Yongchao Tian

Monitoring leaf nitrogen in wheat using canopy reflectance spectra

Non-destructive monitoring of leaf nitrogen (N) status can assist in growth diagnosis, N management and productivity forecast in field crops. The objectives of this study were to determine the relationships of leaf nitrogen concentration on a leaf dry weight basis (LNC) and leaf nitrogen accumulation per unit soil area (LNA) to ground-based canopy reflectance spectra, and to derive regression equations for monitoring N nutrition status in wheat (Triticum aestivum L.). Four field experiments were conducted with different N application rates and wheat cultivars across four growing seasons, and time-course measurements were taken on canopy spectral reflectance, LNC and leaf dry weights under the various treatments. In these studies, LNC and LNA in wheat increased with increasing N fertilization rates. The canopy reflectance differed significantly under varied N rates, and the pattern of response was consistent across the different cultivars and years. Overall, an integrated regression equation of LNC to norm...

Quantitative relationships of leaf nitrogen status to canopy spectral reflectance in rice

Non-destructive and quick methods for assessing leaf nitrogen (N) status are helpful for precision N management in field crops. The present study was conducted to determine the quantitative relationships of leaf N concentration on a leaf dry weight basis (LNC) and leaf N accumulation per unit soil area (LNA) to ground-based canopy spectral reflectance in rice (Oryza sativa L.). Time-course measurements were taken on canopy spectral reflectance, LNC, and leaf dry weights, with 4 field experiments under different N application rates and rice cultivars across 4 growing seasons. All possible ratio vegetation indices (RVI), difference vegetation indices (DVI), and normalised difference vegetation indices (NDVI) of key wavebands from the MSR16 radiometer were calculated. The results showed that LNC, LNA, and canopy reflectance spectra all markedly varied with N rates, with consistent change patterns among different rice cultivars and experiment years. There were highly significant linear correlations between LNC and canopy reflectance in the visible region from 560 to 710 nm (|r| > 0.85), between LNA and canopy reflectance from 760 to 1100 nm (|r| > 0.79), and from 460 to 710 nm wavelengths (|r| > 0.70). Among all possible RVI, DVI, and NDVI of key wavebands from the MSR16 radiometer, NDVI of 1220 and 710 nm was most highly correlated to LNC, and RVI of 950 and 660 nm and RVI of 950 and 680 nm were the best spectral indices for quantitative monitoring of LNA in rice. The average relative root mean square errors (RRMSE) between the predicted LNC and LNA and the observed values with independent data were no more than 11% and 25%, respectively. These results indicated that the canopy spectral reflectance can be potentially used for non-destructive and real-time monitoring of leaf N status in rice.

Comparison of different hyperspectral vegetation indices for canopy leaf nitrogen concentration estimation in rice

Background and aimsVariations in the water and soil background in the signal path can cause variations in canopy spectral reflectance, which leads to uncertainty in estimating the canopy nitrogen (N) status. The primary objective of this study was to explore the optimum vegetation indices that were highly correlated with canopy leaf N concentration (LNC) but less influenced by the canopy leaf area index (LAI) and vegetation coverage (VC) in rice.MethodsA systematic analysis of the quantitative relationships between various hyperspectral vegetation indices and LNC, VC and LAI was conducted based on 4-year rice field experiments using different rice varieties, N rates and planting densities. New spectral indices were derived to estimate LNC in rice under variable vegetation coverage.ResultsAlthough the newly developed simple green ratio indices, SR (R553, R537) and SR (R545, R538), and the three-band index (R605-R521-R682)/(R605+R521+R682) correlated well with the LNC. Only SR (R553, R537) was less influenced by VC/LAI and showed a stable performance in both the independent calibration and validation datasets. For the published indices tested in the present study, NDVIg-b and ND (R503, R483) showed a good predictive ability for the LNC. However, both of these indices and other published indices were found to be significantly dominated by the VC/LAI.ConclusionSR (R553, R537) was the best index to reliably estimate the LNC in rice under various cultivation conditions, and is recommended for this use. However, other spectral indices need to be examined to determine if they influenced by factors such as VC/LAI. Such studies will improve the applicability of these indices to different types of rice cultivars and production systems.

Estimating leaf chlorophyll content using red edge parameters.

Abstract Hyperspectral remote sensing makes it possible to non-destructively monitor leaf chlorophyll content (LCC). This study characterized the geometric patterns of the first derivative reflectance spectra in the red edge region of rapeseed ( Brassica napus L.) and wheat ( Triticum aestivum L.) crops. The ratio of the red edge area less than 718 nm to the entire red edge area was negatively correlated with LCC. This finding allowed the construction of a new red edge parameter, defined as red edge symmetry (RES). Compared to the commonly used red edge parameters (red edge position, red edge amplitude, and red edge area), RES was a better predictor of LCC. Furthermore, RES was easily calculated using the reflectance of red edge boundary wavebands at 675 and 755 nm ( R 675 and R 755 ) and reflectance of red edge center wavelength at 718 nm ( R 718 ), with the equation RES = ( R 718 -R 675 ) / ( R 755 - R 675 ). In addition, RES was simulated effectively with wide wavebands from the airborne hyperspectral sensor AVIRIS and satellite hyperspectral sensor Hyperion. The close relationships between the simulated RES and LCC indicated a high feasibility of estimating LCC with simulated RES from AVIRIS and Hyperion data. This made RES readily applicable to common airborne and satellite hyperspectral data derived from AVIRIS and Hyperion sources, as well as ground-based spectral reflectance data.

Evaluation of Six Algorithms to Monitor Wheat Leaf Nitrogen Concentration

The rapid and non-destructive monitoring of the canopy leaf nitrogen concentration (LNC) in crops is important for precise nitrogen (N) management. Nowadays, there is an urgent need to identify next-generation bio-physical variable retrieval algorithms that can be incorporated into an operational processing chain for hyperspectral satellite missions. We assessed six retrieval algorithms for estimating LNC from canopy reflectance of winter wheat in eight field experiments. These experiments represented variations in the N application rates, planting densities, ecological sites and cultivars and yielded a total of 821 samples from various places in Jiangsu, China over nine consecutive years. Based on the reflectance spectra and their first derivatives, six methods using different numbers of wavelengths were applied to construct predictive models for estimating wheat LNC, including continuum removal (CR), vegetation indices (VIs), stepwise multiple linear regression (SMLR), partial least squares regression (PLSR), artificial neural networks (ANNs), and support vector machines (SVMs). To assess the performance of these six methods, we provided a systematic evaluation of the estimation accuracies using the six metrics that were the coefficients of determination for the calibration (R2C) and validation (R2V) sets, the root mean square errors of prediction (RMSEP) for the calibration and validation sets, the ratio of prediction to deviation (RPD), the computational efficiency (CE) and the complexity level (CL). The following results were obtained: (1) For the VIs method, SAVI(R1200, R705) produced a more accurate estimation of the LNC than other indices, with R²C, R²V, RMSEP, RPD and CE values of 0.844, 0.795, 0.384, 2.005 and 0.10 min, respectively; (2) For the SMLR, PLSR, ANNs and SVMs methods, the SVMs using the first derivative canopy spectra (SVM-FDS) offered the best accuracy in terms of R²C, R²V, RMSEP, RPD, and CE, at 0.96, 0.78, 0.37, 2.02, and 21.17, respectively; (3) The PLSR-FDS, ANN-OS and SVM-FDS methods yield similar accuracies if the CE and CL are not considered, however, ANNs and SVMs performed better on calibration set than the validation set which indicate that we should take more caution with the two methods for over-fitting. Except PLS method, the performance for most methods did not enhance when the spectrum were operated by the first derivative. Moreover, the evaluation of the robustness demonstrates that SVM method may be better suited than the other methods to cope with potential confounding factors for most varieties, ecological site and growth stage; (4) The prediction accuracy was found to be higher when more wavelengths were used, though at the cost of a lower CE. The findings are of interest to the remote sensing community for the development of improved inversion schemes for hyperspectral applications concerning other types of vegetation. The examples provided in this paper may also serve to illustrate the advantages and shortcomings of empirical hyperspectral models for mapping important vegetation biophysical properties of other crops.

Monitoring leaf photosynthesis with canopy spectral reflectance in rice

Non-destructive and rapid method for assessment of leaf photosynthetic characteristics is needed to support photosynthesis modelling and growth monitoring in crop plants. We determined the quantitative relationships between leaf photosynthetic characteristics and canopy spectral reflectance under different water supply and nitrogen application rates. The responses of reflectance at red radiation (wavelength 680 nm) to different water contents and nitrogen rates were parallel to those of leaf net photosynthetic rate (PN). The relationships of reflectance at 680 nm and ratio index of R(810,680) (near infrared/red, NIR/R) to PN of different leaf positions and leaf layers in rice indicated that the top two full leaves were the best leaf positions for quantitative monitoring of leaf PN with remote sensing technique, and the ratio index R(810,680) was the best ratio index for evaluating leaf photosynthetic characteristics in rice. Testing of the models with independent data sets indicated that R(810,680) could well estimate PN of top two leaves and canopy leaf photosynthetic potential in rice, with the root mean square error of 0.25, 0.16, and 4.38, respectively. Hence R(810,680) can be used to monitor leaf photosynthetic characteristics at different growth stages of rice under diverse growing conditions.

Integrating remotely sensed leaf area index and leaf nitrogen accumulation with RiceGrow model based on particle swarm optimization algorithm for rice grain yield assessment

Abstract A regional rice (Oryza sativa) grain yield prediction technique was proposed by integration of ground-based and spaceborne remote sensing (RS) data with the rice growth model (RiceGrow) through a new particle swarm optimization (PSO) algorithm. Based on an initialization/parameterization strategy (calibration), two agronomic indicators, leaf area index (LAI) and leaf nitrogen accumulation (LNA) remotely sensed by field spectra and satellite images, were combined to serve as an external assimilation parameter and integrated with the RiceGrow model for inversion of three model management parameters, including sowing date, sowing rate, and nitrogen rate. Rice grain yield was then predicted by inputting these optimized parameters into the reinitialized model. PSO was used for the parameterization and regionalization of the integrated model and compared with the shuffled complex evolution–University of Arizona (SCE-UA) optimization algorithm. The test results showed that LAI together with LNA as the integrated parameter performed better than each alone for crop model parameter initialization. PSO also performed better than SCE-UA in terms of running efficiency and assimilation results, indicating that PSO is a reliable optimization method for assimilating RS information and the crop growth model. The integrated model also had improved precision for predicting rice grain yield.

Common Spectral Bands and Optimum Vegetation Indices for Monitoring Leaf Nitrogen Accumulation in Rice and Wheat

Real-time monitoring of nitrogen status in rice and wheat plant is of significant importance for nitrogen diagnosis, fertilization recommendation, and productivity prediction. With 11 field experiments involving different cultivars, nitrogen rates, and water regimes, time-course measurements were taken of canopy hyperspectral reflectance between 350-2 500 nm and leaf nitrogen accumulation (LNA) in rice and wheat. A new spectral analysis method through the consideration of characteristics of canopy components and plant growth status varied with phenological growth stages was designed to explore the common central bands in rice and wheat. Comprehensive analyses were made on the quantitative relationships of LNA to soil adjusted vegetation index (SAVI) and ratio vegetation index (RVI) composed of any two bands between 350-2 500 nm in rice and wheat. The results showed that the ranges of indicative spectral reflectance were largely located in 770-913 and 729-742 nm in both rice and wheat. The optimum spectral vegetation index for estimating LNA was SAVI (R822, R738) during the early-mid period (from jointing to booting), and it was RVI (R822, R738) during the mid-late period (from heading to filling) with the common central bands of 822 and 738 nm in rice and wheat. Comparison of the present spectral vegetation indices with previously reported vegetation indices gave a satisfactory performance in estimating LNA. It is concluded that the spectral bands of 822 and 738 nm can be used as common reflectance indicators for monitoring leaf nitrogen accumulation in rice and wheat.

Analysis of Common Canopy Reflectance Spectra for Indicating Leaf Nitrogen Concentrations in Wheat and Rice

Abstract Non-destructive monitoring and diagnosis of plant nitrogen (N) concentration are of significant importance for precise N management and productivity forecasting in field crops. The present study was conducted to identify the common spectra wavebands and canopy reflectance spectral parameters for indicating leaf nitrogen concentration (LNC, mg N g-1 DW) and to determine quantitative relationships of LNC to canopy reflectance spectra in both rice (Oryza sativa L.) and wheat (Triticum aestivum L.). Ground-based canopy spectral reflectance and LNC were measured with seven field experiments consisting of seven different wheat cultivars and five different rice cultivars and varied N fertilization levels across three growing seasons for wheat and four growing seasons for rice. All possible ratio vegetation indices (RVI), difference vegetation indices (DVI), and normalized difference vegetation indices (NDVI) of key wavebands from the MSR16 radiometer were calculated. The results showed that LNC of wheat and rice increased with increasing N fertilization rates. Canopy reflectance, however, was a more complicated relationship under different N application rates. In the near infrared portion of the spectrum (760−1220 nm), canopy spectral reflectance increased with increasing N supply, whereas in the visible region (460−710 nm), canopy reflectance decreased with increasing N supply. For both rice and wheat, LNC was best estimated at 610, 660 and 680 nm. Among all possible RVI, DVI and NDVI of key bands from the MSR16 radiometer, NDVI(1220, 610) and RVI(1220, 610) were most highly correlated to LNC in both wheat and rice. In addition, the correlations of NDVI(1220, 610) and RVI(1220, 610) to LNC were found to be higher than those of individual wavebands at 610, 660 and 680 nm in both wheat and rice. Thus LNC in both wheat and rice could be indicated with common wavebands and vegetation indices, but separate regression equations are necessary for precisely describing the dynamic change patterns of LNC in wheat and rice. When independent data were fit to the derived equations, the root mean square error (RMSE) values for the predicted LNC with NDVI(1220, 610) and RVI(1220, 610) relative to the observed values were 10.50% and 10.52% in wheat, and 13.04% and 12.61% in rice, respectively, indicating a good fit. These results should improve the knowledge on non-destructive monitoring of leaf N status in cereal crops.

Optimal Leaf Positions for SPAD Meter Measurement in Rice

The Soil Plant Analysis Development (SPAD) chlorophyll meter is one of the most commonly used diagnostic tools to measure crop nitrogen status. However, the measurement method of the meter could significantly affect the accuracy of the final estimation. Thus, this research was undertaken to develop a new methodology to optimize SPAD meter measurements in rice (Oryza sativa L.). A flatbed color scanner was used to map the dynamic chlorophyll distribution and irregular leaf shapes. Calculus algorithm was adopted to estimate the potential positions for SPAD meter measurement along the leaf blade. Data generated by the flatbed color scanner and SPAD meter were analyzed simultaneously. The results suggested that a position 2/3 of the distance from the leaf base to the apex (2/3 position) could represent the chlorophyll content of the entire leaf blade, as indicated by the relatively low variance of measurements at that position. SPAD values based on di-positional leaves and the extracted chlorophyll a and b contents were compared. This comparison showed that the 2/3 position on the lower leaves tended to be more sensitive to changes in chlorophyll content. Finally, the 2/3 position and average SPAD values of the fourth fully expanded leaf from the top were compared with leaf nitrogen concentration. The results showed the 2/3 position on that leaf was most suitable for predicting the nitrogen status of rice. Based on these results, we recommend making SPAD measurements at the 2/3 position on the fourth fully expanded leaf from the top. The coupling of dynamic chlorophyll distribution and irregular leaf shapes information can provide a promising approach for the calibration of SPAD meter measurement, which can further benefit the in situ nitrogen management by providing reliable estimation of crops nitrogen nutrition status.