forest tends to be brighter in the visible wavelength when the canopy above is more closed because of the specific properties of plants grown in low light. When leaf-area-based chlorophyll content falls below approximately 150 mg•m-2 , reflectance in the red region of the spectrum increases compared with that of the background. The best descriptors of the herb-moss layer for deriving optical parameters are herb layer dry mass for the visible wavelength range and total aboveground water mass for near-infrared (NIR) reflectance. In addition, chlorophyll content per leaf area considerably improves the red reflectance estimate. Résumé. Le signal de réflectance de la canopée des arbres forestiers est influencé par les propriétés optiques du second plan formé par la végétation des strates inférieures. Notre étude montre que la couche d'herbe/de mousse dans la forêt a tendance à être plus lumineuse dans le spectre visible lorsque la canopée au-dessus est plus fermée. Ceci est dû aux propriétés spécifiques des plantes adaptées à la faible luminosité. Lorsque la teneur en chlorophylle par unité de surface foliaire descend en dessous d'approximativement 150 mg•m-2 , la réflectance dans le spectre rouge augmente par rapport à celle du second plan. Les meilleurs descripteurs de la couche d'herbe/de mousse pour dériver des paramètres optiques sont : la masse sèche de la couche d'herbes pour estimer le spectre visible, et la masse totale d'eau de surface pour la réflectance dans le proche infrarouge (NIR). De plus, connaître la teneur en chlorophylle par surface foliaire améliore considérablement l'estimation de la réflectance dans le spectre rouge.
Review of spectral vegetation indices and methods for estimation of crop biophysical variables
Aerospace Research in Bulgaria, 2017
In present article a brief overview is presented on spectral vegetation indices and methods for estimation of crop main biophysical variables and their proxies. The main VIs used in estimation of nitrogen and chlorophyll, biomass, LAI and fAPAR, fCover, and photosynthesis are summarized. Biophysical variables and vegetation indices A number of techniques have evolved to derive the biophysical variables of vegetation using remote sensing data; these can be grouped into three broad categories: the inversion of radiative transfer models [39], machine learning (for example neural networks) [4] and the use of vegetation Indices. There are generally few ways of deriving the biophysical estimates using empirical or semi-empirical relationships: 1) single regression; 2) stepwise linear regression; 3) partial least squares (PLS) regression; 4) artificial neural networks [12]. Methods based on vegetation indices (VIs) have the benefit of being computationally simple while they are generally less site specific and more universally applicable than the other methods. The performance of the different indices and selected "optimal" wavebands depends on vegetation and land cover type, the variables to be retrieved, sun/view geometry to name but a few [12]. Satellite spectral data has the potential to measure the reflected radiation from many plants, thus making assessment of biophysical variables feasible on canopy level. The regression models relate in situ measurements and VIs. The VIs are mathematical transformations of the original spectral reflectance that are designed to reduce the additive and multiplicative errors associated with atmospheric effects, solar illumination, soil background effects, and sensor viewing geometry [29].
A broad-band leaf chlorophyll vegetation index at the canopy scale
Precision Agriculture, 2008
An assessment of the sensitivity at the canopy scale to leaf chlorophyll concentration of the broad-band chlorophyll vegetation index (CVI) is carried out for a wide range of soils and crops conditions and for different sun zenith angles by the analysis of a large synthetic dataset obtained by using in the direct mode the coupled PROS-PECT ? SAILH leaf and canopy reflectance model. An optimized version (OCVI) of the CVI is proposed. A single correction factor is incorporated in the OCVI algorithm to take into account the different spectral behaviors due to crop and soil types, sensor spectral resolution and scene sun zenith angle. An estimate of the value of the correction factor and of the minimum leaf area index (LAI) value of applicability are given for each considered condition. The results of the analysis of the synthetic dataset indicated that the broad-band CVI index could be used as a leaf chlorophyll estimator for planophile crops in most soil conditions. Results indicated as well that, in principle, a single correction factor incorporated in the OCVI could take into account the different spectral behaviors due to crop and soil types, sensor spectral resolution and scene sun zenith angle.
Estimation of Canopy Photosynthetic and Nonphotosynthetic Components from Spectral Transmittance
Ecology, 2000
Spectral transmittance signatures (expressed as absorbances) were studied as a potential indicator of photosynthetic and nonphotosynthetic contributions to the canopyabsorbed photosynthetically active radiation (PAR). An analytical approach was made under laboratory conditions using synthetic canopies in an integrating sphere. This approach provided the basis for identifying spectral (absorbance-based) features and indices to estimate green (photosynthetic) and nongreen (structural and dead materials) contributions to canopy absorbance. A strong relationship was found between the amplitude of the first derivative of the absorbance (A RE) and green area, while the integrated absorbance in the PAR region (A PAR) mainly responded to variations in total area. The ratio A RE /A PAR was closely correlated to the fraction of photosynthetic area to total area (i.e., the canopy green fraction). Similarly, the ratio and normalized difference of the absorbances at 680 and 900 nm (A SR and A NDVI) closely tracked variations in the canopy green fraction. Subsequently, these indices were tested in field plots with contrasting structural characteristics. Under field conditions, A RE was a good indicator of green biomass. The indices A SR and A NDVI were also reliable indicators of green biomass but were affected by changes in sampling conditions. As in the lab study, A RE /A PAR was a good indicator of canopy green fraction. Thus, ground-based measurements of canopy spectral transmittance provided a tool for determining the photosynthetic contribution to canopy-absorbed PAR by correcting for nonphotosynthetic canopy components. Moreover, A RE showed a strong correlation with conventional vegetation indices derived from spectral reflectance measurements. This technique could be a useful tool for plant ecophysiology studies and a field-validation method for remote-sensing studies.
Evaluation of the relationship between NDVI and LAI in cool-temperate deciduous forest
2005
Leaf area index (LAI) is a key biophysical variable influencing land surface processes such as photosynthesis, transpiration and it is a required input for various ecological models. A cool-temperate deciduous forest in Takayama, central Japan was selected as experimental site for this study. The dominant tree species at the study site are birch (Betula ermanii, Betula platyphylla) and oak (Quercus crispula). In addition, the dominant species of understory is evergreen dwarf bamboo (Sasa senanensis). The moderate resolution imaging spectroradiometer (MODIS) LAI products (MOD15) and ground-measured LAI was compared. Because more than half of MODIS LAI data used in this study was derived from Normalized Difference Vegetation Index (NDVI)-LAI relationship, MODIS NDVI derived from MOD9 reflectance data and ground-measured LAI was also compared. As results of comparison between NDVI derived from MOD9 and ground-measured LAI, the date of beginning of increase in NDVI value is earlier than the date of beginning of leaf expansion detected by ground-measured LAI. Furthermore, the date of beginning of decrease in NDVI value is earlier than the date of beginning of leaf fall detected by ground-measured LAI. In order to understand the relationship between NDVI and LAI in this study area, the relationship between NDVI calculated by scattering from arbitrarily inclined leaves (SAIL) radiative transfer model and LAI was evaluated. This led to two results: (a) understory plant (Sasa senanensis) affected canopy level NDVI during leaf expansion and leaf senescent periods, and (b) the relationships between NDVI and LAI of summer and that of autumn are different because of discoloration of the leaf during leaf senescent period. These results indicate that it is necessary to take into account the influence of understory plant for estimating canopy LAI from NDVI during leaf expansion and leaf senescent periods, and it is also necessary to consider discoloration of the leaf during leaf senescent period.
Canopy spectral invariants for remote sensing and model applications
Remote Sensing of Environment, 2007
The concept of canopy spectral invariants expresses the observation that simple algebraic combinations of leaf and canopy spectral transmittance and reflectance become wavelength independent and determine a small set of canopy structure specific variables. This set includes the canopy interceptance, the recollision and the escape probabilities. These variables specify an accurate relationship between the spectral response of a vegetation canopy to the incident solar radiation at the leaf and the canopy scale and allow for a simple and accurate parameterization for the partitioning of the incoming radiation into canopy transmission, reflection and absorption at any wavelength in the solar spectrum. This paper presents a solid theoretical basis for spectral invariant relationships reported in literature with an emphasis on their accuracies in describing the shortwave radiative properties of the three-dimensional vegetation canopies. The analysis of data on leaf and canopy spectral transmittance and reflectance collected during the international field campaign in Flakaliden, Sweden, June 25 -July 4, 2002 supports the proposed theory. The results presented here are essential to both modeling and remote sensing communities because they allow the separation of the structural and radiometric components of the measured/modeled signal. The canopy spectral invariants offer a simple and accurate parameterization for the shortwave radiation block in many global models of climate, hydrology, biogeochemistry, and ecology. In remote sensing applications, the information content of hyperspectral data can be fully exploited if the wavelength independent variables can be retrieved, for they can be more directly related to structural characteristics of the three dimensional vegetation canopy.
Remote Sensing of Environment, 1995
relationships and the synaptic weights and biases of two-layer neural networks were successfully adjusted on an experimental data set and on data derived from radiative transfer model simulations. Three experimental data sets were acquired over sugar beet canopies. They expressed a large range of canopy architecture and soil background reflectance. Two of them that correspond to similar experimental procedures were merged together and then randomly split into two subsets: one for fitting the parameters of the VI-Po(0) relationships or to train the neural network, and the other for the evaluation of the predictive performances. The third experiment is used as an additional independent data set to test the robustness of the approaches. The modelsimulated data set was generated using the SAIL and PROSPECT radiative transfer models, with input parameter values that were chosen to have similar distributions as observed over sugar beet canopies. As for the experimental data set, the model-simulated data set was split into a training (calibration) and a test (validation) data set. Results show that the gap fraction can be accurately estimated from the red and near-infrared reflectance without any external information except maybe the crop type (here sugar beet) and the soil line characteristics required for some of the vegetation indices. The best predictive performances were observed for the SAVI-like vegetation indices (SAVI, TSAVI, MSAVI) and the poorest for the NDVI, PVI, and GEMI having intermediate although satisfactory results. Neural networks trained on the simulated data set appeared to be the most robust approach. It allows us to implicitly incorporate our knowledge about the physics of the radiative transfer in the interpretation of remote sensing data.
The use of vegetation indices in forested regions: issues of linearity and saturation
IGARSS'97. 1997 IEEE International Geoscience and Remote Sensing Symposium Proceedings. Remote Sensing - A Scientific Vision for Sustainable Development
Numerous problems and difficulties have been reported with the use of vegetation indices in high biomass, forested regions. In this study we analyzed Landsat-Thematic Mapper (TM) scenes from various temperate and tropical forested biomes, representing needleleafand broadleaf canopy structures in the Pacific Northwest (Oregon), Eastern U.S. (Harvard Forest), southern Chile, the Amazon, and Central America. The TM scenes were atmospherically corrected and reduced to MODIS surface reflectance data at 250 m pixel sizes. Various vegetation indices (VIS) were then computed including the normalized difference vegetation index (NDVI), simple ratio, soil-adjusted vegetation index (SAVI), enhanced vegetation index (EVI), and green vegetation index (GVI). The NDVI was also tested utilizing the green and middle-infrared (MIR) bands. All of the NDVIs were non-linear and were fairly saturated across the forested biomes. In contrast, the remaining indices remained sensitive to canopy structure variations over all of the forested biomes with minimal saturation problems. The high 'penetrating' capability of the near-i&a& band through forested canopies was the dominant factor in vegetation index sensitivity and performance. We found that indices with higher weighing coefficients in the "near-i&a&" to be the best approach in extending vegetation index performance over forested and dense vegetated canopies.
Journal of Geophysical Research, 1992
Leaves of the dominant grass species of the First International Satellite Land Surface Climatology Project (ISLSCP) Field Experiment (FIFE) site reflect and transmit radiation in a similar manner to other healthy green leaves. Visible reflectance factors (RFs) and transmittance factors (TFs) were lower for older leaves than younger leaves except during senescence, when RF and TF values were higher. Near-infrared (NIR) RF values increased and TF values decreased with leaf age, with the reverse occurring as the leaf underwent senescence. Leaf optical properties were not found to be dependent on leaf water potential in the range from-0.5 to-3.0 MPa. Canopy bidirectional reflectance factor (BRF) values generally increased with increasing view zenith angle (Ov). Maximum values were in the backscatter direction, whereas BRF values in the visible region were lowest at oblique off-nadir Ov in the forward scatter direction and at or near nadir in the NIR region. Solar principal plane BRF values varied most at large solar zenith angles (Os). Visible and mid-infrared canopy BRF values decreased and NIR BRF values increased with leaf area index (LAI). Soil BRF distributions in the solar principal plane varied slightly with Os and Ov and varied considerably for wet and dry surfaces. Spectral vegetation indices (SVIs) varied with Os and Ov; values were lowest in the backscatter direction and highest in the forward scatter direction. The fraction of absorbed photosynthetically active radiation (APAR) increased with increasing Os. APAR had a strong linear relationship to nadir-derived SVI values but not to oblique off-nadir-derived SVI values. The relatively small dependence of off-nadir SVI values on Os should allow daily APAR values to be estimated from measurements made at any time of the day. Factors which affect reflectance from vegetated surfaces and contribute to the non-Lambertian nature of these surfaces include (1) spectral properties of canopy elements and substrate; (2) the canopy architecture (that is, leaf area index (3) illumination and viewing directions [Norman et al., 1985]. Canopy architecture plays a vital role in determining the BRFs from a vegetative canopy. Leaves are oriented at a variety of inclination angles, thereby varying effective illumination and viewing angles. The result is a complex pattern of reflected and transmitted radiation. Canopy architecture can change due to wind, heliotropism, and water stress.