Spectral Components Analysis: A Bridge between Spectral Observations and Agrometeorological Crop Models

Spectral observations have been acknowledged to indicate general plant conditions over large areas but have yet to be exploited in connection with agrometeorological crop models. One reason is that it is not yet appreciated how periodic spectral observations of row-cropped and natural plant canopies, as expressed by vegetation indices (VI), can provide information on important crop model parameters, such as leaf area index (LAI) and absorbed photosynthetically active radiation (APAR). Two experiments were conducted under AgRISTARS sponsorship, one with cotton and one with spring wheat, specifically to determine the relationships for each term in the " spectral components analysis" identity begin{equation*} LAI/VI times APAR/LAI = APAR/VI.end{equation*} LAI and APAR could, indeed, be well estimated from vegetation indices such as normalized difference(ND) and perpendicular vegetation index (PVI)?apparently because of the close relation between the VI and amount of photosynthetically active tissue in the canopy. APAR and VI measurements are similarly affected by solar zenith angle (SZA), and LAI can be divided by cos SZA at the time of the VI and APAR measurements to achieve correspondence. APAR, ND, and PVI plotted against LAI all asymptote to limiting values in the same way yield does as LAI exceeds 5, further linking canopy development to yield capability. In summary, the spectral components analysis results presented add credence to the information conveyed by spectral canopy observations about plant development and yield, and establish a bridge between remote observations and agrometeorological crop modeling through the variables of mutual concern, LAI, biomass, and yield.     

基于小波变换与支持向量机回归的冬小麦叶面积指数估算

叶面积指数(LAI)是作物长势诊断及产量预测的重要参数。通过对冬小麦采样点的高光谱曲线进行连续小波变换(CWT),然后利用小波系数与LAI 建立支持向量机回归(SVR)模型,实现冬小麦不同生育时期的叶面积指数估算。通过对所研究方法与选取的植被指数、偏最小二乘(PLS)回归等5种方法的反演结果进行统计分析。结果表明:利用连续小波变换确定的LAI 的敏感波段为680、739、802、895 nm,对应尺度分别为8、4、9 和8,对应小波系数的LAI 回归确定系数(R2)明显高于冠层反射率的回归确定系数;利用小波系数与LAI 建立的SVR 模型的反演精度最高,模型实测值与预测值的检验精度(R2)为0.86,均方根误差(RMSE)为0.43;而常用植被指数(归一化植被指数,NDVI;比值植被指数,RVI)建立的估测模型对冬小麦多个生育时期LAI 反演精度最低(R2 0.76,RMSE0.56)。因此利用连续小波变换进行数据预处理,能更好地筛选出对叶面积指数敏感的信息,LAI 回归方法比较结果表明,SVR 比PLS 更适合于LAI 的估测,通过将CWT 与SVR 结合(CWT-SVR)能实现不同生育时期冬小麦叶面积指数的遥感估算。     

Estimating leaf area index from satellite imagery using Bayesian networks

In this study, we investigated the use of Bayesian networks for inferring tropical dry forest leaf area index (LAI) from satellite imagery in dry and wet seasons. LAI was chosen as the variable of interest because leaf area is the exchange surface between the photosynthetically active component of the canopy and the atmosphere. Initial network estimates were obtained from ground truth plot data with known forest structure, LAI, and satellite reflectance in the red and near-infrared bands (as observed by the Landsat 7 Enhanced Thematic Mapper Plus sensor). We tested the performance of the Bayesian networks with scoring rules and also with confidence and surprise scores. We evaluated the networks on a per-pixel basis and created both LAI maps of the study area as well predicted the probability maps for the highest LAI states. Results not only demonstrate the predictive power of a Bayesian network but also its explanatory power which is far beyond what is typically available with current pixel classifier approaches such as spectral vegetation indices or other approaches such as neural networks.     

最小二乘支持向量机用于时间序列叶面积指数预测

遥感反演的叶面积指数（LAI）时间序列被广泛应用于气候模拟、作物长势监测等研究。但遥感数据受天气等因素影响,时间序列的LAI 数据存在缺失。支持向量机（SVM）是一种有效的数据分类和回归预测工具,而最小二乘支持向量机（LS-SVM）是对SVM 的有效改进。以西藏那曲县为例,使用2003-2011 年MODIS LAI 产品,分别用LS-SVM 和SVM 两种方法对研究区域2011 年LAI 时间序列进行预测,并用MODIS 原始LAI 以及部分地面实验样点值进行验证。结果表明,基于LS-SVM 的LAI 时间序列预测算法的精度比基于SVM 的算法高,从而证明LS-SVM 方法能够弥补遥感反演时间序列LAI 数据的缺失问题,对提高时间序列的LAI 遥感产品质量具有重要意义。     

Light Interception and Leaf Area Estimates from Measurements of Grass Canopy Reflectance

Grassland is a major component of the Earth's available land. The vast area and remoteness of this ecosystem makes it difficult to assess its condition and monitor productivity by traditional niethods. Remote sensing potentially offers a rapid nondestructive approach for monitoring such ecosystems. A study was carried out in a tallgrass prairie site near Manhattan, Kansas, during the 1983 and 1984 seasons to investigate the feasibility of estimating light interception and green leaf area index (LAI) from measurements of canopy multispectral reflectance. Greenness (G, i) index was found to be strongly correlated with intercepted photosynthetically active radiation ( PAR). Two methods, a direct regression (RGR) and an indirect approach (IND), were used to estimate LAI from Goi index. The LAI values estimated by RGR method were consistently lower than the measured ones; however, good agreement was obtained between the LAI values estimated by IND method and the measured LAI. This suggests that Goi transformation of canopy spectral reflectance is more closely related to the fraction of intercepted PAR by green foliage than the quantity of green LAI.     

基于EnMAP卫星和深度神经网络的LAI遥感反演方法

区域叶面积指数(Leaf Area Index,LAI)定量反演是开展大尺度农作物长势监测和产量估算的重要基础。针对当前区域LAI遥感定量反演存在的反演精度不理想和模型稳定性弱等问题,提出了一种基于少量训练样本进行LAI高精度反演的深度神经网络(Small Simple Learning LAI-Net,SSLLAI-Net)。该网络由2个卷积层、1个池化层和3个全连接层构成,将光谱反射率数据作为网络输入端、输出端得到LAI反演值,且该网络模型可支持小样本数据量的训练。以德国阿尔卑斯山麓高光谱遥感卫星影像Environmental Mapping and Analysis Program(EnMAP)为数据源,以该区域的谷物、玉米、油菜、其他作物为研究对象,数值实验结果表明当各作物类别的训练样本量均为50时,基于SSLLAI-Net的LAI反演精度分别为0. 95、0. 99、0. 98、0. 90;且在添加噪声的情况下,各作物类别的LAI反演精度分别为0. 95、0. 98、0. 96、0. 89。综上,提出的基于深度神经网络的区域LAI遥感定量反演方法 SSLLAI-Net是鲁棒可靠的,且该模型能够支持稳定的小样本建模。     

Total solution approach using IRS-1C and IRS-P3 data

High spectral (10 nm) and radiometric (16 bits) resolutions of IRS-P3:MOS-B coupled with moderate spatial resolution (188 m) of IRS-P3:WiFS provide unique solutions to many problems related to sustainable management of ecosystems. While the high spatial resolutions of IRS-1C PAN and IRS-1C LISS-3 help in identifying the structural attributes of the biosphere, a synthetic product of MOS-B and WiFS offers immense potential to address several crucial issues including improved classification accuracy in heterogeneous land covers, environmental stress, improved vegetation signal-to-noise ratio, etc. In this paper, the operational issues such as multisensor calibration and validation, registration and merging of multisensor data from different platforms, identification of red edge using IRS-P3:MOS-B data, resolving subpixel heterogeneity, scale anomalies and uncertainty in spectral estimates of biophysical variables are discussed. With the integration of parameters sensitive to atmospheric scattering and soil background reflectance into NDVI derived from the synthetic image, the spectral index called soil adjusted and atmospheric resistant vegetation index (SARVI) has been found to be more sensitive to biophysical variables such as leaf area index (LAI) and fraction of absorbed photosynthetically active radiation (FPAR). It has also reduced, up to certain extent, the uncertainty related to the spectral measurements of bio-physical variables. Further study, in this regard, aims at evaluating the changes in entropy with the fusion of high spectral, radiometric, spatial, and temporal data     

Algorithm for global leaf area index retrieval using satellite imagery

Leaf area index (LAI) is one of the most important Earth surface parameters in modeling ecosystems and their interaction with climate. Based on a geometrical optical model (Four-Scale) and LAI algorithms previously derived for Canada-wide applications, this paper presents a new algorithm for the global retrieval of LAI where the bidirectional reflectance distribution function (BRDF) is considered explicitly in the algorithm and hence removing the need of doing BRDF corrections and normalizations to the input images. The core problem of integrating BRDF into the LAI algorithm is that nonlinear BRDF kernels that are used to relate spectral reflectances to LAI are also LAI dependent, and no analytical solution is found to derive directly LAI from reflectance data. This problem is solved through developing a simple iteration procedure. The relationships between LAI and reflectances of various spectral bands (red, near infrared, and shortwave infrared) are simulated with Four-Scale with a multiple scattering scheme. Based on the model simulations, the key coefficients in the BRDF kernels are fitted with Chebyshev polynomials of the second kind. Spectral indices - the simple ratio and the reduced simple ratio - are used to effectively combine the spectral bands for LAI retrieval. Example regional and global LAI maps are produced. Accuracy assessment on a Canada-wide LAI map is made in comparison with a previously validated 1998 LAI map and ground measurements made in seven Landsat scenes.     

The role of canopy structure in the spectral variation oftransmission and absorption of solar radiation in vegetation canopies

This paper presents empirical and theoretical analyses of spectral hemispherical reflectances and transmittances of individual leaves and the entire canopy sampled at two sites representative of equatorial rainforests and temperate coniferous forests. The empirical analysis indicates that some simple algebraic combinations of leaf and canopy spectral transmittances and reflectances eliminate their dependencies on wavelength through the specification of two canopy-specific wavelength-independent variables. These variables and leaf optical properties govern the energy conservation in vegetation canopies at any given wavelength of the solar spectrum. The presented theoretical development indicates these canopy-specific wavelength-independent variables characterize the capacity of the canopy to intercept and transmit solar radiation under two extreme situations, namely, when individual leaves 1) are completely absorptive and 2) totally reflect and/or transmit the incident radiation. The interactions of photons with the canopy at red and near-infrared (IR) spectral bands approximate these extreme situations well. One can treat the vegetation canopy as a dynamical system and the canopy spectral interception and transmission as dynamical variables. The system has two independent states: canopies with totally absorbing and totally scattering leaves. Intermediate states are a superposition of these pure states. Such an interpretation provides powerful means to accurately specify changes in canopy structure both from ground-based measurements and remotely sensed data. This concept underlies the operational algorithm of global leaf area index (LAI), and the fraction of photosynthetically active radiation absorbed by vegetation developed for the moderate resolution imaging spectroradiometer (MODIS) and multiangle imaging spectroradiometer (MISR) instruments of the Earth Observing System (EOS) Terra mission     

Estimation of global leaf area index and absorbed par usingradiative transfer models

A simple method for the estimation of global leaf area index (LAI) and fraction of photosynthetically active radiation absorbed by the vegetation (FAPAR) from atmospherically corrected Normalized Difference Vegetation Index (NDVI) observations is described. Recent improvements to the authors' three dimensional radiative transfer model of a vegetated surface are described. Example simulation results and a validation exercise are discussed. The model was utilized to derive land cover specific NDVI-LAI and NDVI-FAPAR relations. The method therefore requires stratification of global vegetation into cover types that are compatible with the radiative transfer model. Such a classification based on vegetation structure is proposed and a simple method for its derivation is presented. Proof-of-concept results are given to illustrate the feasibility of the proposed method     

Spectral inputs to crop identification and condition assessment

This review discusses, from an agronomic perspective, contemporary remote sensing research on crop identification and condition assessment. The paper begins with a review of the basic relationships of reflectance and biophysical properties of crop canopies. Leaf area index is shown to be a key parameter linking multispectral reflectance and crop physiology. Major advancements in capability, particularly the development of spectral-temporal profile models, for crop identification have been made in the past decade. The same model form has been used for estimating crop development stage, leaf area index, and canopy light interception as inputs to crop growth and yield models. The paper concludes with a discussion of potential advancements in capability with respect to new sensors.     

Spectral-Temporal Classification Using Vegetation Phenology

In this paper we describe a multitemporal classification procedure for crops in Landsat scenes. The method involves the creation of crop signatures which characterize multispectral observations as functions of phenological growth states. The phenological signature models spectral reflectance explicitly as a function of crop maturity rather than a function of date. This means that instead of stacking spectral vectors of one observation on another, as is usually done for multitemporal data, for each possible crop category a correspondence of time to growth state is established which minimizes the smallest difference between the given multispectral multitemporal vector and the category mean vector indexed by growth state. The results of applying it to winter wheat show that the method is capable of discrimination with about the same degree of accuracy as more traditional multitemporal classifiers. It shows some potential to label degree of maturity of the crop without crop condition information in the training set.     

一种改进的叶绿素提取植被指数

对MERIS陆地叶绿素指数(MTCI)进行了改进,提出了改性的MERIS陆地叶绿素指数(M-MTCI).该指数在植被覆盖度较高时(LAI>1),可获得比MTCI和双峰冠层氮素指数(DCNI)更好的叶绿素提取精度,且比MTCI有更好的抗LAI干扰能力.这些结果的验证是建立在模拟以及实测数据的基础上,具有较好的可靠性,所以M-MTCI具有很好地监测植被叶绿素含量的潜力.     

Analysis of interannual changes in northern vegetation activityobserved in AVHRR data from 1981 to 1994

This paper reports on the analysis of Pathfinder AVHRR land (PAL) data set that spans the period July 1981 to September 1994. The time series of normalized difference vegetation index (NDVI) data for land areas north of 45° N assembled by correcting the PAL data with spectral methods confirms the northerly greening trend and extension of the photosynthetically active growing season. Analysis of the channel reflectance data indicates that the interannual changes in red and near-infrared reflectances are similar to seasonal changes in the spring time period when green leaf area increases and photosynthetic activity ramps up. Model calculations and theoretical analysis of the sensitivity of NDVI to background reflectance variations confirm the hypothesis that warming driven reductions in snow cover extent and earlier onset of greening are responsible for the observed changes in spectral reflectances over vegetated land areas     

Theoretical limits to the estimation of the leaf area index on thebasis of visible and near-infrared remote sensing data

The Leaf Area Index (LAI) of a plant canopy is an important environmental parameter required by various applications. It would be highly desirable to be able to estimate this parameter on the basis of satellite remote sensing data in the optical spectral range. However, LAI affects the propagation of light in a plant canopy (and therefore its measurable reflectance factor) exclusively through a boundary condition of the equation of radiation transfer. It is shown that LAI may be retrievable accurately and reliably only when the canopy is optically thin enough to allow a significant illumination of the underlying soil, and when the optical properties of this soil are such that the radiance field emerging from this level is sufficiently different from that which would be exhibited by a deeper canopy. The combinations of radiative conditions (soil and plant properties) necessary for the reliable and accurate retrieval of the LAI on the basis of remote sensing reflectance data acquired above the canopy in the red and near-infrared spectral regions are investigated and documented with the help of simulation studies. These results show the retrievability of LAI from remote sensing data in optimal situations, however     

Visible/Infrared/Microwave Agriculture Classification, Biomass, and Plant Height Algorithms

This paper describes the results of a study to determine if visible, infrared, and microwave data is correlated to crop-canopy characteristics (biomass and crop height) and can improve estimates of crop acreage. The objectives were to 1) determine if different crops can be discriminated using multifrequency microwave data, and 2) determine which visible, infrared, and microwave spectral regions can classify crops and correlate well to plant biomass, crop height, and the perpendicular vegetation index (PVI). The study was conducted at Dalhart, Texas, in 1980. Aircraft multispectral data collected during the study included visible and infrared data and active multifrequency microwave data. Ground-truth data from each field consisted of soil moisture, total plant biomass, and crop height. Results indicated that C- (4.75 GHz), L- (1.6 GHz), and P- (0.4 GHz) band active microwave data combined with visible and infrared data maintained or improved crop-discrimination accuracy compared to models using only visible and infrared data. The active microwave frequencies were more sensitive to plant height differences than total biomass differences; the K- (13.3 GHz) and C-bands were sensitive to height variations in short plants, while the P-band was sensitive to differences between tall and short plants.     

Canopy architecture and remote sensing of the fraction ofphotosynthetically active radiation absorbed by boreal conifer forests

Measurements of the fraction of photosynthetically active radiation (FPAR) absorbed by the forest overstory were made at 20 sites in black spruce (Picea mariana) and jack pine (Pinus banksiana) boreal forests located in Saskatchewan and Manitoba, Canada. Canopies of both species have similar vertical tree crown structure but different branch and shoot architecture. Intensive investigation was made on the effect of these canopy architecture on the penetration of total visible radiation into the canopy at various solar zenith angles &thetas;, quantified using the projection coefficient G_t(&thetas;). Based on experimental evidence, constant values of G_t(&thetas;) and the above- and below-canopy PAR reflectivities are suggested for these two species for the calculation of daily green FPAR. The calculation then requires only a single stand parameter: the effective green leaf area index (LAI) L_eg, which is similar to the effective LAI L_e measured using optical instruments but reduced by a small fraction to remove the contribution of woody material to the total above-ground plant area. Daily green FPAR of the sites was correlated with the Simple Ratio (SR) and the Normalized Difference Vegetation Index (NDVI) obtained from Landsat 5 TM images. The correlation was better in late-spring than in mid-summer, suggesting spring images are more useful for obtaining FPAR of the overstory. Comparisons of the present with previous results suggest that the background (understory and ground cover) signal and the tree crown shadows are important in satellite measurements of FPAR     

Development of the Adjoint Model of a Canopy Radiative Transfer Model for Sensitivity Study and Inversion of Leaf Area Index

Many canopy reflectance models have been developed in the last decades and used for estimating land surface biogeophysical variables, such as leaf area index (LAI), from satellite observations through optimization procedures. In most studies, the derivative information of the canopy reflectance model has not been used effectively, which limits this approach for regional and global applications. The final solutions are often converged to the local minima. To address these issues, the adjoint model of a canopy radiative transfer model is developed in this study through the automatic differentiation technique. The developed adjoint model is used for sensitivity study, and a combination of the adjoint model with the trust region global optimization method is performed to retrieve LAI from the Enhanced Thematic Mapper Plus (ETM+). This study demonstrates that this method can be reliably used for inverting LAI efficiently and is suitable for global applications.