Xingang Xu

Exploring the Best Hyperspectral Features for LAI Estimation Using Partial Least Squares Regression

Abstract: The use of spectral features to estimate leaf area index (LAI) is generally considered a challenging task for hyperspectral data. In this study, the hyperspectral reflectance of winter wheat was se lected to optimize the selection of spectral features and to evaluate their performance in modeling LAI at various grow th stages during 2008 and 2009. We extracted hyperspectral featur es using different techniques, including reflectance spectra and first derivative spectra, absorption and reflectance position and vegetation indices. In order to find the best subset of features with the best predictive accuracy, partial least squares regression (PLSR) and variable importance in projection (VIP) were applied to estimated LAI values. The results indicated that the red edge–NIR spectral region (680 nm–1300 nm) was the most sensitive to LAI. Most features in this region exhibited a high correlation with LAI and had higher VIP values, especially the first derivative waveband at 750 nm (

Assessment of the AquaCrop model for use in simulation of irrigated winter wheat canopy cover, biomass, and grain yield in the North China Plain.

Improving winter wheat water use efficiency in the North China Plain (NCP), China is essential in light of current irrigation water shortages. In this study, the AquaCrop model was used to calibrate, and validate winter wheat crop performance under various planting dates and irrigation application rates. All experiments were conducted at the Xiaotangshan experimental site in Beijing, China, during seasons of 2008/2009, 2009/2010, 2010/2011 and 2011/2012. This model was first calibrated using data from 2008/2009 and 2009/2010, and subsequently validated using data from 2010/2011 and 2011/2012. The results showed that the simulated canopy cover (CC), biomass yield (BY) and grain yield (GY) were consistent with the measured CC, BY and GY, with corresponding coefficients of determination (R2) of 0.93, 0.91 and 0.93, respectively. In addition, relationships between BY, GY and transpiration (T), (R2 = 0.57 and 0.71, respectively) was observed. These results suggest that frequent irrigation with a small amount of water significantly improved BY and GY. Collectively, these results indicate that the AquaCrop model can be used in the evaluation of various winter wheat irrigation strategies. The AquaCrop model predicted winter wheat CC, BY and GY with acceptable accuracy. Therefore, we concluded that AquaCrop is a useful decision-making tool for use in efforts to optimize wheat winter planting dates, and irrigation strategies.

Combined Multi-Temporal Optical and Radar Parameters for Estimating LAI and Biomass in Winter Wheat Using HJ and RADARSAR-2 Data

Leaf area index (LAI) and biomass are frequently used target variables for agricultural and ecological remote sensing applications. Ground measurements of winter wheat LAI and biomass were made from March to May 2014 in the Yangling district, Shaanxi, Northwest China. The corresponding remotely sensed data were obtained from the earth-observation satellites Huanjing (HJ) and RADARSAT-2. The objectives of this study were (1) to investigate the relationships of LAI and biomass with several optical spectral vegetation indices (OSVIs) and radar polarimetric parameters (RPPs), (2) to estimate LAI and biomass with combined OSVIs and RPPs (the product of OSVIs and RPPs (COSVI-RPPs)), (3) to use multiple stepwise regression (MSR) and partial least squares regression (PLSR) to test and compare the estimations of LAI and biomass in winter wheat, respectively. The results showed that LAI and biomass were highly correlated with several OSVIs (the enhanced vegetation index (EVI) and modified triangular vegetation index 2 (MTVI2)) and RPPs (the radar vegetation index (RVI) and double-bounce eigenvalue relative difference (DERD)). The product of MTVI2 and DERD (R2 = 0.67 and RMSE = 0.68, p < 0.01) and that of MTVI2 and RVI (R2 = 0. 68 and RMSE = 0.65, p < 0.01) were strongly related to LAI, and the product of the optimized soil adjusted vegetation index (OSAVI) and DERD (R2 = 0.79 and RMSE = 148.65 g/m2, p < 0.01) and that of EVI and RVI (R2 = 0. 80 and RMSE = 146.33 g/m2, p < 0.01) were highly correlated with biomass. The estimation accuracy of LAI and biomass was better using the COSVI-RPPs than using the OSVIs and RPPs alone. The results revealed that the PLSR regression equation better estimated LAI and biomass than the MSR regression equation based on all the COSVI-RPPs, OSVIs, and RPPs. Our results indicated that the COSVI-RPPs can be used to robustly estimate LAI and biomass. This study may provide a guideline for improving the estimations of LAI and biomass of winter wheat using multisource remote sensing data.

A method of estimating soil moisture based on the linear decomposition of mixture pixels

Abstract The objective of this study was to estimate soil moisture with soil red and nir (near-infrared) band reflectance from TM/ETM+ remotely sensed images acquired over vegetated fields. Based on linear decomposition algorithm of mixture pixel, first the soil reflectance from red–nir bands were directly and computationally derived by combining soil line equation with a developed empirical relationship between vegetation canopy and mixture pixel reflectance in red–nir spectral feature space. Then, a remote sensing image from TM with measurement data from experimental fields in Beijing, China, was used to establish the retrieval relationships between soil moisture and soil reflectance from the red and nir bands, and the results showed that the retrieval of soil moisture was better with nir band reflectance than that of red. Finally, the soil moisture retrieved method was further evaluated and validated with two images from ETM + and ground measurements from fields in Walnut Creek, America, and the analysis showed that the proposed method could be used to monitor soil moisture well, with the correlation coefficient exceeding 0.80. The preliminary results with such acceptable accuracy indicate that the method of estimating soil moisture based on the linear decomposition of mixture pixels is reasonable and suitable for being widely applied in different temporal and spatial scaled fields.

Estimation of Winter Wheat Biomass and Yield by Combining the AquaCrop Model and Field Hyperspectral Data

Knowledge of spatial and temporal variations in crop growth is important for crop management and stable crop production for the food security of a country. A combination of crop growth models and remote sensing data is a useful method for monitoring crop growth status and estimating crop yield. The objective of this study was to use spectral-based biomass values generated from spectral indices to calibrate the AquaCrop model using the particle swarm optimization (PSO) algorithm to improve biomass and yield estimations. Spectral reflectance and concurrent biomass and yield were measured at the Xiaotangshan experimental site in Beijing, China, during four winter wheat-growing seasons. The results showed that all of the measured spectral indices were correlated with biomass to varying degrees. The normalized difference matter index (NDMI) was the best spectral index for estimating biomass, with the coefficient of determination (R2), root mean square error (RMSE), and relative RMSE (RRMSE) values of 0.77, 1.80 ton/ha, and 25.75%, respectively. The data assimilation method (R2 = 0.83, RMSE = 1.65 ton/ha, and RRMSE = 23.60%) achieved the most accurate biomass estimations compared with the spectral index method. The estimated yield was in good agreement with the measured yield (R2 = 0.82, RMSE = 0.55 ton/ha, and RRMSE = 8.77%). This study offers a new method for agricultural resource management through consistent assessments of winter wheat biomass and yield based on the AquaCrop model and remote sensing data.

Newly Combined Spectral Indices to Improve Estimation of Total Leaf Chlorophyll Content in Cotton

The total leaf chlorophyll content (TLCC) provides valuable information about the physiological status of crops. The objectives of this study were 1) to analyze the leaf area index (LAI) and soil factors that influences the estimation of TLCC using the PROSAIL model, which is a combination of the PROSPECT leaf model and the SAIL canopy model; 2) to propose newly combined spectral indices that reduce the influence of LAI and soil factors in order to improve the TLCC estimation; and 3) to test and validate the relationship between TLCC and the newly combined spectral indices. Ground-based hyperspectral data and concurrent TLCC parameters of samples were acquired at the Shihezi University Experiment Site, Xinjiang Province, China, during the 2009 and 2010 cotton growing seasons. The results showed that the newly combined spectral indices [double-peak canopy nitrogen index I (DCNI I), the ratio of the structure insensitive pigment index to the ratio vegetation index III (SIPI/RVI III), the ratio of the plant pigment ratio to the normalized difference vegetation index (PPR/NDVI), and the modified MERIS terrestrial chlorophyll index (MMTCI)] were more sensitive to chlorophyll and more resistant to LAI than the PPR, SIPI, and MERIS terrestrial chlorophyll index alone. In this study, DCNI I proved to be the best spectral index for estimating chlorophyll content, with determination coefficients (R2) and root mean square error (RMSE) values of 0.80 and 8.31μg · cm- 2, respectively. PPR/NDVI was also strongly correlated with chlorophyll content, with corresponding R2 and RMSE values of 0.79 and 9.45 μg · cm-2, respectively. This study concluded that DCNI I and PPR/NDVI, in association with indices related to nitrogen, have good potential for assessing nitrogen content.

Estimating winter wheat Triticum aestivum LAI and leaf chlorophyll content from canopy reflectance data by integrating agronomic prior knowledge with the PROSAIL model

Leaf area index (LAI) and leaf chlorophyll content (LCC) are major considerations in management decisions, agricultural planning, and policy-making. When a radiative transfer model (RTM) was used to retrieve these biophysical variables from remote-sensing data, the ill-posed problem was unavoidable. In this study, we focused on the use of agronomic prior knowledge (APK), constructing the relationship between LAI and LCC, to restrict and mitigate the ill-posed inversion results. For this purpose, the inversion results obtained using the SAILH+PROSPECT (PROSAIL) canopy reflectance model alone (no agronomic prior knowledge, NAPK) and those linked with APK were compared. The results showed that LAI inversion had high accuracy. The validation results of the root mean square error (RMSE) between measured and estimated LAI were 0.74 and 0.69 for NAPK and APK, respectively. Compared with NAPK, APK improved LCC estimation; the corresponding RMSE values of NAPK and APK were 13.36 µg cm–2 and 9.35 µg cm–2, respectively. Our analysis confirms the operational potential of PROSAIL model inversion for the retrieval of biophysical variables by integrating APK.

Assessing Rice Chlorophyll Content with Vegetation Indices from Hyperspectral Data

Leaf chlorophyll content is not only an important biochemical parameter for determinating the capacity of rice photosynthesis, but also a good indicator of crop stress, nutritional state. Due to the reliable, operational and non-destructive advantages, hyperspectral remote sensing plays a significant role for assessing and monitoring chlorophyll content. In the study, a few of typical vegetation indices (VI) with the combination of 670nm and 800nm band reflectance, Normalized Difference Vegetation Index (NDVI), Modified Simple Ratio index (MSR), Modified Chlorophyll Absorption Ratio Index (MCARI), Transformed Chlorophyll Absorption Ratio Index (TCARI), and Optimized Soil-Adjusted Vegetation Index (OSAVI) are modified by using 705nm and 750nm band reflectance so as to reduce the effect of spectral saturation in 660-680nm absorptive band region, and then used to assess the rice chlorophyll content. The result shows that the five mentioned VIs have better correlation with rice chlorophyll content while using 705nm and 750nm. In addition, in the study the Weight optimization combination (WOC) principle is utilized to further assess the capacity of the five modified VIs for estimating rice chlorophyll content, it is proved that OSAVI and MSR display the better performance.

Hyperspectral Discrimination and Response Characteristics of Stressed Rice Leaves Caused by Rice Leaf Folder

Detecting plant health condition plays an important role in controlling disease and insect pest stresses in agricultural crops. In this study, we applied support vector classification machine (SVC) and principal components analysis (PCA) techniques for discriminating and classifying the normal and stressed paddy rice (Oryza sativa L.) leaves caused by rice leaf folder (Cnaphalocrocis medinalis Guen). The hyperspectral reflectance of paddy rice leaves was measured through the full wavelength range from 350 to 2500nm under the laboratory condition. The hyperspectral response characteristic analysis of rice leaves indicated that the stressed leaves presented a higher reflectance in the visible (430~470 nm, 490~610 nm and 610~680 nm) and one shortwave infrared (2080~2350 nm) region, and a lower reflectance in the near infrared (780~890 nm) and the other shortwave infrared (1580~1750 nm) region than the normal leaves. PCA was performed to obtain the principal components (PCs) derived from the raw and first derivative reflectance (FDR) spectra. The nonlinear support vector classification machine (referred to as C-SVC) was employed to differentiate the normal and stressed leaves with the front several PCs as the independent variables of C-SVC model. Classification accuracy was evaluated using overall accuracy (OA) and Kappa coefficient. OA of C-SVC with PCA derived from both the raw and FDR spectra for the testing dataset were 100%, and the corresponding Kappa coefficients were 1. Our results would suggest that it’s capable of discriminating the stressed rice leaves from normal ones using hyperspectral remote sensing data under the laboratory condition.

The Estimation of Winter Wheat Yield Based on MODIS Remote Sensing Data

A yield estimation method by remote sensing was used to estimate the yield of winter wheat in Jiangsu province, China. The first step of this study was to extract the planting area of winter wheat from environmental satellite images and land -use map of Jiangsu province, meanwhile, correlation analyses were performed by using 8-day of composite Leaf Area Index (LAI) data from Moderate Resolution Imaging Spectroradiometer (MODIS) and statistical yield of corresponding counties. Secondly, the average LAI was calculated at the optimal growth period, and the statistical yields of wheat for all counties were collected, in which the former was chosen as the independent variable and the latter was the dependent variable, and the regression model was established. Finally, the accuracy and stability of the regression model were validated using the data of another year. The results indicated that the yield estimation model at provincial level was reliable, the Root Mean Square Error (RMSE) and the Mean Absolute Error (MAE) of the model was 12.1% and 9.7%, respectively. In addition, the yield estimation system of winter wheat in Jiangsu province was constructed and published based on ArcMap and ArcGIS Server.