Radiative Transfer Model PROSPECT Research Articles

Nitrogen retrieval in grapevine (Vitis vinifera L.) leaves by hyperspectral sensing

Nitrogen (N) is a critical macronutrient that directly affects grapevine yield and quality but is also highly mobile in soil and can cause environmental contamination when over-applied. Monitoring leaf N content is useful for assessing vine N status and thereby improving N management plans, and the ability to assess N by remote sensing is desirable. Remote sensing technology has been utilized for retrieving N in agronomic crops since the 1990s. However, remote sensing has not been widely adopted for N management mainly due to variability in spectral patterns that emerge when training a model on one dataset and then applying it to a different dataset. Differences in environmental conditions, phenological stages, and plant varieties contribute to variability. This study aimed to understand the major factors limiting the generalizability of N-level retrieval algorithms in grapevine: the impact of leaf age and how calibrating nitrogen content, using either leaf dry mass (Nmass) or leaf area (Narea), influence the outcomes. This study also compares the performance among five major N retrieval approaches. We used spectral data from a hand-held hyperspectral sensor measuring wavelengths from 350 to 2500 nm from 664 individual leaf samples obtained from two grapevine varieties on three different sampling campaigns. Our findings indicate that while machine learning and chemometric approaches offer high accuracy levels (R2 = 0.78), the physical modeling approach utilizing the PROSPECT radiative transfer model (RTM) is capable of consistently retrieving N independent of specific dataset conditions and with an acceptable level of accuracy (R2 = 0.45) while using only 50% of the hyperspectral data. The estimation of Nmass using Proteinmass, which was calculated from Proteinarea retrieved by RTM and ground truth LMA, demonstrates a high potential for consistent retrieval of Total N (TN) in grapes. We also found that while calibrating a model by Narea can eliminate the negative impact of leaf age on prediction accuracy, Nmass still performed better in four out of five approaches tested. This study shows that a hybrid model that combines machine learning and RTM can alleviate the impact of leaf age on prediction accuracy while reducing 50% of spectral bands required and maintaining an acceptable level of accuracy (R2 = 0.54).

Generality of leaf spectroscopic models for predicting key foliar functional traits across continents: A comparison between physically- and empirically-based approaches

Leaf spectroscopy provides an efficient way of predicting foliar functional traits, commonly using physically- and empirically-based models. However, the generality of both models has not been fully investigated, and it is not clear if inversion strategies of physically-based models can be transferred across datasets. In this study, we evaluated the generality of leaf spectroscopic models for predicting key foliar functional traits and compared the performance of physically- and empirically-based approaches. Two extensive datasets compiling a total of 3861 foliar samples were collected from 24 field sites in eastern United States and south China. The leaf radiative transfer model PROSPECT was coupled with COSINE (PROCOSINE) to retrieve foliar traits from leaf bidirectional reflectance factor (BRF). A commonly used empirically-based model, partial least squares regression (PLSR) was performed as a comparison. Results showed that both PROSPECT and PROCOSINE can accurately estimate leaf mass per area (LMA) and equivalent water thickness (EWT). Inversion strategies including the selection of optimal spectral domains and the use of prior information (IS3) greatly improved the estimation accuracy of leaf nitrogen, leaf chlorophyll a + b and carotenoids. The estimation accuracies were similar when transferring inversion strategies across datasets, indicating a high level of transferability of physically-based models. PLSR and interval PLSR (iPLSR, via feature selection) could predict foliar traits with high accuracies when cross-validation was performed, and iPLSR achieved higher accuracies. But both the empirical approaches demonstrated low transferability when applied to an independent dataset. Our findings highlight the importance of generalized traits models with respect to development and calibration of leaf radiative transfer model, as well as incorporating representative samples in training empirical models. This study can help us to better understand the variation of foliar traits among and within species, their response to environmental change, as well as plant biodiversity.

Close-range, heavy-duty hyperspectral imaging for tracking drought impacts using the PROCOSINE model

Leaf response to stressors is heterogeneously distributed within the canopy, creating complex light-matter interactions that hinder translation of spectra into meaningful plant traits at larger scales. A deeper understanding of the linkages between leaf traits and canopy spectra has been achieved by a variety of close-range hyperspectral imaging systems. Yet the majority are scaled for small plants in pots, and thereby face numerous challenges in replicating field conditions for temporal drought monitoring. In monocots, drought alters foliar biochemistry and canopy structure, leading to the eventual senescence and decay of individual leaves, yet few radiative transfer models have been applied to track drought stress at the submillimeter scale in mixed canopies of photosynthetic and non-photosynthetic vegetation. Our objective was to compare the retrieved parameter set for PROCOSINE and PROSPECT radiative transfer model inversions utilizing two decay pigments. Although the average inversion values for both models indicated an increase in initial decay pigments after drought, the per-pixel inversion results for the PROCOSINE and PROSPECT inversions were unique (Mann-Whitney U-test p<0.05). Model retrievals for PROCOSINE were consistently higher for N, and lower for chlorophyll and carotenoids. In PROCOSINE, the Thatai and BSpec trended towards smaller values due to the increasing percentage of non-photosynthetic leaves, suggesting that mixed canopies are less prone to bidirectional reflectance. Regardless, the spatial-temporal trend of the inversion estimated parameter set was similar between both models, demonstrating the efficacy of the close-range imaging platform to monitor drought effects over time.

Application of Landsat ETM+ and OLI Data for Foliage Fuel Load Monitoring Using Radiative Transfer Model and Machine Learning Method

Foliage fuel load (FFL) is a critical factor affecting crown fire intensity and rate of spread. Satellite observations provide the potential for monitoring FFL dynamics across large areas. Previous studies commonly used empirical methods to estimate FFL, which potentially lacks reproducibility. This study applied Landsat 7 ETM+ and 8 OLI data for FFL retrieval using radiative transfer model (RTM) and machine learning method. To this end, the GeoSail, SAIL, and PROSPECT RTMs were first coupled together to model the near-realistic scenario of a two-layered forest structure. Second, available ecological information was applied to constrain the coupled RTM modeling phases in order to decrease the probability of generating unrealistic simulations. Third, the coupled RTMs were linked to three machine learning models-random forest, support vector machine, and multilayer perceptron-as well as the traditional lookup table. Finally, the performance of each method was validated by FFL measurements from Southwest China and Sweden. The resulting multilayer perceptron (R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> = 0.77, RMSE = 0.13, and rRMSE = 0.43) outperformed the other three methods. The evaluation of the applicability of the FFL estimates was conducted in a southwest China forest where two occurred in 2014 and 2020. The FFL dynamics from 2013 through 2020 showed that the fire was likely to occur when the FFL accumulated to a critical point (around 27 × 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">6</sup> kg), highlighting the relevance of remote sensing derived FFL estimates for understanding potential fire occurrence.

Spectral subdomains and prior estimation of leaf structure improves PROSPECT inversion on reflectance or transmittance alone

Leaf biochemical and structural traits are vegetation characteristics related to various physiological processes. Taking advantage of the physical relationship between optical properties and leaf biochemistry, field-based spectroscopy has allowed for the rapid estimation of leaf biochemical constituents and repeated non-destructive measurements through time. Leaf constituent retrieval from leaf optical properties following inversion of the physically-based radiative transfer model PROSPECT is now a popular method, but some cases prompt poor retrieval success and this approach requires a strict inversion procedure. We investigated the performances of different inversion procedures for the estimation of leaf constituents, specifically chlorophyll a and b, carotenoids, water (EWT), and dry matter (LMA) from >1400 broadleaf samples, including the definition of optimal spectral subdomains, and the use of leaf reflectance or transmittance alone. We also developed a strategy to obtain prior information on the leaf structure parameter (N) in PROSPECT, when only reflectance or transmittance is measured, and examined the influence of this prior information in combination with different inversion procedures. We found that using the full domain of reflectance or transmittance only systematically leads to suboptimal estimation of chlorophyll a and b, carotenoids, EWT, and LMA, due to either the combined absorption of multiple constituents or inaccurate estimation of the N parameter. Our study confirms that the selection of optimal spectral subdomains leads to improved estimation of all leaf constituents, from 700 to 720 nm for chlorophyll a and b, 520–560 nm for carotenoids, and from 1700 to 2400 nm for EWT and LMA. Prior information on N, computed directly from the spectra, leads to systematic improved estimation of leaf constituents when only reflectance or transmittance is measured, with reductions in normalized root mean square error from 8 to 37%. We strongly recommend using optimal subdomains when inverting PROSPECT to retrieve leaf constituents, and with the availability of only reflectance or transmittance we further recommend the use of prior information on the N parameter.

Developing a New Spectral Index for Detecting Cadmium-Induced Stress in Rice on a Regional Scale.

In natural farmland ecosystems, cadmium (Cd) pollution in rice has attracted increasing attention because of its high toxicity, relative mobility, and high water solubility. This study aims to develop a spectral index for detecting Cd stress in rice on a regional scale. Three experimental sites are selected in Zhuzhou City, Hunan Province. The hyperspectral data, chlorophyll (Chl) content, leaf area index, average leaf angle, Cd concentration in soil, and Sentinel-2A images from 2017 and 2018 are collected. A new spectral index sensitive to Cd stress in rice is established based on the global sensitivity analysis of the radiative transfer model PROSPECT + SAIL (commonly called PROSAIL) model with the auxiliary of the field-measured data. The heavy metal Cd stress-sensitive spectral index (HCSI) is devised as an indicator of the degree of Cd stress in rice. Results indicate that (1) the HCSI developed based on Chl is a good indicator of rice damage caused by Cd stress, that is, low values of HCSI occur in rice subject to relatively high pollution; (2) compared with common spectral indices, such as red-edge position and red-edge Chl index, HCSI is more sensitive to Chl content with higher Pearson correlation coefficients with respect to Chl content, ranging from 0.85 to 0.95; (3) HCSI is successfully applied in Sentinel-2A images from the two different years of monitoring rice Cd stress on a regional scale. Cd stress levels in rice stabilized, and the largest area percentage of each pollution levels of Cd decreased in the following order: No pollution (i.e., 40%), low pollution (i.e., 35%), and high pollution (i.e., 25%). This study indicates that a combination of simulation data from the PROSAIL model and measured data appears to be a promising method for establishing a sensitivity spectral index to heavy metal stress, which can accurately detect regional Cd stress in crops.

Evaluating the performance of PROSPECT in the retrieval of leaf traits across canopy throughout the growing season

Leaf traits and subsequently leaf spectral properties depend on the leaf phenological stage and light conditions within a canopy. The PROSPECT radiative transfer model has been extensively and successfully used to retrieve leaf traits for mature, sunlit leaves at peak vegetation growth, i.e. summer. However, research on the quantification of leaf traits using PROSPECT across the canopy vertical profile throughout the growing season is still lacking. Therefore, this study aims at examining the effect of leaf position on the performance of the PROSPECT model in modelling leaf optical properties and retrieving leaf chlorophyll content (Cab), equivalent water thickness (EWT), and leaf mass per area (LMA) throughout the growing season. To achieve this objective, we collected 588 leaf samples from the upper and lower canopies of deciduous stands over three seasons (i.e., spring, summer and autumn) in Bavaria Forest National Park, Germany. Leaf traits including Cab, EWT and LMA, were measured for all the samples, and their reflectance spectra were obtained using an ASD FieldSpec-3 Pro FR spectroradiometer coupled with an Integrating Sphere. We initially assessed the performance of the PROSPECT model by comparing reflectance spectra generated in forward mode against reflectance spectra measured on leaf samples collected in the field. We subsequently inverted the PROSPECT model to retrieve Cab, EWT and LMA using the look-up-table (LUT) approach. Our results consistently demonstrated that the measured reflectance of leaf samples collected from the lower canopy had a stronger match with PROSPECT simulated reflectance spectra, especially in the NIR spectrum compared to leaf samples collected from the upper canopy throughout the growing season. This observation concurred with the pattern of Cab and EWT retrieval accuracies across the canopy i.e. the retrieval accuracy for the lower canopy was consistently higher (NRMSE = 0.1-0.2 for Cab; NRMSE = 0.125-0.16 for EWT) when compared to the upper canopy (NRMSE = 0.122 - 0.269 for Cab; NRMSE = 0.162 -0.0.258 for EWT) across all seasons. In contrast, LMA retrieval accuracies for the upper canopy (NRMSE = 0.146 - 0.184) were higher compared to the lower canopy (NRMSE = 0.162 - 0.239) for all seasons except for the spring season. For all the leaf traits examined in this study, the range in retrieval accuracy between the upper and lower canopy was greater in summer (compared to other seasons). We report for the first time that although the PROSPECT model provides reasonable retrieval accuracy of Cab, EWT and LMA, variations in leaf biochemistry and morphology through the vertical canopy profile affects the performance of the model over the growing season. Findings of this study have important implications on field sampling protocols and upscaling leaf traits to canopy and landscape level using multi-layered physical models coupled with PROSPECT.

Wavelength selection of the multispectral lidar system for estimating leaf chlorophyll and water contents through the PROSPECT model

The estimation of leaf biochemical constituents is of high interest for the physiological and ecological applications of remote sensing. The multispectral lidar (MSL) system emerges as a promising active remote sensing technology with the ability to acquire both three-dimensional and spectral characteristics of targets. The detection wavelengths of the MSL system can be geared toward the specific application purposes. Therefore, it’s important to conduct the wavelength selection work to maximize the potential of the MSL system in vegetation monitoring. Traditional strategies of wavelength selection attempt to establish an empirical relationship between large quantities of observed reflectance and foliar biochemical constituents. By contrast, this study proposed to select wavelengths through the radiative transfer model PROSPECT. A five-wavelength combination was established to estimate leaf chlorophyll and water contents: 680, 716, 1104, 1882 and 1920 nm. The consistency of the wavelengths selected were tested by running different versions of PROSPECT model. Model inversion using simulated and experimental datasets showed that the selected wavelengths have the ability to retrieve leaf chlorophyll and water contents accurately. Overall, this study demonstrated the potential of the MSL system in vegetation monitoring and can serve as a guide in the design of new MSL systems for the application community.

Modelling leaf spectral properties in a soybean functional-structural plant model by integrating the prospect radiative transfer model.

Currently, functional-structural plant models (FSPMs) mostly resort to static descriptions of leaf spectral characteristics, which disregard the influence of leaf physiological changes over time. In many crop species, including soybean, these time-dependent physiological changes are of particular importance as leaf chlorophyll content changes with leaf age and vegetative nitrogen is remobilized to the developing fruit during pod filling. PROSPECT, a model developed to estimate leaf biochemical composition from remote sensing data, is well suited to allow a dynamic approximation of leaf spectral characteristics in terms of leaf composition. In this study, measurements of the chlorophyll content index (CCI) were linked to leaf spectral characteristics within the 400-800 nm range by integrating the PROSPECT model into a soybean FSPM alongside a wavelength-specific light model. Straightforward links between the CCI and the parameters of the PROSPECT model allowed us to estimate leaf spectral characteristics with high accuracy using only the CCI as an input. After integration with an FSPM, this allowed digital reconstruction of leaf spectral characteristics on the scale of both individual leaves and the whole canopy. As a result, accurate simulations of light conditions within the canopy were obtained. The proposed approach resulted in a very accurate representation of leaf spectral properties, based on fast and simple measurements of the CCI. Integration of accurate leaf spectral characteristics into a soybean FSPM leads to a better, dynamic understanding of the actual perceived light within the canopy in terms of both light quantity and quality.

Field spectroscopy and radiative transfer modelling to assess impacts of petroleum pollution on biophysical and biochemical parameters of the Amazon rainforest

Biophysical and biochemical plant foliage parameters play a key role in assessing vegetation health. Those plant parameters determine the spectral reflectance and transmittance properties of vegetation; therefore, hyperspectral remote sensing, particularly imaging spectroscopy, can provide estimates of leaf and canopy chemical properties. Based on the relationship between spectral response and biochemical/biophysical properties of the leaves and canopies, the PROSPECT radiative transfer model simulates the interaction of light with leaves. In this study, more than 1100 leaf samples from the Amazon forest of Ecuador were collected at several study sites, some of which are affected by petroleum pollution, and across the vertical profile of the forest. For every sample, field spectroscopy at leaf level was conducted with a spectroradiometer. The goal of this study was to assess leaf optical properties of polluted and unpolluted rainforest canopies across the vertical profile and identify vegetation stress expressed in changes of biophysical and biochemical properties of vegetation. An ANOVA followed by Holme’s multiple comparisons of means and a principal component analysis showed that photosynthetic pigments, chlorophyll and carotenoids have significantly lower levels across the vertical profile of the forest, particularly in sites affected by petroleum pollution. On the other hand, foliar water content showed significantly higher levels in the polluted site. Those findings are symptoms of vegetation stress caused by reduced photosynthetic activity and consequently decreased transpiration and water-use efficiency of the plants. Cross-comparison between SPAD-502 chlorophyll content meter index and chlorophyll content showed strong positive correlation coefficients (r = 0.71 and r2 = 0.51) which suggests that using the SPAD-502 chlorophyll index itself is sensitive enough to detect vegetation stress in a multispecies tropical forest. Therefore, the SPAD-502 can be used to assess chlorophyll content of vegetation across polluted and non-polluted sites at different canopy layers. The results presented in this paper contribute to the very limited literature on field spectroscopy and radiative transfer models applied to the vertical profile of the Amazon forest.

Estimating leaf functional traits by inversion of PROSPECT: Assessing leaf dry matter content and specific leaf area in mixed mountainous forest

Assessments of ecosystem functioning rely heavily on quantification of vegetation properties. The search is on for methods that produce reliable and accurate baseline information on plant functional traits. In this study, the inversion of the PROSPECT radiative transfer model was used to estimate two functional leaf traits: leaf dry matter content (LDMC) and specific leaf area (SLA). Inversion of PROSPECT usually aims at quantifying its direct input parameters. This is the first time the technique has been used to indirectly model LDMC and SLA. Biophysical parameters of 137 leaf samples were measured in July 2013 in the Bavarian Forest National Park, Germany. Spectra of the leaf samples were measured using an ASD FieldSpec3 equipped with an integrating sphere. PROSPECT was inverted using a look-up table (LUT) approach. The LUTs were generated with and without using prior information. The effect of incorporating prior information on the retrieval accuracy was studied before and after stratifying the samples into broadleaf and conifer categories. The estimated values were evaluated using R2 and normalized root mean square error (nRMSE).Among the retrieved variables the lowest nRMSE (0.0899) was observed for LDMC. For both traits higher R2 values (0.83 for LDMC and 0.89 for SLA) were discovered in the pooled samples. The use of prior information improved accuracy of the retrieved traits. The strong correlation between the estimated traits and the NIR/SWIR region of the electromagnetic spectrum suggests that these leaf traits could be assessed at canopy level by using remotely sensed data.

MAPPING OF FOLIAR CONTENT USING RADIATIVE TRANSFER MODELING AND VIS-NIR HYPERSPECTRAL CLOSE-RANGE REMOTE-SENSING

Abstract. Most methods for retrieving foliar content from hyperspectral data are well adapted either to remote-sensing scale, for which each spectral measurement has a spatial resolution ranging from a few dozen centimeters to a few hundred meters, or to leaf scale, for which an integrating sphere is required to collect the spectral data. In this study, we present a method for estimating leaf optical properties from hyperspectral images having a spatial resolution of a few millimeters or centimeters. In presence of a single light source assumed to be directional, it is shown that leaf hyperspectral measurements can be related to the directional hemispherical reflectance simulated by the PROSPECT radiative transfer model using two other parameters. The first one is a multiplicative term that is related to local leaf angle and illumination zenith angle. The second parameter is an additive specular-related term that models BRDF effects. Our model was tested on visible and near infrared hyperspectral images of leaves of various species, that were acquired under laboratory conditions. Introducing these two additional parameters into the inversion scheme leads to improved estimation results of PROSPECT parameters when compared to original PROSPECT. In particular, the RMSE for local chlorophyll content estimation was reduced by 21% (resp. 32%) when tested on leaves placed in horizontal (resp. sloping) position. Furthermore, inverting this model provides interesting information on local leaf angle, which is a crucial parameter in classical remote-sensing.

Prospect inversion for indirect estimation of leaf dry matter content and specific leaf area

Abstract. Quantification of vegetation properties plays an indispensable role in assessments of ecosystem function with leaf dry mater content (LDMC) and specific leaf area (SLA) being two important vegetation properties. Methods for fast, reliable and accurate measurement of LDMC and SLA are still lacking. In this study, the inversion of the PROSPECT radiative transfer model was used to estimate these two leaf parameters. Inversion of PROSPECT traditionally aims at quantifying its direct input parameters rather than identifying the parameters which can be derived indirectly from the input parameters. The technique has been tested here to indirectly model these parameters for the first time. Biophysical parameters such as leaf area, as well as fresh and dry weights of 137 leaf samples were measured during a field campaign in July 2013 in the mixed mountain forests of the Bavarian Forest National Park, Germany. Reflectance and transmittance of the leaf samples were measured using an ASD field spec III equipped with an integrating sphere. The PROSPECT model was inverted using a look-up table (LUT) approach for the NIR/SWIR region of the spectrum. The retrieved parameters were evaluated using their calculated R2 and normalized root mean square error (nRMSE) values with the field measurements. Among the retrieved variables the lowest nRMSE (0.0899) was observed for LDMC. For both traits higher R2 values (0.83 for LDMC and 0.89 for SLA) were discovered. The results indicate that the leaf traits studied can be quantified as accurately as the direct input parameters of PROSPECT. The strong correlation between the estimated traits and the NIR/SWIR region of the electromagnetic spectrum suggests that these leaf traits could be assessed at canopy and in the landscape by using hyperspectral remote sensing data.

Inverse Retrieval of Chlorophyll From Reflected Spectra for Assimilating Branches of Drought-Tolerant Tamarix ramosissima

Leaf chlorophyll content is a critical indicator for the better understanding of the featured carbon cycle and ecosystem functions in arid regions. The potential applicability of the inverse retrieval of chlorophyll via the radiative transfer model PROSPECT has been intensively examined in this study for a dominant species in desert ecosystems, Tamarix ramosissima, in central Asia, which has distinct structural features of xeromorphism and its young twig functioned as “assimilating organisms.” This study revealed that the performance of the inverse retrieval based on the original version of PROSPECT-4 was poor with very low accuracy. As a comparison, its calibrated version was capable of accurate retrieval of chlorophyll content, which recorded an R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> (coefficient of determination) of 0.47 with a root mean square error (RMSE) of 28.79 mg/m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> , over the entire measured chlorophyll range, from 57.37 to 202.27 mg/m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> . The model calibration is not considered to be overfitting since the tenfold cross-validation shows a close RMSE value of 26.19 mg/m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> . The effects of the percentage of training dataset for calibration and the number of calibration repeats have also been intensively investigated in this study. Despite some inherent defects of applying broadleaforiented PROSPECT on assimilating branches, it remains a feasible selection to inversely retrieve the chlorophyll content from the reflectance and thus provides a base for investigating the chlorophyll content of desert plants using hyperspectral remote sensing.

Retrieval of Chlorophyll Content from Leaf Reflectance Spectra Using Support Vector Machine

The chlorophyll content in crop leaf is an indicator of health situation and the crop yield. Hence, it is very important to retrieval of accurate chlorophyll content in paddy rice. This research selected Zhengyi town of Suzhou city as the study area, measurements were acquired during the summer of 2009, in a field campaign in which for 288 rice leaf samples, rice hyperspectral data was measured by ASD FieldSpec3 spectrometer, chlorophyll content was measured by using a SPAD-502 chlorophyll meter. And the parameters of support vector machine were optimized by genetic algorithm, then support vector machine and PROSPECT radiative transfer model were adopted to build estimation model, which used to retrieve the chlorophyll content of rice. The results indicate that: the coefficient of determination for the rice chlorophyll estimation model is 0.8825, and RMSE is 8.7491. Research of this paper provides some reference for quickly and accurately estimating the chlorophyll content in rice.

A reflectance spectra model of heavy metal stressed leaves: advances in the PROSPECT model adding specific absorption coefficients of heavy metal ion

This article aims to investigate the reflectance model of heavy metal copper stressed crop. Forty-six groups of copper-treated leaves were measured during a laboratory experiment in order to obtain the leaf biochemical component information and its corresponding scanning electron microscopy image. Then a new reflectance spectral model was developed on the basis of the classical broadleaf radiative transfer model-PROSPECT. Comparing with the PROSPECT model, new models mainly consider adding specific absorption coefficient of copper ion. The scattering process is described by a refractive index (n) and a leaf structure parameter (N). Absorption is modeled using pigment concentration, water content, dry matter content, copper ion contamination and the corresponding specific spectral absorption coefficients (Kab, Kw, Kd, KCu). Thus, reflectance spectral modeling is an inversion procedure to calculate the above 6 parameters accurately. To validate the model 16 leaves were tested in laboratory experiment. This experiment showed that the inversion values of KCu had very strong agreement with the published absorption spectra of Cupric Chloride. The linear regression analysis between simulated and measured reflectance provides a correlation coefficient of 0.93 and a root mean squares of 0.067.

Optimal modalities for radiative transfer-neural network estimation of canopy biophysical characteristics: Evaluation over an agricultural area with CHRIS/PROBA observations

Neural networks trained over radiative transfer simulations constitute the basis of several operational algorithms to estimate canopy biophysical variables from satellite reflectance measurements. However, only little attention was paid to the training process which has a major impact on retrieval performances. This study focused on the several modalities of the training process within neural network estimation of LAI, FCOVER and FAPAR biophysical variables. Performances were evaluated over both actual experimental observations and model simulations. The SAIL and PROSPECT radiative transfer models were used here to simulate the training and the synthetic test datasets. Measurements of LAI, FCOVER and FAPAR were achieved over the Barrax (Spain) agricultural site for a range of crop types concurrently to CHRIS/PROBA satellite image acquisition. Results showed that the spectral band selection was specific to LAI, FCOVER and FAPAR variables. The optimal band set provided significantly improved performances for LAI, while only small differences were observed for the other variables. Gaussian distributions of the radiative transfer model input variables performed better than uniform distributions for which no prior information was exploited. Including moderate uncertainties in the reflectance simulations used in the training process improved the flexibility of the neural network in cases where simulations departed slightly from observations. Simple neural network architecture with a single hidden layer of five tangent sigmoid transfer functions was performing as good as more complex architectures if the training dataset was larger than ten times the number of coefficients to tune. Small sensitivity of performances was observed depending on the way the solution was selected when several networks were trained in parallel. Finally, comparison with a NDVI based approach showed the generally better retrieval accuracy of neural networks.

Biophotonic in situ sensor for plant leaves

Knowledge of the water concentration of plants can be helpful in several environmental and agricultural domains. There are many methods for the determination of water content in plant leaves; however, most of them give a relative moisture level or an analytical measure after a previous calibration procedure. Even for other biochemical compounds such as dry matter or chlorophyll, the measurement techniques could be destructive. For this reason, a nondestructive method has been developed to measure the biochemical compounds of a plant leaf, using an infrared spectroscopy technique. One important advantage is the simplicity of the device (RAdiomètre portatif de Mesure In Situ, RAMIS) and its capability to perform measurements in situ. The prototype is a leaf-clip configuration and is made of LEDs at five wavelengths (656, 721, 843, 937, and 1550 nm), and a silicon/germanium photosensor. To compute the water content of vegetative leaves, the radiative transfer model PROSPECT was implemented. This model can accurately predict spectral transmittances in the 400 nm to 2500 nm spectral region as a function of the principal leaf biochemical contents: water, dry matter, and chlorophyll. Using the transmittance measured by RAMIS into an inversion procedure of A Model of Leaf Optical Properties Spectra, we are able to compute the values of water contents that show an agreement with the water contents measured directly using dry weight procedures. This method is presented as a possibility to estimate other leaf biochemical compounds using appropriate wavelengths.

Physically based retrieval of crop characteristics for improved water use estimates

Abstract. The increasing scarcity of water from local to global scales requires the efficient monitoring of this valuable resource, especially in the context of a sustainable management in irrigated agriculture. In this study, a two-source energy balance model (TSEB) was applied to the Barrax test site. The inputs of leaf area index (LAI) and fractional vegetation cover (fCover) were estimated from CHRIS imagery by using the traditional scaled NDVI and a look-up table (LUT) inversion approach. The LUT was constructed by using the well established SAILH + PROSPECT radiative transfer model. Simulated fluxes were compared with tower measurements and vegetation characteristics were evaluated with in situ LAI and fCover measurements of a range of crops from the SPARC campaign 2004. Results showed a better retrieval performance for the LUT approach for canopy parameters, affecting flux predictions that were related to land use.

Retrieval of vegetation equivalent water thickness from reflectance using genetic algorithm (GA)-partial least squares (PLS) regression

This study aimed to investigate the performance of genetic algorithms coupled with partial least squares (GA-PLS) modeling of spectral reflectance in retrieving equivalent water thickness (EWT) at leaf and canopy level. A genetic algorithm was used to identify a subset of spectral bands sensitive to the variation in EWT, and PLS was then applied to relate the identified bands to EWT values. GA-PLS was applied to leaf level reflectance available from LOPEX dataset, and canopy data includes reflectance simulated by a leaf radiative transfer model PROSPECT and a canopy radiative transfer model SAILH and acquired by airborne visible/infrared imaging spectrometer (AVIRIS). The results indicate that GA-PLS has the capability of retrieving EWT from leaf and canopy reflectance, and achieved good estimation accuracy, i.e. low root mean square errors (RMSE) and high squared correlation coefficients ( R 2). For the retrieval at leaf level, the estimation accuracy can be as good as RMSE = 0.0019 g/cm 2 and R 2 = 0.939 or better. For the retrieval at canopy level, the model accuracy is RMSE = 0.0061 g/cm 2 and R 2 = 0.966 or better when PROSPECT-SAILH simulated reflectance was used; when AVIRIS image spectra were used, the model accuracy is RMSE = 0.0094 g/cm 2 and R 2 = 0.8734 for the calibration, and RMSE = 0.0132 g/cm 2 and R 2 = 0.7756 for the validation. These results from GA-PLS modeling support the conclusion that GA-PLS has the potential to be applied to AVIRIS, Hyperion and HyMap imagery for retrieving EWT. The selected bands for the AVIRIS datasets differ from those for the LOPEX and PROSPECT-SAILH simulated datasets, and this inconsistence of the selected bands for different datasets indicates that the GA-PLS method has the advantage of tuning the optimum bands for PLS regression and accommodating the effects of confounding factors.