Table 3 - uploaded by Luis Alonso
Content may be subject to copyright.
Source publication
The SPARC campaign has been organized in coincidence of CHRIS/Proba multi-angular and hyper- spectral acquisition over the agricultural test site of Barrax in Spain. Radiometric and biophysical vegetation parameters measurements have been carried out for different crops in both field and laboratory. The aim of this preliminary study is to assess th...
Context in source publication
Context 1
... this study, no particular distribution function for the vegetation parameters was assumed in order to sample the parameters space. They were sampled uniformly according to Table 3. C w was fixed since the leaf water content doesn't affect the spectrum in the 400÷1000 nm range and 5 different leaf inclination distribution, corresponding to 5 different plant architecture (Planophile, Plagiophile, Extremophile, Erectophile, Spherical) were taken into consideration. ...
Citations
... where ET 0 is the reference evapotranspiration (cm), k is the constant for the radiation extinction by canopy and LAI is the leaf area index adapted from data obtained by the study of Antonopoulos [27] for maize, [28] for cotton and [29] for alfalfa. Root growth was modelled using the Verhulst-Pearl logistic growth function [Eq. ...
Accurate estimation of the hydrological features in the unsaturated zone is mandatory for the effective planning of irrigation strategies. Irrigation scheduling depends on crop and soil type as well as climatic characteristics and is usually empirically conducted. This paper simulates the water flow in order to model the soil water balance in three agricultural fields (maize, cotton, alfalfa) located in the River Strymonas basin using the HYDRUS-1D model. The model is fed with meteorological data, soil data and soil moisture measurements. After the calibration, through HYDRUS-1D’s inverse solution, model results were used to evaluate the irrigation activities applied in the pilot application fields in terms of irrigation dose, irrigation interval and soil moisture variation for the cultivation period. In addition, in order to measure the efficiency of the irrigation method evaluated in this work, water productivities for all three fields were compared with productivities yielded from similar applications and experiments as well as precision irrigation experiments found around the world at similar climates with the one at Nigrita.
... However, for LAI and CCC obtained from LUTreg, the accuracies were considerably improved (R 2 = 0.77 and NRMSE = 9.18% for LAI; R 2 = 0.62 and NRMSE = 12.16% for CCC) rather than LUTstd (R 2 = 0.61 and NRMSE = 14.45% for LAI; R 2 = 0.46 and NRMSE = 18.28% for CCC). LUTreg underestimated the high values of LAI (above 4) due to saturation; this result is in line with previous studies [79,89,90]. The SLC model, as an extension of PROSAIL, did not take the row-structure of the potato crop into account; therefore, underestimation often occurred. ...
Hyperspectral cameras onboard unmanned aerial vehicles (UAVs) have recently emerged for monitoring crop traits at the sub-field scale. Different physical, statistical, and hybrid methods for crop trait retrieval have been developed. However, spectra collected from UAVs can be confounded by various issues, including illumination variation throughout the crop growing season, the effect of which on the retrieval performance is not well understood at present. In this study, four retrieval methods are compared, in terms of retrieving the leaf area index (LAI), fractional vegetation cover (fCover), and canopy chlorophyll content (CCC) of potato plants over an agricultural field for six dates during the growing season. We analyzed: (1) The standard look-up table method (LUTstd), (2) an improved (regularized) LUT method that involves variable correlation (LUTreg), (3) hybrid methods, and (4) random forest regression without (RF) and with (RFexp) the exposure time as an additional explanatory variable. The Soil–Leaf–Canopy (SLC) model was used in association with the LUT-based inversion and hybrid methods, while the statistical modelling methods (RF and RFexp) relied entirely on in situ data. The results revealed that RFexp was the best-performing method, yielding the highest accuracies, in terms of the normalized root mean square error (NRMSE), for LAI (5.36%), fCover (5.87%), and CCC (15.01%). RFexp was able to reduce the effects of illumination variability and cloud shadows. LUTreg outperformed the other two retrieval methods (hybrid methods and LUTstd), with an NRMSE of 9.18% for LAI, 10.46% for fCover, and 12.16% for CCC. Conversely, LUTreg led to lower accuracies than those derived from RF for LAI (5.51%) and for fCover (6.23%), but not for CCC (16.21%). Therefore, the machine learning approaches—in particular, RF—appear to be the most promising retrieval methods for application to UAV-based hyperspectral data.
... However, for LAI and CCC obtained from LUTreg, the accuracies were considerably improved (R 2 = 0.77 and NRMSE = 9.18% for LAI; R 2 = 0.62 and NRMSE = 12.16% for CCC) rather than LUTstd (R 2 = 0.61 and NRMSE = 14.45% for LAI; R 2 = 0.46 and NRMSE = 18.28% for CCC). LUTreg underestimated the high values of LAI (above 4) due to saturation; this result is in line with previous studies [79,89,90]. The SLC model, as an extension of PROSAIL, did not take the row-structure of the potato crop into account; therefore, underestimation often occurred. ...
Hyperspectral cameras onboard unmanned aerial vehicles (UAVs) have recently emerged for monitoring crop traits at the sub-field scale. Different physical, statistical, and hybrid methods for crop trait retrieval have been developed. However, spectra collected from UAVs can be confounded by various issues, including illumination variation throughout the crop growing season, the effect of which on the retrieval performance is not well understood at present. In this study, four retrieval methods are compared, in terms of retrieving the leaf area index (LAI), fractional vegetation cover (fCover), and canopy chlorophyll content (CCC) of potato plants over an agricultural field for six dates during the growing season. We analyzed: (1) The standard look-up table method (LUTstd), (2) an improved (regularized) LUT method that involves variable correlation (LUTreg), (3) hybrid methods, and (4) random forest regression without (RF) and with (RFexp) the exposure time as an additional explanatory variable. The Soil–Leaf–Canopy (SLC) model was used in association with the LUT-based inversion and hybrid methods, while the statistical modelling methods (RF and RFexp) relied entirely on in situ data. The results revealed that RFexp was the best-performing method, yielding the highest accuracies, in terms of the normalized root mean square error (NRMSE), for LAI (5.36%), fCover (5.87%), and CCC (15.01%). RFexp was able to reduce the effects of illumination variability and cloud shadows. LUTreg outperformed the other two retrieval methods (hybrid methods and LUTstd), with an NRMSE of 9.18% for LAI, 10.46% for fCover, and 12.16% for CCC. Conversely, LUTreg led to lower accuracies than those derived from RF for LAI (5.51%) and for fCover (6.23%), but not for CCC (16.21%). Therefore, the machine learning approaches—in particular, RF—appear to be the most promising retrieval methods for application to UAV-based hyperspectral data.
... The campaign was realized near Barrax in the Spanish region La Mancha in July 2003. The test site is characterized by flat topography and irrigated fields of alfalfa, maize, potato, winter wheat, sugarbeet, garlic, onion and papaver (D'Urso et al. 2004). Airborne hyperspectral images of the Barrax test site were captured with the HyMap sensor in four flight strips A, B, C & D, from which we merged B and C into one image to contain all the sampling spots of ground data (Fig. 2). ...
... Very good results were instead obtained for homogeneous crops like alfalfa, and also for maize which was already found to be well suited for estimation based on PROSAIL in previous studies (Danner et al. 2019). D'Urso et al. (2004 used a LUT-based inversion approach and were able to predict Barrax LAI of alfalfa with RMSE 0.3 m 2 m − 2 (our study: RMSE = 0.24 m 2 m − 2 ) and potato with RMSE = 0.4 m 2 m − 2 (our study: RMSE = 1.6 m 2 m − 2 ). In contrast to the study of D'Urso et al. (2004), we neither constrained PROSAIL inputs to the conditions at Barrax, nor u ...
With an upcoming unprecedented stream of imaging spectroscopy data, there is a rising need for tools and software applications exploiting the spectral possibilities to extract relevant information on an operational basis. In this study, we investigate the potential of a scientific processor designed to quantify biophysical and biochemical crop traits from spectroscopic imagery of the upcoming Environmental Mapping and Analysis Program (EnMAP) satellite. Said processor relies on a hybrid retrieval workflow executing pre-trained machine learning regression models fast and efficiently based on training data from a lookup table of synthetic vegetation spectra and their associated parameterization of the well-known radiative transfer model (RTM) PROSAIL. The established models provide spatial information about leaf area index (LAI), average leaf inclination angle (ALIA), leaf chlorophyll content (Cab) and leaf mass per area (Cm). In contrast to using site-specific training data, the approach facilitates a universal application without the need to integrate a priori information into the processor. Four machine learning algorithms, namely artificial neural networks (ANN), random forest regression (RFR), support vector machine regression (SVR), and Gaussian process regression (GPR), were found to estimate biophysical and biochemical variables of unseen targets with high performance (relative error scores < 10%). ANNs excelled in terms of accuracy, model size and execution time when the 242 spectral bands were transformed into 15 principal components, the signals of which were scaled by a z-transformation. Validation using in situ data from the SPARC03 Barrax campaign dataset revealed an overall good estimation of measured functional traits, for instance for LAI with root mean squared error (RMSE) of 0.81 m² m⁻², and for Cab RMSE of 6.2 µg cm⁻² with the ANN model. Moreover, both crop traits could be successfully mapped using a pseudo-EnMAP scene revealing plausible within-field patterns. Conformity with LAI output of the SNAP biophysical processor was found especially for grassland and maize in the vegetative stages. Based on these findings, ANN models are considered the best choice for implementation of a hybrid retrieval workflow within the context of operational agricultural crop traits monitoring from future satellite imaging spectroscopy.
... On the other hand, for the plots that relatively covered the soil, LUTreg retained the high values of LAI overestimation (above 3), which are poorly estimated by a 1D turbid medium RTM (i.e., SLC) as a result of the saturation effect [20,59]. This result is consistent with other studies that have used the PROSAIL model and overestimated high LAI values (above 3.5) [37,39,68]. ...
Look-up table (LUT)-based canopy reflectance models are considered robust methods to estimate vegetation attributes from remotely sensed data. However, the LUT inversion approach is sensitive to measurements and model uncertainties, which raise the ill-posed inverse problem. Therefore, regularization options are needed to mitigate this problem and reduce the uncertainties of estimates. In this study, we introduce a new method to regularize the LUT inversion approach to improve the accuracy of biophysical parameters (leaf area index (LAI) and fractional vegetation cover (fCover)). This was achieved by incorporating known variable correlations that existed at the test site into the LUT approach to correlate the model variables of the Soil–Leaf–Canopy (SLC) model using the Cholesky decomposition algorithm. The retrievals of 27 potato plots obtained from the regularized LUT (LUTreg) were compared with the standard LUT (LUTstd), which did not consider variable correlations. Different solutions from both types of LUTs (LUTreg and LUTstd) were utilized to improve the quality of the model outputs. Results indicate that the present method improved the accuracy of LAI estimation, with the coefficient of determination R2 = 0.74 and normalized root-mean-square error NRMSE = 24.45% in LUTreg, compared with R2 = 0.71 and NRMSE = 25.57% in LUTstd. In addition, the variability of LAI decreased in LUTreg (5.10) compared with that in LUTstd (12.10). Hence, our results give new insight into the impact of adding the correlation between variables to the LUT inversion approach to improve the accuracy of estimations. In this study, only two correlated variables (LAI and fCover) were examined; in subsequent studies, the full correlation matrix based on the Cholesky algorithm should be explored.
... Rapid developments in remote sensing technologies over the last two decades inspired scientists to probe into the relationships between structural parameters of a vegetation canopy and multi-angular remote sensing data. In particular, the combination of multi-angular and hyperspectral data is very useful for retrieving canopy structure (Lewis, Barnsley, and Cutter 2001;Urso et al. 2004;Bach et al. 2005;Begiebing et al. 2005;Schlerf and Hill 2005;Rautiainen et al. 2008;Simic and Chen 2008;Vuolo, Dini, and D'Urso 2008;Zhang et al. 2008). The improved multiangle-based LAI was found to have a significant impact on chlorophyll content retrieval in the studies of Simic et al. (2008) and Simic, Chen, and Noland (2011). ...
The impact of structural parameters of agricultural crops on the retrieval of chlorophyll content presents a real challenge for the remote-sensing community. Canopy reflectance can differ between crops of different canopy structure even when they have the same canopy chlorophyll content. Thus, structural properties should be incorporated in chlorophyll mapping to reduce modelling errors. The empirical relationships between vegetation indices and chlorophyll content are well established and commonly used in precision agriculture. Recent advances in using unmanned aerial vehicle (UAV/drone) technology offer successful retrieval of crop structural and biochemical parameters. However, transfer of empirical algorithms derived from satellite to UAV-based analyses introduces new challenges mainly due to fine spatial resolution and details such as crop rows and between- and within-canopy gaps that are more pronounced in UAV images. There are two components of the analysis in this study. The first part is related to heterogeneity of leaf area index (LAI) and chlorophyll content of corn under four agricultural treatments (conventional ploughed, conventional with no tilling, biological with reduced chemical inputs, and certified organic) at the Kellogg Biological Station Long-Term Ecological Research (KBS LTER) site in Michigan, USA. The second part examines the necessity and importance of LAI in chlorophyll mapping using UAV images collected over the heterogeneous KBS LTER parcels at peak growing season. The UAV-derived normalized difference red edge index (NDRE) is found to be highly correlated with canopy chlorophyll, calculated as a product of leaf chlorophyll content and LAI. The coefficient of determination changes from R² = 0.177 to R² = 0.774 when LAI is added to the empirical model. NDRE is also found to be highly correlated with LAI (R² = 0.620). The findings suggest that the conventional corn treatment with no-tilled soil exhibits the highest crop vigour during the peak growing season. The herbicide management applied earlier in the season may have a strong effect on weeds, reducing the crop–weeds competition for nutrients.
... State variables can be defined as the main variables governing radiative transfer processes within the canopy (Verstraete et al., 1996). For instance; combining multi-temporal data from identical sensor (Hagolle et al., 2005;Koetz, Kneubuehler, Huber, Schopfer, & Baret, 2007;Lauvernet, Baret, Hascoët, Buis, & Le Dimet, 2008;Menenti et al., 2005;Yang et al., 2006), synthetic data (Verhoef, 2007), a synergy of LIDAR data with spectral measurements (Kötz, 2006), combining hyperspectral and synthetic aperture radar (Treuhaft, Asner, Law, & Van Tuyl, 2002;Treuhaft, Law, & Asner, 2004), integrating spectral and directional measurements (D'Urso, Dini, Vuolo, Alonso, & Guanter, 2004;Widlowski et al., 2004) and a synergy of an ensemble of imaging instruments (Geiger et al., 2004;Jin et al., 2002;Samain et al., 2006). Although the retrieval of vegetation state variables from these multi-sensor approaches represents a promising advance in improving the quality of the retrieved products, the differences in sensor characteristics (spectral, radiometric, spatial and angular) and acquisition geometry pose a number of challenges for the implementation of such an approach. ...
The synergy of a time series of optical satellite observations from a variety of sensors can be exploited to improve the retrieval of biophysical variables. Information from different sensors may assist in the variable retrieval by limiting potential ambiguities. This involves observations at different spatial, spectral, temporal and angular resolutions, etc. Furthermore, using timely data is of much importance for vegetation monitoring in environmental modeling. While the other necessary variables for such models can be collected daily (e.g. meteorological variables), the temporal resolution of optical sensors (high to intermediate spatial resolutions) does not allow having temporally frequent products of vegetation characteristics due to the revisit time of the sensors and cloud coverage. A multi-temporal, multi-sensor approach applied to a temporal sequence of radiometric data acquired by different sensors can improve mapping and monitoring of vegetation state variables over time. Even when no observations are available due to cloudiness or orbital configuration, the prior retrievals are taken.
... For instance; combining multi-temporal data from identical sensor (Hagolle et al., 2005;Koetz et al., 2007;Lauvernet et al., 2008;Menenti et al., 2005;Yang et al., 2006), synthetic data (Verhoef, 2007), a synergy of LIDAR data with 4.1. Introduction 4 57 spectral measurements (Koetz, 2006), combining hyperspectral and synthetic aperture radar (Treuhaft et al., 2002(Treuhaft et al., , 2004, integrating spectral and directional measurements (D'Urso et al., 2004;Widlowski et al., 2004) and a synergy of an ensemble of imaging instruments (Geiger et al., 2004;Jin et al., 2002;Samain et al., 2006). Although the retrieval of vegetation state variables from these multi-sensor approaches represents a promising advance in improving the quality of the retrieved products, the differences in sensor characteristics (spectral, radiometric, spatial and angular) and acquisition geometry pose a number of challenges for the implementation of such an approach. ...
... This is because vegetation properties are not generally perfectly diffuse reflectors, and sunlight reflected from vegetation exhibits a significant degree of anisotropy (Settle, 2004). The surface reflectance anisotropy can be estimated by directional reflectance analysis through the collection of multi-angular spectral data (Chooping et al., 2003;D'Urso et al., 2003;Widen, 2004). Proper characterization of the surface anisotropy is an important element in the successful interpretation of remotely sensed signals. ...
The surface reflectance anisotropy can be estimated by directional reflectance analysis through the collection of multi-angular spectral data. Proper characterization of the surface anisotropy is an important element in the successful interpretation of remotely sensed signals. A signal received by a sensor from a vegetation canopy is affected by several factors. One of them is the sensor zenith angle. Functional data analysis can be used to assess the distribution and variation of spectral reflectance due to sensor zenith angle. This paper examines the effect of sensor zenith angles on the spectral reflectance of vegetation, example on cotton leaves. The spectra were acquired in a green house trial in order to address the question ‘how much information can be obtained from multi-angular hyperspectral remote sensing on vegetation?’ The goals of the functional data analysis applied in this paper is to examine the Functional Data Analysis approach was applied to analysis multi-angular hyperspectral data on cotton, highlighting various characteristics of cotton spectra due to sensor view angles, and to infer directional variation in an outcome or dependent variable with different zenith angles.
... Rapid developments in modelling approaches and remote sensing technologies over the last two decades have inspired scientists to probe the relationships between structural parameters of a vegetation canopy and multi-angle, multispectral remote sensing data. The concept of combining multi-angle and hyperspectral remote sensing, in particular, is useful to retrieve both canopy structure and biochemical parameters of vegetation (Lewis et al. 2001, Rautiainen et al. 2004, Urso et al. 2004, Schlerf and Hill 2005, Simic and Chen 2008, Vuolo et al. 2008. At fine spectral resolution, Leaf chlorophyll content retrieval 3 hyperspectral data provide a unique way to retrieve chlorophyll content (Curran 1989, Elvidge and Chen 1995, Broge and Leblanc 2001, Zarco-Tejada et al. 2001, Gitelson et al. 2005. ...
The retrieval of total chlorophyll content (chla + b) per unit leaf area and unit ground area was investigated for a boreal forest near Sudbury in northern Ontario, Canada. The retrieval was based on inversions of the 5-Scale and PROSPECT models using canopy structure parameters, leaf area index (LAI) and clumping index, generated from off-nadir (multi-angle) multispectral data. Findings support the validity of combining nadir hyperspectral and multi-angle multispectral remote sensing in simultaneous retrieval of structural and biochemical vegetation parameters. Chlorophyll retrievals are improved once the improved structural parameters are obtained from multispectral data at two optimal off-nadir angles, the hotspot and darkspot. The estimated leaf chlorophyll contents agree well with the field measured values (R 2 = 0.89 and root mean square error (RMSE) = 8.1 μg cm−2). When the clumping index is excluded from the inversion, the coefficient of determination, R 2, decreases to 0.53 and the RMSE, increases to 13.4 μg cm−2.
