Canopy Hyperspectral Reflectance Research Articles

Mining sensitive hyperspectral feature to non-destructively monitor biomass and nitrogen accumulation status of tea plant throughout the whole year

Rapid and non-destructive estimation of tea plant growth and nitrogen (N) nutrition status using hyperspectral remote sensing is crucial for precise management of tea gardens. This study aimed to mine and fuse sensitive hyperspectral features to achieve an accurate estimation of tea plant growth parameters (biomass and N accumulation) throughout the whole year. An ASD Handheld 2 sensor was used to collect canopy hyperspectral reflectance of tea plants across four periods (Period 1–4) within a year, with tea plant biomass and N accumulation indicators acquired synchronously. The measured spectral reflectance and its first derivative, and wavelet feature were extracted and used to establish quantitative relationships with tea plant growth parameters. Random forest and LASSO algorithms were employed to combine sensitive hyperspectral features and construct the biomass and N accumulation monitoring models. The results showed that wavelet features (R2 = 0.35–0.58) had a stronger correlation with tea plant biomass and N accumulation parameters compared with the measured reflectance or first derivative spectral features. Similarly, the hyperspectral indices (R2 = 0.51–0.69) derived from sensitive wavelet features performed an accurate estimation of tea plant growth parameters. Furthermore, the combination of sensitive hyperspectral indices derived from measured reflectance, first derivative, and wavelet feature using random forest (R2 = 0.67–0.76) and LASSO (R2 = 0.61–0.72) algorithms achieved the greatest accuracy for monitoring tea plant biomass and N accumulation compared with individual hyperspectral feature. Additionally, the above estimation models obtained higher accuracy in period 4 compared to periods 1–3. This study provides valuable remote sensing technical support for predicting biomass and N accumulation status of tea plant throughout the whole year.

Spectral-genomic chain-model approach enhances the wheat yield component prediction under the Mediterranean climate.

In light of the changing climate that jeopardizes future food security, genomic selection is emerging as a valuable tool for breeders to enhance genetic gains and introduce high-yielding varieties. However, predicting grain yield is challenging due to the genetic and physiological complexities involved and the effect of genetic-by-environment interactions on prediction accuracy. We utilized a chained model approach to address these challenges, breaking down the complex prediction task into simpler steps. A diversity panel with a narrow phenological range was phenotyped across three Mediterranean environments for various morpho-physiological and yield-relatedtraits. The results indicated that a multi-environment model outperformed a single-environment model in prediction accuracy for most traits. However, prediction accuracy forgrain yield was not improved. Thus, in an attempt to ameliorate the grain yield prediction accuracy, we integrated a spectral estimation of spike number, being a major wheat yield component, with genomic data. A machine learning approach was used for spike number estimation from canopy hyperspectral reflectance captured by an unmanned aerial vehicle. The spectral-based estimated spike number was utilized as a secondary trait in a multi-trait genomic selection, significantly improving grain yield prediction accuracy. Moreover, the ability to predict the spike number based on data from previous seasons implies that it could be applied to new trials at various scales, even in small plot sizes. Overall, we demonstrate here that incorporating a novel spectral-genomic chain-model workflow, which utilizes spectral-based phenotypes as a secondary trait, improves the predictive accuracy of wheat grain yield.

Improving potato AGB estimation to mitigate phenological stage impacts through depth features from hyperspectral data

The accurate estimation of above ground biomass (AGB) is valuable in grasping the status of potato growth and assessing yield. Spectral analysis techniques play an important role in AGB estimation by virtue of its non-destructive and rapid advantages. However, the models constructed based on spectral features at the entire growth stages due to differences in phenological stages, such as vegetation indices (VIs) and locations at fixed wavelengths, can be not applicable to single growth stages. Therefore, the main objective of this study is to improve the applicability of the AGB estimation model to mitigate the impacts of phenological stages by mining depth features from hyperspectral data. The canopy hyperspectral and AGB data of potato tuber formation, tuber growth and starch accumulation stages were measured in 2017–2019. The acquired canopy hyperspectral reflectance was smoothed by Savitzky-Golay (SG) filter, and the green edge (GE), the red edge (RE), red edge chlorophyll index (CIred-edge), optimized soil adjusted vegetation index (OSAVI), green normalized difference vegetation index (GNDVI), and normalized difference vegetation index (NDVI) were extracted. The spectral changes were analyzed at different growth stages of each year. To obtain features in response to AGB changes, a new approach is proposed to analyze the hidden depth features between sensitive wavelengths by combining a successive projections algorithm (SPA) and a long-short-term memory network (LSTM). The AGB estimation models were constructed by partial least squares regression (PLSR) based on traditional VIs + GE + RE, full spectrum, sensitive wavelengths obtained by SPA, and depth features obtained by SPA-LSTM. The results showed that GE, RE and VIs alone explained only 24 %-36 % of the variation in AGB. The accuracy of estimating AGB at sensitive wavelengths obtained after SPA was higher (R2 = 0.70, RMSE = 406 kg/hm2, and NRMSE = 27.14 %) than that of conventional VIs + GE + RE (R2 = 0.51, RMSE = 536 kg/hm2, and NRMSE = 35.81 %) and full spectrum (R2 = 0.63, RMSE = 452 kg/hm2, and NRMSE = 30.19 %) in the three-year dataset. The depth features extracted by the SPA-LSTM method proposed in this study achieved the best AGB estimation (R2 = 0.82, RMSE = 318 kg/hm2, and NRMSE = 21.25 %). The applicability of the model was validated in different years and growth stages. The R2 of each year was in the range of 0.48–0.77, the RMSE was in the range of 290–367 kg/hm2, and the NRMSE was in the range of 16.69 %-25.81 %. The R2 of each growth stage within three year was in the range of 0.48–0.78, the RMSE was in the range of 184–386 kg/hm2, and the NRMSE was in the range of 14.91 %-29.45 %. The method proposed in this study mitigates the impacts of phenological stage and improves the robustness and accuracy of the AGB estimation model, which provides remote sensing technical support for realizing potato growth monitoring and yield assessment in the field.

Dual sampling linear regression ensemble to predict wheat yield across growing seasons with hyperspectral sensing

In the domain of crop yield assessment, the combination of remote sensing and advanced regression techniques has gained significant attention. This study focused on predicting wheat yield across growing seasons using hyperspectral reflectance data. The performance of two decision tree-based algorithms, deep forest (DF) and random forest (RF), as well as a novel ensemble method called dual sampling linear regression ensemble (DSLRE), was evaluated. Canopy hyperspectral reflectance data were collected at two stages of wheat growth: early grain filling (EGF) and mid-grain filling (MGF), in three distinct growing environments. Prediction accuracy was assessed using the Pearson correlation coefficient (PCC) and mean absolute error (MAE). The results showed that the DF algorithm outperformed the RF algorithm in most scenarios. In modeling framework A, DF achieved the highest accuracy at the MGF stage using truncated reflectance (PCC = 0.60, MAE = 0.82 t/ha). In modeling framework B, DF performed better than RF in three out of four conditions, while RF exhibited higher prediction accuracy during the EGF stage using raw reflectance (PCC = 0.59, MAE = 1.48 t/ha). Moreover, the DSLRE algorithm outperformed DF and RF in six of the eight modeling scenarios and demonstrated robustness in coping with noisy features. The heritability analysis indicated that DSLRE predictions effectively captured the effects of genetic factors on yield variation. These findings illustrate the significance of the proposed DSLRE algorithm in improving yield prediction, providing valuable guidance for effective crop yield management.

Hyperspectral-to-image transform and CNN transfer learning enhancing soybean LCC estimation

Leaf chlorophyll content (LCC) is a distinct indicator of crop health status used to estimate nutritional stress, diseases, and pests. Thus, accurate LCC information can assist in the monitoring of crop growth. The combined use of hyperspectral and deep learning techniques (e.g., convolutional neural network [CNN] and transfer learning [TL]) can improve the performance of crop LCC estimation. We propose a hyperspectral-to-image transform (HIT) technique for converting canopy hyperspectral reflectance into 2D images. We designed a CNN architecture called LCCNet that fuses the deep and shallow features of CNNs to improve soybean LCC estimation. This study evaluated the combined use of hyperspectral remote sensing (RS), HIT, CNN, and TL techniques to estimate soybean LCC for multiple growth stages. The LCCNet was pre-trained based on a simulated dataset (n = 114,048) from the PROSAIL radiative transfer model (RTM) and used as prior knowledge for this work. The soybean canopy hyperspectral RS dataset (n = 910) was obtained using a FieldSpec 3 spectrometer. The knowledge gained while learning to estimate LCC from PROSAIL RTM was applied when estimating field soybean LCC (Dualex readings). TL was used to enhance the soybean estimation model, called the Soybean-LCCNet (RTM + HIT + CNN + TL) model. We tested the LCC (Dualex readings) estimation performance using (a) HIT + CNN, (b) LCCNet (RTM + HIT + CNN), (c) Soybean-LCCNet (RTM + HIT + CNN + TL), and (d) widely used LCC spectral features + partial least squares regression (PLSR). Four methods were ranked based on their LCC estimation performance: Soybean-LCCNet (R2 = 0.78, RMSE = 4.13 (Dualex readings)) > HIT + CNN (R2 = 0.75, RMSE = 4.41 (Dualex readings)) > PLSR-based method (R2 = 0.61, RMSE = 5.39 (Dualex readings)) > LCCNet (R2 = 0.53, RMSE = 7.11 (Dualex readings)). The main conclusions of this work are as follows: (1) HIT + CNN can provide a more robust LCC estimation performance than the widely used LCC SIs; (2) Fusing the deep and shallow features of CNNs can improve the performance of RS soybean LCC (Dualex readings) estimation; and (3) Soybean-LCCNet can reuse the CNN layer information of a pre-trained LCCNet based on a PROSAIL RTM dataset and improve the soybean LCC estimation performance.

From spectra to plant functional traits: Transferable multi-trait models from heterogeneous and sparse data

Large-scale information on several vegetation properties (‘plant traits’) is critical to assess ecosystem functioning, functional diversity and their role in the Earth system. Hyperspectral remote sensing of plant canopies offers a key tool to map multiple plant traits. However, we are still lacking generalized methods to translate hyperspectral reflectance into a suite of relevant plant traits across biomes, land cover and sensor types. The absence of globally representative data sets and the gap between the available reflectance data with corresponding in-situ measurements have hampered such approaches. In recent years, the scientific community acquired multiple data sets encompassing canopy hyperspectral reflectance and plant traits from different plant types and sensors. To combine these heterogeneous data sets, we propose three multi-trait modeling approaches based on Convolutional Neural Networks (CNNs) to simultaneously infer a broad set of 20 structural and chemical traits (e.g. leaf mass per area, leaf area index, pigments, nitrogen). The performance of these multi-trait CNN models predicting these traits was compared against single-trait CNN as well as single-trait partial least squares regression (PLSR). We found that the multi-trait CNNs performances significantly increased from single-trait CNNs (nRMSE = 0.027–19.61%) and the state-of-the-art PLSR models (nRMSE = 1.94–40.07%) across a broad range of vegetation types (crops, forest, tundra, grassland, shrubland) and sensor types. Thus, providing a single model for multiple traits not only proved to be computationally more efficient, but also more accurate, since it enabled the model to incorporate traits' co-variation. Despite the data heterogeneity of the merged data set, our models performances' were comparable or exceeded those of previous studies. Overall, this study highlights the potential of weakly supervised approaches to overcome the scarcity of in-situ measurements and take a step forward in creating efficient predictive models of multiple biochemical and biophysical vegetation properties.

Evaluation of Hyperspectral Monitoring Model for Aboveground Dry Biomass of Winter Wheat by Using Multiple Factors

The aboveground dry biomass (AGDB) of winter wheat can reflect the growth and development of winter wheat. The rapid monitoring of AGDB by using hyperspectral technology is of great significance for obtaining the growth and development status of winter wheat in real time and promoting yield increase. This study analyzed the changes of AGDB based on a winter wheat irrigation experiment. At the same time, the AGDB and canopy hyperspectral reflectance of winter wheat were obtained. The effect of spectral preprocessing algorithms such as reciprocal logarithm (Lg), multiple scattering correction (MSC), standardized normal variate (SNV), first derivative (FD), and second derivative (SD); sample division methods such as the concentration gradient method (CG), the Kennard–Stone method (KS), and the sample subset partition based on the joint X–Y distances method (SPXY); sample division ratios such as 1:1 (Ratio1), 3:2 (Ratio2), 2:1 (Ratio3), 5:2 (Ratio4), and 3:1 (Ratio5); dimension reduction algorithms such as uninformative variable elimination (UVE); and modeling algorithms such as partial least-squares regression (PLSR), stepwise multiple linear regression (SMLR), artificial neural network (ANN), and support vector machine (SVM) on the hyperspectral monitoring model of winter wheat AGDB was studied. The results showed that irrigation can improve the AGDB and canopy spectral reflectance of winter wheat. The spectral preprocessing algorithm can change the original spectral curve and improve the correlation between the original spectrum and the AGDB of winter wheat and screen out the bands of 1400 nm, 1479 nm, 1083 nm, 741 nm, 797 nm, and 486 nm, which have a high correlation with AGDB. The calibration sets and validation sets divided by different sample division methods and sample division ratios have different data-distribution characteristics. The UVE method can obviously eliminate some bands in the full-spectrum band. SVM is the best modeling algorithm. According to the universality of data, the better sample division method, sample division ratio, and modeling algorithm are SPXY, Ratio4, and SVM, respectively. Combined with the original spectrum and by using UVE to screen bands, a model with stable performance and high accuracy can be obtained. According to the particularity of data, the best model in this study is FD-CG-Ratio4-Full-SVM, for which the R2c, RMSEc, R2v, RMSEv, and RPD are 0.9487, 0.1663 kg·m−2, 0.7335, 0.3600 kg·m−2, and 1.9226, respectively, which can realize hyperspectral monitoring of winter wheat AGDB. This study can provide a reference for the rational irrigation of winter wheat in the field and provide a theoretical basis for monitoring the AGDB of winter wheat by using hyperspectral remote sensing technology.

Hyperspectral proximal sensing of leaf chlorophyll content of spring maize based on a hybrid of physically based modelling and ensemble stacking

Leaf chlorophyll content (LCC) is an important indicator for evaluating crop nutritional status and environmental stress. For the purpose of achieving rapid, non-destructive, and real-time LCC monitoring, experiments of spring maize under three drip irrigation levels and three film mulching conditions were conducted in the Shiyang River Basin of Northwest China in 2020 and 2021. We measured LCC and canopy hyperspectral reflectance during the maize growth period. Several two-layer ensemble stacking models were built to predict LCC from field hyperspectral measurement. The layer-1 base regression algorithms (i.e., support vector regression (SVR), back propagation neural network (BPNN), partial least square regression (PLSR)) were trained with synthetic data from the physically-based PROSAIL modelling. Multiple linear regression (MLR) was adopted for the layer-2 meta model, which was trained with the layer-1 modelling output and field observed LCC. The effects of the selected sensitive hyperspectral bands were also examined via three methods: recursive feature elimination (RFE), correlation analysis (CA) and variable importance in the projection analysis (VIP). Finally, the predictability and robustness of 16 models, composed of four types of input variables and four regression algorithms (three base regression algorithms and ensemble stacking algorithm), were tested using three field measured datasets. The results showed that the wavelengths at 698, 705, 693, 695, 697, 699, 700, 703, 709 and 712 nm were most sensitive for LCC. In the base regression models, the BPNN was the best performing model, followed by PLSR and SVR. The ensemble stacking models (R2: 0.78–0.91, RMSE: 3.27–9.34 μg cm−2, NRMSE: 7–20%) performed substantially better than the base models (R2: 0.28–0.68, RMSE: 10.53–30.83 μg cm−2 and NRMSE: 22–65%) did. No statistically significant difference was found in the stacking model performance between the two field years. We highly recommend using PROSAIL modelling to train machine-learning based stacking models for the proximal sensing of crop LCC.

Estimation of Chlorophyll Content in Soybean Crop at Different Growth Stages Based on Optimal Spectral Index

Chlorophyll is an important component of crop photosynthesis as it is necessary for the material exchange between crops and the atmosphere. The amount of chlorophyll present reflects the growth and health status of crops. Spectral technology is a feasible method for obtaining crop chlorophyll content. The first-order differential spectral index contains sufficient spectral information related to the chlorophyll content and has a high chlorophyll prediction ability. Therefore, in this study, the hyperspectral index data and chlorophyll content of soybean canopy leaves at different growth stages were obtained. The first-order differential transformation of soybean canopy hyperspectral reflectance data was performed, and five indices, highly correlated with soybean chlorophyll content at each growth stage, were selected as the optimal spectral index input. Four groups of model input variables were divided according to the following four growth stages: four-node (V4), full-bloom (R2), full-fruit (R4), and seed-filling stage (R6). Three machine learning methods, support vector machine (SVM), random forest (RF), and back propagation neural network (BPNN) were used to establish an inversion model of chlorophyll content at different soybean growth stages. The model was then verified. The results showed that the correlation coefficient between the optimal spectral index and chlorophyll content of soybean was above 0.5, the R2 period correlation coefficient was above 0.7, and the R4 period correlation coefficient was above 0.8. The optimal estimation model of soybean and chlorophyll content is established through the combination of the first-order differential spectral index and RF during the R4 period. The optimal estimation model validation set determination coefficient (R2) was 0.854, the root mean square error (RMSE) was 2.627, and the mean relative error (MRE) was 4.669, demonstrating high model accuracy. The results of this study can provide a theoretical basis for monitoring the growth and health of soybean crops at different growth stages.

Bayesian model averaging to improve the yield prediction in wheat breeding trials

Accurate pre-harvest prediction of wheat yield through secondary traits helps to facilitate plant breeding and reduce costs. Machine learning (ML) algorithms are increasingly applied to grain yield with remote sensing data. However, the performance of individual ML algorithms varies for different species in different environments due to different sources of uncertainty. This study proposed a novel wheat yield prediction framework based on canopy hyperspectral reflectance (350–2500 nm) and adopted the ensemble Bayesian model averaging (EBMA) method to improve model performance. To develop the yield prediction models, important bands extracted by the Boruta feature selection method were fed into four linear ML models and four nonlinear ML models. Meanwhile, Bayesian model averaging (BMA) weights obtained based on model cross-validation performance were used to combine the predictions of individual ML models. Compared to the best-performing individual model, the EBMA models obtained a weak accuracy improvement by integrating only the linear models or the nonlinear models. Additionally, the integration of two linear models and two non-linear models simultaneously was analyzed. Results indicate that most EBMA combinations of mixed linear and non-linear models achieved higher prediction accuracy than those integrating a single type of model and the best-performing individual model. The advantage of the EBMA method is that it produces a prediction distribution that reflects the uncertainty associated with deterministic predictions. With full consideration of the model diversity of ensemble members, the EBMA modeling framework provides an alternative method for predicting grain yield in plant breeding trials.

Combining novel feature selection strategy and hyperspectral vegetation indices to predict crop yield

BackgroundWheat is an important food crop globally, and timely prediction of wheat yield in breeding efforts can improve selection efficiency. Traditional yield prediction method based on secondary traits is time-consuming, costly, and destructive. It is urgent to develop innovative methods to improve selection efficiency and accelerate genetic gains in the breeding cycle.ResultsCrop yield prediction using remote sensing has gained popularity in recent years. This paper proposed a novel ensemble feature selection (EFS) method to improve yield prediction from hyperspectral data. For this, 207 wheat cultivars and breeding lines were grown under full and limited irrigation treatments respectively, and their canopy hyperspectral reflectance was measured at the flowering, early grain filling (EGF), mid grain filling (MGF), and late grain filling (LGF) stages. Then, 115 vegetation indices were extracted from the hyperspectral reflectance and combined with four feature selection methods, i.e., mean decrease impurity (MDI), Boruta, FeaLect, and RReliefF to train deep neural network (DNN) models for yield prediction. Next, a learning framework was developed by combining the predicted values of the selected and the full features using multiple linear regression (MLR). The results show that the selected features contributed to higher yield prediction accuracy than the full features, and the MDI method performed well across growth stages, with a mean R2 ranging from 0.634 to 0.666 (mean RMSE = 0.926–0.967 t ha−1). Also, the proposed EFS method outperformed all the individual feature selection methods across growth stages, with a mean R2 ranging from 0.648 to 0.679 (mean RMSE = 0.911–0.950 t ha−1).ConclusionsThe proposed EFS method can improve grain yield prediction from hyperspectral data and can be used to assist wheat breeders in earlier decision-making.

Using the SCOPE model for potato growth, productivity and yield monitoring under different levels of nitrogen fertilization

Most applications of remote sensing in agricultural crop monitoring use multispectral imaging techniques, but with upcoming hyperspectral missions, the opportunity arises to better estimate pigment absorption and crop structure by exploiting the full solar reflective spectrum. In this study, we demonstrate how hyperspectral time series can be used with the Soil Canopy Observation of Photochemistry and Energy fluxes (SCOPE) model to estimate crop yield variability among fields, crop varieties and nitrogen treatments generically, i.e. without a calibration with in situ , data. Field experiments were conducted in two potato fields in the Netherlands between May and September 2019. The fields were planted with five varieties of potato, under three nitrogen fertilization treatments. By fitting the model to the full VNIR-SWIR spectrum of measured hyperspectral reflectance, we retrieved the model input parameters of Leaf Area Index (LAI), leaf chlorophyll content (Cab) and leaf water content (Cw) and simulated the photosynthesis throughout the season using data of local Automatic Weather Stations (AWS). Statistical analysis of measured and retrieved traits of LAI, Cab and canopy water content showed that two fields responded differently to the treatments, exhibiting fewer classes than were expected based on the experimental design. Potato yield, which was estimated as the sum of photosynthesis flux multiplied by the harvest index of 0.64, correlated with the measured tuber dry weight with R 2 0.36 and RMSE 2.5 t ha −1 . This study demonstrates that even in the absence of crop or variety specific information, hyperspectral reflectance and local weather data ingested into SCOPE can explain a substantial part of the observed variability in yield among fields. • Time series of hyperspectral canopy reflectance were recorded in two potato fields. • The variation in traits was created by nitrogen fertilization treatments. • SCOPE model retrieved plant traits contained information about treatments. • SCOPE-simulated accumulated photosynthesis correlated with the final dry yield. • SCOPE model is more generic than dimensionality reduction methods.

Nitrogen Balance Index Prediction of Winter Wheat by Canopy Hyperspectral Transformation and Machine Learning

Nitrogen balance index (NBI) is an important indicator for scientific diagnostic and quantitative research on crop growth status. The quick and accurate assessment of NBI is necessary for farmers to make timely N management decisions. The objective of the study was to estimate winter wheat NBI based on canopy hyperspectral features between 400–1350 nm combined with machine learning (ML) methods in the individual and whole growth stages. In this study, 3 years of winter wheat plot experiments were conducted. Ground-level canopy hyperspectral reflectance and corresponding plant NBI values were measured during the jointing, booting, flowering and filling stages. Continuous removal spectra (CRS) and logarithmic transformation spectra (LOGS) were derived from the original canopy spectra. Sensitive bands and vegetation indices (VIs) highly correlated with NBI under different spectral transformations were selected as hyperspectral features to construct the NBI estimation models combined with ML algorithms. The study indicated that the spectral transformation significantly improved the correlation between the sensitive bands, VIs and the NBI. The correlation coefficient of the sensitive band in CRS in the booting stage increased by 27.87%, reaching −0.78. The leaf chlorophyll index (LCI) in LOGS had the highest correlation with NBI in the filling stage, reaching a correlation coefficient of −0.96. The NBI prediction accuracies based on the sensitive band combined with VIs were generally better than those based on the univariate hyperspectral feature, and the prediction accuracy of each growth stage was better than that of the whole growth stage. The random forest regression (RFR) method performed better than the support vector regression (SVR) and partial least squares regression (PLS) methods. The NBI estimation model based on the LOGS-RFR method in the filling stage could explain 95% of the NBI variability with relative prediction deviation (RPD) being 3.69. These results will provide a scientific basis for better nitrogen nutrition monitoring, diagnosis, and later for field management of winter wheat.

High-throughput screening of soybean di-nitrogen fixation and seed nitrogen content using spectral sensing

Symbiotic di-nitrogen fixation of grain legumes has a substantial impact on crop performance, harvest product quality, and nitrogen (N) balance of crop rotations, particularly under organic management regimes. In soybean breeding, selection for increased nitrogen fixation is desirable for improving seed protein content and N balance of cropping systems. However, the lack of high-throughput screening methods for direct measurement of N2 fixation rates prohibits practical breeding efforts. Therefore, hyperspectral canopy reflectance measurement as a field-based phenotyping method was evaluated in three environments for indirect estimation of N fixation and uptake of soil nitrogen in a set of early maturity soybean genotypes exhibiting a wide range in seed protein content. Reflectance spectra were collected in repeated measurements during flowering and early seed filling stages. Subsequently, various spectral reflectance indices (SRIs) were calculated for characterizing nitrogen accumulation of individual genotypes. Moreover, prediction models for seed protein content as an end-of-season target trait were developed utilizing full spectral information in partial-least-square regression (PLSR) models. A number of N-related SRIs calculated from spectral reflectance data recorded at the beginning of the seed filling stage were significantly correlated to seed protein content. The best prediction of seed protein content, however, was achieved in PLSR models (validation R2 = 0.805 across all three environments). Environments lower in initial soil mineral N content appeared as more favorable selection sites in terms of prediction accuracy, because N fixation is not masked by soil N uptake in such environments. Hyperspectral reflectance data proved to be a valuable method for determining genetic variation in crop N accumulation, which might be implemented in high-throughput screening protocols for N fixation in plant breeding programs.

Hyperspectral Reflectance Characteristics of Rice Canopies under Changes in Diffuse Radiation Fraction

To analyze the hyperspectral reflectance characteristics of rice canopies under changes in diffuse radiation fraction, experiments using different cover materials were performed in Nanjing, China, during 2016 and 2017. Each year, two treatments with different reduction ratios of diffuse radiation fraction but with similar shading rates were set in the field experiment: In T1, total solar radiation shading rate was 14.10%, and diffuse radiation fraction was 31.09%; in T2, total solar radiation shading rate was 14.42%, and diffuse radiation fraction was 39.98%, respectively. A non-shading treatment was included as a control (CK). Canopy hyperspectral reflectance, soil and plant analyzer development (SPAD), and leaf area index (LAI) were measured under shading treatments on different days after heading. The red-edge parameters (position, λ0; maximum amplitude, Dλ; area, α0; width, σ) were calculated, as well as the area, depth, and width of three absorption bands. The location of the first absorption band appeared in the range of 553–788 nm, and the second and third absorption bands appeared in the range of 874–1257 nm. The results show that the shading treatment had a significant effect on the rice canopy’s hyperspectral reflectance. Compared with CK, the canopy reflectance of T1 (the diffuse radiation fraction was 31.09%) and T2 (the diffuse radiation fraction was 39.98%) decreased in the visible light range (350–760 nm) and increased in the near-infrared range (800–1350 nm), while the red-edge parameters (λ0, Dλ, α0), SPAD, and LAI increased. On the other hand, under shading treatment, the increase in diffuse radiation fraction also had a significant impact on the hyperspectral spectra of the rice canopy, especially at 14 days after heading. Compared with T1, the green peak (550 nm) of T2 reduced by 16.12%, and the average reflectance at 800–900 nm increased by 10%. Based on correlation analysis, it was found that these hyperspectral reflectance characteristics were mainly due to the increase in SPAD (2.31%) and LAI (7.62%), which also led to the increase in Dλ (8.70%) and α0 (13.89%). Then, the second and third absorption features of T2 were significantly different from that of T1, which suggests that the change in diffuse radiation fraction could affect the process of water vapor absorption by rice.

Estimation of leaf nitrogen levels in sugarcane using hyperspectral models

ABSTRACT: Sugarcane is a good source of renewable energy and helps reduce the emission of greenhouse gases. Nitrogen has a critical role in plant growth; therefore,estimating nitrogen levels is essential, and remote sensing can improve fertilizer management. This field study selects wavelengths from hyperspectral data on a sugarcane canopy to generate models for estimating leaf nitrogen concentrations. The study was carried out in the municipalities of Piracicaba, Jaú, and Santa Maria da Serra, state of São Paulo, in the 2013/2014 growing season. The experiments were carried out using a completely randomized block design with split plots (three sugarcane varieties per plot [variety SP 81-3250 was common to all plots] and four nitrogen concentrations [0, 50, 100, and 150 kgha-1] per subplot) and four repetitions. The wavelengths that best correlated with leaf nitrogen were selected usingsparse partial least square regression. The wavelength regionswere combinedby stepwise multiple linear regression. Spectral bands in the visible (700-705 nm), red-edge (710-720 nm), near-infrared (725, 925, 955, and 980 nm), and short-wave infrared (1355, 1420, 1595, 1600, 1605, and 1610 nm) regions were identified. The R² and RMSE of the model were 0.50 and 1.67 g.kg-1, respectively. The adjusted R² and RMSE of the models for Piracicaba, Jaú, and Santa Maria were 0.31 (unreliable) and 1.30 g.kg-1, 0.53 and 1.96 g.kg-1, and 0.54 and 1.46 g.kg-1, respectively. Our results showed that canopy hyperspectral reflectance can estimate leaf nitrogen concentrations and manage nitrogen application in sugarcane.

Improvement of Wheat Grain Yield Prediction Model Performance Based on Stacking Technique

Crop growth and development is a dynamic and complex process, and the essence of yield formation is the continuous accumulation of photosynthetic products from multiple fertility stages. In this study, a new stacking method for integrating multiple growth stages information was proposed to improve the performance of the winter wheat grain yield (GY) prediction model. For this purpose, crop canopy hyperspectral reflectance and leaf area index (LAI) data were obtained at the jointing, flagging, anthesis and grain filling stages. In this case, 15 vegetation indices and LAI were used as input features of the elastic network to construct GY prediction models for single growth stage. Based on Stacking technique, the GY prediction results of four single growth stages were integrated to construct the ensemble learning framework. The results showed that vegetation indices coupled LAI could effectively overcome the spectral saturation phenomenon, the validated R2 of each growth stage was improved by 10%, 22.5%, 3.6% and 10%, respectively. The stacking method provided more stable information with higher prediction accuracy than the individual fertility results (R2 = 0.74), and the R2 of the model validation phase improved by 236%, 51%, 27.6%, and 12.1%, respectively. The study can provide a reference for GY prediction of other crops.

Hyperspectral Estimation of Winter Wheat Leaf Area Index Based on Continuous Wavelet Transform and Fractional Order Differentiation.

Leaf area index (LAI) is highly related to crop growth, and the traditional LAI measurement methods are field destructive and unable to be acquired by large-scale, continuous, and real-time means. In this study, fractional order differential and continuous wavelet transform were used to process the canopy hyperspectral reflectance data of winter wheat, the fractional order differential spectral bands and wavelet energy coefficients with more sensitive to LAI changes were screened by correlation analysis, and the optimal subset regression and support vector machine were used to construct the LAI estimation models for different growth stages. The precision evaluation results showed that the LAI estimation models constructed by using wavelet energy coefficients combined with a support vector machine at the jointing stage, fractional order differential combined with support vector machine at the booting stage, and wavelet energy coefficients combined with optimal subset regression at the flowering and filling stages had the best prediction performance. Among these, both flowering and filling stages could be used as the best growth stages for LAI estimation with modeling and validation R2 of 0.87 and 0.71, 0.84 and 0.77, respectively. This study can provide technical reference for LAI estimation of crops based on remote sensing technology.

Monitoring of Wheat Powdery Mildew under Different Nitrogen Input Levels Using Hyperspectral Remote Sensing

Both wheat powdery mildew severities and nitrogen input levels can lead to changes in spectral reflectance, but they have been rarely studied simultaneously for their effect on spectral reflectance. To determine the effects and influences of different nitrogen input levels on monitoring wheat powdery mildew and estimating yield by near-ground hyperspectral remote sensing, Canopy hyperspectral reflectance data acquired at Feekes growth stage (GS) 10.5.3, 10.5.4, and 11.1 were used to monitor wheat powdery mildew and estimate grain yield under different nitrogen input levels during the 2016–2017, 2017–2018, 2018–2019 and 2019–2020 seasons. The relationships of powdery mildew and grain yield with vegetation indices (VIs) derived from spectral reflectance data across the visible (VIS) and near-infrared (NIR) regions of the spectrum were studied. The relationships of canopy spectral reflectance or first derivative spectral reflectance with powdery mildew did not differ under different nitrogen input levels. However, the dynamics of VIs differed in their sensitivities to nitrogen input levels, disease severity, grain yield, The area of the red edge peak (Σdr680–760 nm) was a better overall predictor for both disease severity and grain yield through linear regression models. The slope parameter estimates did not differ between the two nitrogen input levels at each GSs. Hyperspectral indices can be used to monitor wheat powdery mildew and estimate grain yield under different nitrogen input levels, but such models are dependent on GS and year, further research is needed to consider how to incorporate the growth stage and year-to-year variation into future applications.

Better revisiting chlorophyll content retrieval with varying senescent material and solar-induced chlorophyll fluorescence simulation on paddy rice during the entire growth stages

Solar-induced chlorophyll fluorescence at 760 nm (SIF) is a promising proxy of photosynthesis and can help improving plant stress monitoring. The Soil-Canopy Observation of Photochemistry and Energy fluxes (SCOPE) model combines radiative transfer and enzyme kinetics of photosynthesis and is widely used to interpret SIF at different temporal and spatial scales. In this study, growing season canopy hyperspectral reflectance between 400 nm and 900 nm was used to retrieve chlorophyll content (Cab) and leaf inclination (LIDFa) using radiative transfer models (RTMs) combined with the shuffled complex evolution-University of Arizona (SCE-UA) method. These parameters were then used to simulate diurnal and seasonal trends of SIF for paddy rice. The results showed that the accuracy of Cab retrieval was improved when the variation in senescent material (Cs) was considered, especially in the later growth stages. The SCOPE model was able to reliably interpret the diurnal cycle and seasonal trend of SIF with a correlation coefficient of 0.92 and RMSE of 0.12 w m−2 sr-1 um−1. Our results revealed that the SCOPE model provides a promising method for interpreting SIF variations but its accuracy should be evaluated in different growth stages. This will serve as a significant reference for detecting plant photosynthetic activity and physiological traits at different growth stages.