Estimating leaf nitrogen concentration in heterogeneous crop plants from hyperspectral reflectance

The estimation of leaf nitrogen concentration (LNC) in crop plants is an effective way to optimize nitrogen fertilizer management and to improve crop yield. The objectives of this study were to (1) analyse the spectral features, (2) explore the spectral indices, and (3) investigate a suitable modelling strategy for estimating the LNC of five species of crop plants (rice (Oryza sativa L.), corn (Zea mays L.), tea (Camellia sinensis), gingili (Sesamum indicum), and soybean (Glycine max)) with laboratory-based visible and near-infrared reflectance spectra (300–2500 nm). A total of 61 leaf samples were collected from five species of crop plant, and their LNC and reflectance spectra were measured in laboratories. The reflectance spectra of plants were reduced to 400–2400 and smoothed using the Savitzky–Golay (SG) smoothing method. The normalized band depth (NBD) values of all bands were calculated from SG-smoothed reflectance spectra, and a successive projections algorithm-based multiple linear regression (SPA-MLR) method was then employed to select the spectral features for five species. The SG-smoothed reflectance spectra were resampled using a spacing interval of 10 nm, and normalized difference spectral index (NDSI) and three-band spectral index (TBSI) were calculated for all wavelength combinations between 400 and 2400 nm. The NDSI and TBSI values were employed to calibrate univariate regression models for each crop species. The leave-one-out cross-validation procedure was used to validate the calibrated regression models. Study results showed that the spectral features for LNC estimation varied among different crop species. TBSI performed better than NDSI in estimating LNC in crop plants. The study results indicated that there was no common optimal TBSI and NDSI for different crop species. Therefore, we suggest that, when monitoring LNC in heterogeneous crop plants with hyperspectral reflectance, it might be appropriate to first classify the data set considering different crop species and then calibrate the model for each species. The method proposed in this study requires further testing with the canopy reflectance and hyperspectral images of heterogeneous crop plants.     

Comparative analysis of four semi-analytical models for estimating chlorophyll-a concentration in case-2 waters using field hyperspectral reflectance

ABSTRACT

Hyperspectral remote sensing can capture the complicated and variable characteristics of inland waters; thus, it is suited for the water quality assessment of Case-2 waters, and it has the potential to attain high estimation accuracy. In the present study, four improved models adapted from published approaches (three-band index, ΔΦ, BNDBI and TCARI) were investigated to estimate chlorophyll-a (chl-a) for the case of Dianshan Lake, China. Calibration and validation were provided from in situ measured chl-a and field hyperspectral measurements. The improved three-band (ITB) model, ΔΦ model, and BNDBI model yielded satisfactory results and enabled the estimation of chl-a for inland Case-2 waters with coefficients of determination (R²) reaching 0.75, 0.76, and 0.86, respectively. In particular, the TCARI/OSAVI model presented the highest accuracy (R² = 0.94) compared to the other models. All of the results provide strong evidence that the hyperspectral models presented in this paper are promising and applicable to estimate chl-a in eutrophic inland Case-2 waters.     

Validation of artificial neural network techniques in the estimation of nitrogen concentration in rape using canopy hyperspectral reflectance data

A systematic comparison of two types of method for estimating the nitrogen concentration of rape is presented: the traditional statistical method based on linear regression and the emerging computationally powerful technique based on artificial neural networks (ANN). Five optimum bands were selected using stepwise regression. Comparison between the two methods was based primarily on analysis of the statistic parameters. The rms. error for the back-propagation network (BPN) was significantly lower than that for the stepwise regression method, and the T-value was higher for BPN. In particular, for the first-difference of inverse-log spectra (log 1/R)′, T-values performed with a 127.71% success rate using BPN. The results show that the neural network is more robust to training and estimating rape nitrogen concentrations using canopy hyperspectral reflectance data.     

Wavelet-based detection of crop zinc stress assessment using hyperspectral reflectance

Accurate detection of heavy metal-induced stress on the growth of crops is essential for agricultural ecological environment and food security. This study focuses on exploring singularity parameters as indicators for a crop's Zn stress level assessment by applying wavelet analysis to the hyperspectral reflectance. The field in which the experiment was conducted is located in the Changchun City, Jilin Province, China. The hyperspectral and biochemistry data from four crops growing in Zn contaminated soils: rice, maize, soybean and cabbage were collected. We performed a wavelet transform to the hyperspectral reflectance (350-1300 nm), and explored three categories of singularity parameters as indicators of crop Zn stress, including singularity range (S_R), singularity amplitude (S_A) and a Lipschitz exponent (α). The results indicated that (i) the wavelet coefficient of the fifth decomposition level by applying Daubechies 5 (db5) mother wavelets proved successful for identifying crop Zn stress; the S_R of crop concentrated on the region was around 550-850 nm of the spectral signal under Zn stress; (ii) the S_R stabilized, but S_A and α had developed some variations at the growth stages of the crop; (iii) the S_R, S_A and α were found among four crop species differentially; and moreover the S_A increased in relation to an increase in the S_R of crop species; (iv) the α had a strong non-linear relationship with the Zn concentration (R²:0.7601-0.9451); the S_A had a strong linear relationship with Zn concentration (R²:0.5141-0.8281). Singularity parameters can be used as indicators for a crop's Zn stress level as well as offer a quantitative analysis of the singularity of spectrum signal. The wavelet transform technique has been shown to be very promising in detecting crops with heavy metal stress.     

Prediction of concentration for microalgae using image analysis

Multimedia Tools and Applications - Maintaining the optimum growth rate and estimating the concentration of microalgae are critical in improving microalgae production. An efficient concentration...     

Evaluating the performance of PC-ANN for the estimation of rice nitrogen concentration from canopy hyperspectral reflectance

In this study, a wide range of leaf nitrogen concentration levels was established in field-grown rice with the application of three fertilizer levels. Hyperspectral reflectance data of the rice canopy through rice whole growth stages were acquired over the 350 nm to 2500 nm range. Comparisons of prediction power of two statistical methods (linear regression technique (LR) and artificial neural network (ANN)), for rice N estimation (nitrogen concentration, mg nitrogen g^?1 leaf dry weight) were performed using two different input variables (nitrogen sensitive hyperspectral reflectance and principal component scores). The results indicted very good agreement between the observed and the predicted N with all model methods, which was especially true for the PC-ANN model (artificial neural network based on principal component scores), with an RMSE?=?0.347 and REP?=?13.14%. Compared to the LR algorithm, the ANN increased accuracy by lowering the RMSE by 17.6% and 25.8% for models based on spectral reflectance and PCs, respectively.     

Identification and mapping of minerals in drill core using hyperspectral image analysis of infrared reflectance spectra

A Portable Infrared Mineral Analyzer II (PIMA II) field spectrometer was used to measure infrared reflectance spectra (1·3-2·5 μm) of split drill core at 1 cm intervals in both the along-core and cross-core directions. These data were formatted into an image cube similar to that acquired by an imaging spectrometer with 600 spectral channels, and multi-spectral and hyperspectral analysis techniques were used for analysis. Colour images and enhancements provided visual displays of the spectral information, while real-time digital extraction of individual spectra allowed identification of minerals. Absorption band-depth mapping and spectral classification were used to map the spatial distribution of specific minerals in the core. Linear spectral unmixing provided estimated mineral abundances. Analysis results demonstrate that multi-spectral and hyperspectral image analysis methods can be used to produce detailed mineralogical maps of drill core. They suggest that the concepts and analytical techniques developed for analysis of hyperspectral image data can be applied to field and laboratory spectra in a variety of disciplines, and raise the question of the use of hyperspectral scanners in the laboratory.     

Estimating sulphide ore grade in broken rock using visible/infrared hyperspectral reflectance spectra

The field of hyperspectral remote sensing has developed rapidly for widespread mineral mapping from airborne platforms. The purpose of the current study was to examine whether hyperspectral spectrometry (0.35-2.5 w m) can be used in an underground mining environment for mapping the grade of sulphide ore in rock faces, hand specimens and core logging. Naturally broken samples of barren and ore-bearing rocks were collected from mines in the Sudbury Basin, Ontario, and dry and wet reflectance were measured. The sulphide minerals exhibit a one-sided absorption band at short wavelengths known as a conductance band. The hydroxyl-bearing silicates exhibit a triple absorption feature near 2.3 w m. Two ratios, one describing the conductance band and one describing the hydroxyl band, can be used to separate high grade ores (>20-25% sulphides) from barren and lower grade rocks. The conductance band ratio can also be used to estimate the concentration of chalcopyrite alone, - 15% chalcopyrite, absolute. Errors are proportional to the concentration of pyrrhotite and pentlandite. Errors can be reduced if total sulphides are estimated by other means, which a parallel study indicates is possible using thermal reflectance wavelengths. The study indicates that there is a high potential to use hyperspectral tools to grade sulphide ores.     

Towards universal broad leaf chlorophyll indices using PROSPECT simulated database and hyperspectral reflectance measurements

Fifty-three leaves were randomly sampled on different deciduous tree species, representing a wide range of chlorophyll contents, tree ages, and leaf structural features. Their reflectance was measured between 400 and 800 nm with a 1-nm step, and their chlorophyll content determined by extraction. A larger simulated database (11,583 spectra) was built using the PROSPECT model, in order to test, calibrate, and obtain universal indices, i.e., indices applicable to a wide range of species and leaf structure. To our knowledge, almost all leaf chlorophyll indices published in the literature since 1973 have been tested on both databases. Fourteen canonical types of indices (published ones and new ones) were identified, and their wavelengths calibrated on the simulated database as well as on the experimental database to determine the best wavelengths and, hence, the best performances in chlorophyll estimation for each index types. These indices go from simple reflectance ratios to more sophisticated indices using reflectance first derivatives (using the Savitzky and Golay method). We also tested other nondestructive methods to obtain total chlorophyll concentration: SPAD (Minolta Camera, Osaka, Japan) and neural networks. The validity of the actual PROSPECT model is challenged by our results: Important discordances are found when the indices are calculated with PROSPECT compared to experimental data, especially for some indices and wavelengths. The discordance is even greater when the indices are determined with PROSPECT and applied on the experimental database. A new calibration of PROSPECT is therefore necessary for any study aiming at using simulated spectra to determine or to calibrate indices. The “peak jump” and the multiple-peak feature observed on the first derivative of the reflectances (e.g., in the Red-Edge Inflection Point [REIP] index) has been investigated. It was shown that chlorophyll absorption alone can explain this feature. The peak jump disqualifies the REIP to be a valuable chlorophyll index. A simple modified difference ratio gave the best results among all published indices (cross-validated RMSE=2.1 μg/cm² on the experimental database). After calibration on the experimental database, modified Simple Ratio (mSR) and modified Normalized Difference (mND) indices gave the best performances (RMSECV=1.8 μg/cm² on the experimental database). The new Double Difference (DD) index, although not the best on the experimental database (RMSECV=2.9 μg/cm²), has the best results on the larger simulated database (RMSE=3.7 μg/cm²) and is expected to give good results on larger experimental databases. The best reflectance-based indices give better performances than the current commercial nondestructive device SPAD (RMSECV=4.5 μg/cm²). In this leaf-level study, the best indices are very near from each other, so that complex methods are useless: REIP-like, neural networks, and derivative-based indices are not necessary and give worst results than simpler properly chosen indices. These conclusions will certainly be different for a canopy-level study, where the derivative-based indices may perform significantly better than the other ones.     

Estimation of foliar pigment concentration in floating macrophytes using hyperspectral vegetation indices

Foliar pigment concentrations of chlorophylls and cartenoids are important indicators of plant physiological status, photosynthesis rate, and net primary productivity. Although the utility of hyperspectral derived vegetation indices for estimating foliar pigment concentration has been documented for many vegetation types, floating macrophytes have not been assessed despite their ecological importance. This study surveyed 39 wetland species (12 floating macrophytes (FM), 8 grasses/sedges/rushes (GSR), and 19 herbs/wildflowers (HWF)) to determine whether foliar pigment concentrations could be estimated from hyperspectral reflectance. Hyperspectral reflectance of samples was recorded using an ASD FieldSpec3 Max portable spectroradiometer with the plant probe attachment or via a typical laboratory set-up. A semi-empirical relationship was established using either a linear, second-degree polynomial or logarithmic function between 13 candidate vegetation indices and chl-a, chl-b, Car, and chl-a + b pigment concentrations. Vegetation indices R-M, CI-Red, and MTCI were strongly correlated with foliar pigment concentrations using a linear fitting function. Chl-a + b and chl-b concentrations for all samples were reasonably estimated by the R-M index (R² = 0.66 and 0.64), although Chl-a and Car concentration estimates using CI-Red were weaker (R² = 0.63 and 0.51). Regression results indicate that pooled samples to estimate individual foliar pigments were less correlated than when each type of vegetation type was treated separately. For instance, chl-a + b was best estimated by CI-Red for FM (R² = 0.80), MTCI for HWF (R² = 0.77), and R-M for GSR (R² = 0.67). Although floating macrophytes feature unique adaptions to their aquatic environment, their foliar pigment concentrations and spectral signatures were comparable to other wetland vegetation types. Overall, vegetation indices that exploit the red-edge region were a reasonable compromise, having good explanatory power for estimation of foliar pigments across the sampled wetland vegetation types and with CI-Red the best suited index for floating macrophytes.     

Modeling canopy reflectance and microwave backscattering coefficient

Various approaches to model canopy reflectance (CR) in the visible/infrared region and backscattering coefficient (BSC) in the microwave region are compared and contrasted. It is noted that BSC can be related to CR in the source direction (the “hot spot” direction). By assuming a frequency dependent leaf reflectance and transmittance it is shown that the observed dependence of BSC on leaf area index, leaf angle distribution, angle of incidence, soil moisture content, and frequency can be simulated by a CR model. Thus both BSC and CR can, in principle, be calculated using a single model which has essentially the same parameters as many CR models do.     

Retrieval of chlorophyll concentration from leaf reflectance spectra using wavelet analysis

The dynamics of foliar chlorophyll concentrations have considerable significance for plant-environment interactions, ecosystem functioning and crop growth. Hyperspectral remote sensing has a valuable role in the monitoring of such dynamics. This study focussed upon improving the accuracy of chlorophyll quantification by applying wavelet analysis to reflectance spectra. Leaf-scale radiative transfer models were used to generate very large spectral data sets with which to develop and rigorously test refinements to the approach and compare it with existing spectral indices. The results demonstrated that by decomposing leaf spectra, the resultant wavelet coefficients can be used to generate accurate predictions of chlorophyll concentration, despite wide variations in the range of other biochemical and biophysical factors that influence leaf reflectance. Wavelet analysis outperformed predictive models based on untransformed spectra and a range of spectral indices. The paper discusses the possibilities for further refining the wavelet approach and for extending the technique to the sensing of a variety of vegetation properties at a range of spatial scales.     

Estimating and mapping grass phosphorus concentration in an African savanna using hyperspectral image data

We tested the utility of imaging spectroscopy and neural networks to map phosphorus concentration in savanna grass using airborne HyMAP image data. We also sought to ascertain the key wavelengths for phosphorus prediction using hyperspectral remote sensing. The remote sensing of foliar phosphorus has received very little attention as compared to nitrogen, yet it plays an equally important role in explaining the distribution and feeding patterns of herbivores. Band depths from two continuum‐removed absorption features as well as the red edge position (REP) were input into a backpropagation neural network. Following a series of experiments to ascertain the optimum wavelengths, the best trained neural network was used to predict and ultimately to map grass phosphorus concentration in the Kruger National Park. The results indicate that the best trained neural network could predict phosphorus distribution with a coefficient of determination of 0.63 and a root mean square error (RMSE) of 0.07 (28% of the mean observed phosphorus concentration) on an independent test data set. Our results also show that the absorption feature located in the shortwave infrared (R _2015–2199) contains more information on phosphorus distribution, a region that has hardly been explored before in most spectroscopic experiments for phosphorus as compared to the visible bands. Overall, the study demonstrates the potential of imaging spectroscopy in mapping grass phosphorus concentration in savanna rangelands.     

Distinguishing mangrove species with laboratory measurements of hyperspectral leaf reflectance

As a first step in developing classification procedures for remotely acquired hyperspectral mapping of mangrove canopies, we conducted a laboratory study of mangrove leaf spectral reflectance at a study site on the Caribbean coast of Panama, where the mangrove forest canopy is dominated by Avicennia germinans, Laguncularia racemosa, and Rhizophora mangle. Using a high‐resolution spectrometer, we measured the reflectance of leaves collected from replicate trees of three mangrove species growing in productive and physiologically stressful habitats. The reflectance data were analysed in the following ways. First, a one‐way ANOVA was performed to identify bands that exhibited significant differences (P value<0.01) in the mean reflectance across tree species. The selected bands then formed the basis for a linear discriminant analysis (LDA) that classified the three types of mangrove leaves. The contribution of each narrow band to the classification was assessed by the absolute value of standardised coefficients associated with each discriminant function. Finally, to investigate the capability of hyperspectral data to diagnose the stress condition across the three mangrove species, four narrow band ratios (R ₆₉₅/R ₄₂₀, R ₆₀₅/R ₇₆₀, R ₆₉₅/R ₇₆₀, and R ₇₁₀/R ₇₆₀ where R ₆₉₅ represents reflectance at wavelength of 695nm, and so on) were calculated and compared between stressed and non‐stressed tree leaves using ANOVA.

Results indicate a good discrimination was achieved with an average kappa value of 0.9. Wavebands at 780, 790, 800, 1480, 1530, and 1550 nm were identified as the most useful bands for mangrove species classification. At least one of the four reflectance ratio indices proved useful in detecting stress associated with any of the three mangrove species. Overall, hyperspectral data appear to have great potential for discriminating mangrove canopies of differing species composition and for detecting stress in mangrove vegetation.     

Absorption spectrum of phytoplankton pigments derived from hyperspectral remote-sensing reflectance

For a data set collected around Baja California with chlorophyll-a concentration ((chl-a)) ranging from 0.16 to 11.3 mg/m³, hyperspectral absorption spectra of phytoplankton pigments were independently inverted from hyperspectral remote-sensing reflectance using a newly developed ocean-color algorithm. The derived spectra were then compared with those measured from water samples using the filter-pad technique, and an average difference of 21.4% was obtained. These results demonstrate that the inversion algorithm worked quite well for the coastal waters observed and suggest a potential of using hyperspectral remote sensing to retrieve both chlorophyll-a and other accessory pigments.     

Retrieval of chlorophyll a concentration from a fluorescence enveloped area using hyperspectral data

It is acknowledged that fluorescence line height (FLH) algorithms are still hampered by the uncertainty of fluorescence peak position. The fluorescence peak moves to longer wavelengths with the increase of chlorophyll a concentration. In this article, the fluorescence enveloped area (FEA), which integrates the fluorescence height and the fluorescence peak position, was used to estimate the chlorophyll a concentration in the coastal waters of the Pearl River Estuary. The FEA algorithm was developed from in situ data of chlorophyll a concentration, total suspended matter (TSM) concentration and above-water remote sensing reflectance, which were collected at 37 sampling stations in the Pearl River Estuary during two cruises. The results showed that the FEA algorithm made a better estimation of chlorophyll a concentration compared with the widely used FLH algorithm and moving fluorescence line height (MFLH) algorithm. These three algorithms were applied to the Pearl River Estuary for retrieval of chlorophyll a concentration from Hyperion data acquired on 21 December 2006. Compared with the FLH and the MFLH, the FEA algorithm showed a rational distribution of the chlorophyll a concentration in the Pearl River Estuary.     

Correlation between soil apparent electroconductivity and plant hyperspectral reflectance in a managed wetland

The apparent electrical conductivity (σ_a) of soil is influenced by a complex combination of soil physical and chemical properties. For this reason, σ_a is proposed as an indicator of plant stress and potential community structure changes in an alkaline wetland setting. However, assessing soil σ_a is relatively laborious and difficult to accomplish over large wetland areas. This work examines the feasibility of using the hyperspectral reflectance of the vegetation canopy to characterize the σ_a of the underlying substrate in a study conducted in a Central California managed wetland. σ_a determined by electromagnetic (EM) inductance was tested for correlation with in-situ hyperspectral reflectance measurements, focusing on a key waterfowl forage species, swamp timothy (Crypsis schoenoides). Three typical hyperspectral indices, individual narrow-band reflectance, first-derivative reflectance and a narrow-band normalized difference spectral index (NDSI), were developed and related to soil σ_a using univariate regression models. The coefficient of determination (R ²) was used to determine optimal models for predicting σ_a, with the highest value of R ² at 2206 nm for the individual narrow bands (R ²?=?0.56), 462 nm for the first-derivative reflectance (R ²?=?0.59), and 1549 and 2205 nm for the narrow-band NDSI (R ²?=?0.57). The root mean squared error (RMSE) and relative root mean squared error (RRMSE) were computed using leave-one-out cross-validation (LOOCV) for accuracy assessment. The results demonstrate that the three indices tested are valid for estimating σ_a, with the first-derivative reflectance performing better (RMSE?=?30.3 mS m^?1, RRMSE?=?16.1%) than the individual narrow-band reflectance (RMSE?=?32.3 mS m^?1, RRMSE?=?17.1%) and the narrow-band NDSI (RMSE?=?31.5 mS m^?1, RRMSE?=?16.7%). The results presented in this paper demonstrate the feasibility of linking plant–soil σ_a interactions using hyperspectral indices based on in-situ spectral measurements.     

Using vegetation reflectance variability for species level classification of hyperspectral data

A study was conducted to investigate whether reflectance data from vegetation in a tropical forest canopy could be used for species level discrimination. Reflectance spectra of 11 species were analysed at the scale of the leaf, branch, tree and species. To enhance separation of species-of-interest spectra from the other spectra in the data, the variation in reflectance values for the species-of-interest were used to create a characteristic spectral shape. With a simple algorithm, the resultant shape-space was used as a data filter that correctly discriminated against 94% of the non-species-of-interest trees.     

Early detection of basal stem rot disease (Ganoderma) in oil palms based on hyperspectral reflectance data using pattern recognition algorithms

Basal stem rot (BSR) is a fatal fungal (Ganoderma) disease of oil palm plantations and has a significant impact on the production of palm oil in Malaysia. Because there is no effective treatment to control this disease, early detection of BSR is vital for sustainable disease management. The limitations of visual detection have led to an interest in the development of spectroscopically based detection techniques for rapid diagnosis of this disease. The aim of this work was to develop a procedure for early and accurate detection and differentiation of Ganoderma disease with different severities, based on spectral analysis and statistical models. Reflectance spectroscopy analysis ranging from the visible to near infrared region (325–1075 nm) was applied to analyse oil palm leaf samples of 47 healthy (G0), 55 slightly damaged (G1), 48 moderately damaged (G2), and 40 heavily damaged (G3) trees in order to detect and quantify Ganoderma disease at different levels of severity. Reflectance spectra were pre-processed, and principal component analysis (PCA) was performed on different pre-processed datasets including the raw dataset, first derivative, and second derivative datasets. The classification models: linear and quadratic discrimination analysis, k-nearest neighbour (kNN), and Naïve–Bayes were applied to PC scores for classifying four levels of stress in BSR-infected oil palm trees. The analysis showed that the kNN-based model predicted the disease with a high average overall classification accuracy of 97% with the second derivative dataset. Results confirmed the usefulness and efficiency of the spectrally based classification approach in rapid screening of BSR in oil palm.