Abstract The continuous application of artificial intelligence (AI) technologies in online education has led to significant progress, especially in the field of Intelligent Tutoring Systems (ITS), online courses and learning management systems (LMS). An important research direction of the field is to provide students with customised learning trajectories via student modelling. Previous studies have shown that customisation of learning trajectories could effectively improve students’ learning experiences and outcomes. However, training an ITS that can customise students’ learning trajectories suffers from cold-start, time-consumption, human labour-intensity, and cost problems. One feasible approach is to simulate real students’ behaviour trajectories through algorithms, to generate data that could be used to train the ITS. Nonetheless, implementing high-accuracy student modelling methods that effectively address these issues remains an ongoing challenge. Traditional simulation methods, in particular, encounter difficulties in ensuring the quality and diversity of the generated data, thereby limiting their capacity to provide intelligent tutoring systems (ITS) with high-fidelity and diverse training data. We thus propose Sim-GAIL, a novel student modelling method based on generative adversarial imitation learning (GAIL). To the best of our knowledge, it is the first method using GAIL to address the challenge of lacking training data, resulting from the issues mentioned above. We analyse and compare the performance of Sim-GAIL with two traditional Reinforcement Learning-based and Imitation Learning-based methods using action distribution evaluation, cumulative reward evaluation, and offline-policy evaluation. The experiments demonstrate that our method outperforms traditional ones on most metrics. Moreover, we apply our method to a domain plagued by the cold-start problem, knowledge tracing (KT), and the results show that our novel method could effectively improve the KT model’s prediction accuracy in a cold-start scenario.

Question answering system is an emerging information service system, which enables people to get answers easily and quickly to their questions. However, most of the existing methods do not effectively utilize semantic information and spatial properties. In this paper, a semantics-guided and spatial-aware framework for natural resources geo-analytical question answering is proposed. First, we use linguistic analysis techniques to translate natural language questions into structured texts. Then, the constructed natural resources semantic knowledge bases and query sample bases are introduced to extract the spatial and semantic information of geographic entities (i.e., precise geographic location and spatial representation). Considering most questions need to be answered based on a combination of multiple data sources, an improved user model that introduces semantic similarity is proposed to select appropriate data sources for a specific question. Finally, geo-analytical workflows are automatically generated by utilizing graph models and rule templates. Additionally, a question answering system was developed based on this framework and applied to natural resources monitoring in Hubei. The proposed framework is validated in both experiments and the case study. The results show that the proposed framework performs favorably on natural resources geo-analytical question answering tasks.

Despite the long-term records of satellite observations, factors controlling the seasonal and interannual variations in canopy reflectance remain poorly understood. Leaf optical properties (LOP, including leaf reflectance and transmittance) changes as leaves age, and thus impact the seasonal pattern of canopy reflectance, i.e., the “leaf age effect”. Here, we combined the Geometric Optical Radiative Transfer (GORT) model with continuous field measurements of leaf- to stand-scale characteristics to simulate canopy reflectance in a Chinese fir plantation with stand development (1-33 yr). We found that canopy structure controls the variations in canopy reflectance during young stages (<; 10 yr) and that leaf age controls the variations in canopy reflectance after canopy closure. Moreover, we found that the “leaf age effect” get enhanced with stand development, with R2 increased from about 0.1 to 0.56, 0.67, 0.92, and 0.82 for young, half-mature, near-mature, and mature stages, respectively. This study reveals the stand age dependence of leaf age effect on canopy reflectance which improves our interpretation and understanding of satellite observations to study the ecosystem function of forests.

Land surface albedo analysis and prediction are of great significance for global energy budget research and global change forecasting. Research has been performed on time series albedo analysis but seldom attempt was performed on land surface albedo prediction. This article develops an effective method for land surface albedo prediction from Moderate-Resolution Imaging Spectroradiometer (MODIS) time series albedo data (MCD43A3). It consists of time series data decomposing and time series data forecasting. The ensemble empirical mode decomposition (EEMD) method decomposes the MODIS historical time series albedo data into several intrinsic mode functions (IMFs) and one residual series, then the nonlinear autoregressive neural network (NARnet) method is used to forecast each IMF component and residue. The predictions of all IMFs and residue are summed to obtain a final forecast for the albedo series. The proposed method was performed on monthly and daily albedo prediction both in snow-free and snowy areas. The results showed that the forecast albedo consists of the MODIS albedo data well, with R2greater than 0.89 and RMSE less than 0.052 for snow-free areas. For snowy areas, the forecasting also performed well during snow cover periods, with R2greater than 0.76 and RMSE less than 0.076. For irregular change periods of snow falling and melting, it is hard to get very high prediction accuracy due to the irregular land surface change. For this problem, more land surface information should be introduced, or adjusting the model over time is necessary.

Leaf area index (LAI) is an important parameter for describing vegetation change and growth trend. Due to various problems of satellite observation and retrieval algorithm, spatiotemporal complete LAI data is limited, which hindered the application of LAI in different areas. In this paper, a data fusion method, multiresolution tree (MRT) model, was used to develop LAI of different spatial resolution. Three LAI data sets of Landsat (30 m), MODIS (450 m) and GLASS (900 m) are used. MRT was performed in Ukraine area including farmland, forestland and grassland. Results indicate that MRT is an efficient algorithm to get spatial complete LAI estimation at different resolution. LAI of different spatial resolution consists well, R2and RMSE are 0.8612 and 0.3986 when compare Landsat LAI with MODIS LAI, which are 0.7568 and 0.4736 when compare Landsat with Glass.

Land surface albedo (LSA) is an important parameter charactering the land surface energy balance. The prevailing LSA products have supplied a well understanding of the global weather change but for regional use, it is difficult to capture the patch-size change induced by human activities. In this paper, we estimate the high resolution LSA from GF-1 WFV data based on a direct estimation algorithm. Results compared with field observation indicate that the estimation accuracy is high with the coefficient of determination of 0.705 and Bias of 0.008. When compared with Landsat LSA data, a high consistency is performed, the coefficients of determination for black and white sky albedo are 0.943 and 0.941 respectively.

Medium spatial resolution biomass is a crucial link from the plot to regional and global scales. Although remote-sensing data-based methods have become a primary approach in estimating forest aboveground biomass (AGB), many difficulties remain in data resources and prediction approaches [1, 2]. Each kind of sensor type and prediction method has its own merits and limitations. To select the proper estimation algorithm and remote-sensing data source, several forest AGB models were developed using different remote-sensing data sources (Geoscience Laser Altimeter System (GLAS) data and Thematic Mapper (TM) data) and 108 field measurements. Three modeling methods (stepwise regression (SR), support vector regression (SVR) and random forest (RF))were used to estimate forest AGB over the Daxing'anling Mountains in northeastern China. The results of models using different datasets and three approaches were compared. The random forest AGB model using Landsat5/TM as input data was shown to estimate AGB reliably by regression (R2=0.96 RMSE=17.73Mg/ha) and cross validation (R2=0.71 RMSE=39.60 Mg/ha).

Leaf area index (LAI) is a critical biophysical parameter to quantify leaves in forests and to study the ecological role of forests in the terrestrial ecosystem. Currently, the most widely-used model to achieve forest canopy structure parameters is the hybird Geometric optical and radiation transfer (GORT) model, by considering the advantages of radiation transfer model and geometric optical model. However, there are some problems when applied in forest areas. The original GORT model divided the forest canopy into several layers, but it ignored the lower shrubs on the forest underlying surface. And the underlstory canopy of forest has effect for the whole canopy reflectance. In this study, we modified hybrid canopy GORT model by considering leaves from both trees and underlying shrubs. Using the modified GORT model, we can improved the accuracy of forest canopy BRF. Then we can get more high-precision forest LAI.

In the validation of MODIS LAI product, much attention has been paid to directly compare the product data and the reference values, few efforts have been made to analyze the bias of validation result comprehensively. In this paper, based on the evaluation of MODIS LAI product, we further divide the bias into three aspects: algorithm, reflectance data and clumping effect. Moreover, the individual influence on the total bias is quantified. The results of assessment indicate MODIS LAI product has an obvious underestimate, as much as 34.14% in this area compared with reference LAI. Individual influence on the total deviation of each bias is 57.50%, 28.33% and 14.17%, respectively.

Leaf area index (LAI) is an important vegetation biophysical variable and has been widely used for crop growth monitoring and yield estimation, land-surface process simulation, and global change studies. Several LAI products currently exist, but most have limited temporal coverage. A long-term high-quality global LAI product is required for greatly expanded application of LAI data. In this paper, a method previously proposed was improved to generate a long time series of Global LAnd Surface Satellite (GLASS) LAI product from Advanced Very High Resolution Radiometer (AVHRR) and Moderate Resolution Imaging Spectroradiometer (MOD!S) reflectance data. The GLASS LAI product has a temporal resolution of eight days and spans from 1981 to 2014. During 1981-1999, the LAI product was generated from AVHRR reflectance data and was provided in a geographic latitude/longitude projection at a spatial resolution of 0.05°. During 2000-2014, the LAI product was derived from MODIS surface-reflectance data and was provided in a sinusoidal projection at a spatial resolution of 1 km. The GLASS LAI values derived from MODIS and AVHRR reflectance data form a consistent data set at a spatial resolution of 0.05°. Comparison of the GLASS LAI product with the MODIS LAI product (MOD15) and the first version of the Geoland2 (GEOV1) LAI product indicates that the global consistency of these LAI products is generally good. However, relatively large discrepancies among these LAI products were observed in tropical forest regions, where the GEOV1 LAI values were clearly lower than the GLASS and MOD15 LAI values, particularly in January. A quantitative comparison of temporal profiles shows that the temporal smoothness of the GLASS LAI product is superior to that of the GEOV1 and MODIS LAI products. Direct validation with the mean values of high-resolution LAI maps demonstrates that the GLASS LAI values were closer to the mean values of the high-resolution LAI maps (RMSE = 0.7848 and R2= 0.8095) than the GEOV1 LAI values (RMSE = 0.9084 and R2= 0.7939) and the MOD15 LAI values (RMSE = 1.1173 and R2= 0.6705).

Leaf area index (LAI) is an important biophysical variable indicating forest growth. A major challenge is to improve the LAI estimates for large forest-covered areas. One way to obtain LAI value is using current LAI products. Current LAI products contain many uncertainties and need improvement. This paper aims to improve forest LAI estimates by combining satellite reflectance derived LAI with forest growth model (physiological principals predicting growth, 3-PG) estimates of LAI. 3-PG can give an accurate estimation of forest inter-annual growing trend, while remote sensing data can provide long time series observation of seasonal variations of forest phenology. We applied this method to Chinese fir forest in China, where the detailed data are available. The combined results were more accurate than either the satellite or the 3-PG estimates. We conclude that we can improve the space-time forest canopy LAI estimates by combining forest growth model with satellite imagery.

Currently available land-surface parameter products are generated using parameter-specific algorithms from various satellite data and contain several inconsistencies. This paper developed a new data assimilation framework for consistent estimation of multiple land-surface parameters from time-series MODerate Resolution Imaging Spectroradiometer (MODIS) surface reflectance data. If the reflectance data showed snow-free areas, an ensemble Kalman filter (EnKF) technique was used to estimate leaf area index (LAI) for a two-layer canopy reflectance model (ACRM) by combining predictions from a phenology model and the MODIS surface reflectance data. The estimated LAI values were then input into the ACRM to calculate the surface albedo and the fraction of absorbed photosynthetically active radiation (FAPAR). For snow-covered areas, the surface albedo was calculated as the underlying vegetation canopy albedo plus the weighted distance between the underlying vegetation canopy albedo and the albedo over deep snow. The LAI/FAPAR and surface albedo values estimated using this framework were compared with MODIS collection 5 eight-day 1-km LAI/FAPAR products (MOD15A2) and 500-m surface albedo product (MCD43A3), and GEOV1 LAI/FAPAR products at 1/112° spatial resolution and a ten-day frequency, respectively, and validated by ground measurement data from several sites with different vegetation types. The results demonstrate that this new data assimilation framework can estimate temporally complete land-surface parameter profiles from MODIS time-series reflectance data even if some of the reflectance data are contaminated by residual cloud or are missing and that the retrieved LAI, FAPAR, and surface albedo values are physically consistent. The root mean square errors of the retrieved LAI, FAPAR, and surface albedo against ground measurements are 0.5791, 0.0453, and 0.0190, respectively.

Forest canopy leaf area index (LAI) is an important biological parameter and an evaluation index of the vegetation canopy. It is difficult to get LAI through direct inversion using Li-Strahler geometric-optical model (GOMS)[1,2]. In this paper, LAI is introduced into the GOMS model as an independent variable to substitute other canopy structure parameters. The verification results well demonstrated the accuracy of the modified GOMS model.

The inconsistent data quality of remote sensing observation, which is mainly caused by atmospheric conditions, presents problems in the application of these data. For the land cover types that cycle yearly, the variations in surface reflectance usually have temporal periodic characteristics. In this study, we modeled the temporal feature of Moderate-Resolution Imaging Spectroradiometer (MODIS) Nadir BRDF-adjusted reflectance (NBAR) time series data of the typical vegetated area using the season-trend statistical method. The fitting values of season-trend model were applied to the recursive estimation of leaf area index (LAI) time series based on nonlinear autoregressive exogenous (NARX) neural network. The results of MODIS NBAR modeling indicate that the season-trend method is effective to model the NBAR time series of the vegetation surface. The NARX neural network works well using the improved NBAR time series as input, and the estimated LAI time series is more continuous than the MODIS LAI.

Land surface albedo is a key input parameter required in the current general circulation models (GCMs). The accurate validation of albedo products is very important for use in various applications by the scientific community. The most common validation methods compare motely sensed albedo products directly with ground-based observed broadband albedo values without considering the band width inconsistency and the narrow-to-broad band transformation error of these products. Unlike previous validation approaches, we propose a spectral albedo direct validation method. By applying the novel instrument of multi-band albedometer, it is possible to obtain MODIS visible and near-infrared bands albedo measurements. The in-field measured data combined with a 16-day interval is compared directly with MODIS narrow band albedo products. The results indicate that the multi-band albedometer measurement and the MODIS narrow band white and black sky albedo has a good consistency. It is an efficient way to get rid of the errors induced by the band conversion and is a labor-saving technique of making possible time series measurements.

Leaf area index (LAI) products at regional and global scales are being routinely generated from individual instrument data acquired at a specific time. As a result of cloud contamination and other factors, these LAI products are spatially and temporally discontinuous and are also inaccurate for some vegetation types in many areas. A better strategy is to use multi-temporal data. In this paper, a method was developed to estimate LAI from time-series remote sensing data using general regression neural networks (GRNNs). A database was generated from Moderate-Resolution Imaging Spectroradiometer (MODIS) and CYCLOPES LAI products as well as MODIS reflectance products of the BELMANIP sites during the period from 2001-2003. The effective CYCLOPES LAI was first converted to true LAI, which was then combined with the MODIS LAI according to their uncertainties determined from the ground-measured true LAI. The MODIS reflectance was reprocessed to remove remaining effects. GRNNs were then trained over the fused LAI and reprocessed MODIS reflectance for each biome type to retrieve LAI from time-series remote sensing data. The reprocessed MODIS reflectance data from an entire year were inputted into the GRNNs to estimate the 1-year LAI profiles. Extensive validations for all biome types were carried out, and it was demonstrated that the method is able to estimate temporally continuous LAI profiles with much improved accuracy compared with that of the current MODIS and CYCLOPES LAI products. This new method is being used to produce the Global Land Surface Satellite LAI products in China.

Time series leaf area index (LAI) derived from remote sensing data is a key parameter for environment researches especially on dynamic changes of land surface. In this study, a new approach was developed to retrieve LAI from time series Moderate Resolution Imaging Spectroradiometer (MODIS) multi-angular remote sensing data. Based on radiative transfer theory, we used Ross Thick-Li Sparse Reciprocal (RTLSR) kernel driven model to generate specific directional BRFs and corresponding anisotropy information, employed Scattering by Arbitrarily Inclined Leaves with Hotspot (SAILH) model to fill in missing data and Data-Based Mechanistic modeling (DBM) procedure to model and estimate time series vegetation LAI. The preliminary results indicated that the LAIs derived in this study have good agreement with ground LAI measurements and their continuity of the time series are superior to MODIS LAI product.

Forest canopy LAI is an important variable to the modeling of energy over regional and global scales. This paper explored the method to retrieve LAI and FAVD directly from full-waveform LiDAR data. At first, he full-waveform LiDAR data, including emitting and return laser waves, were denoised and decomposed. Based on the interaction model between LiDAR and forest, and the parameters of Gaussian decomposition, the forest FAVD was then inverted. Besides, the missing LiDAR data caused by high density forest and LiDAR system restrictions was filled based on the existing LiDARdata and CHM. At last, LAI of the research area was retrieved from the inversed FAVD, and validated by the LAI field measurement (r = 0.73, RMSE = 0.67).

Land surface albedo is a critical physical variable controlling the distribution of radiative energy between the atmosphere and surface. In this paper, we proposed a method to estimate high temporal and spatial resolution land surface shortwave albedo. The method consisted of three main steps: 1. to generate fine-resolution shortwave albedo with Landsat-7 ETM+ data; 2. to generate coarse-resolution shortwave albedo based on MODIS albedo products; 3. to estimate high temporal and spatial resolution albedo using revised STARFM model. For the problem that the estimation accuracy degrades somewhat in heterogeneous regions, we further improved STARFM model by introducing the idea of mixed pixel decomposition. Comparing with ETM+ albedo, the estimated albedo were of reasonable accuracy with root mean square error (RMSE) less than 0.03, and maintained a high level of spatial detail. Results obtained by using improved model showed higher correlation and smaller deviation compared with ETM+ albedo.

In recent years, time series remote sensing data products have been assimilated into the coupled crop growth model and the radiative transfer model to improve the time series LAI estimation. However, due to the large number of input parameters to the crop growth model, the applications of the crop growth models for regional use is restricted. This paper proposed a data-based mechanistic assimilation method for estimation of the time series LAI from Moderate Resolution Imaging Spectroradiometer (MODIS) data. By coupling a revised universal data-based mechanistic model (LAI_UDBM) with a vegetation canopy radiative transfer model (PROSAIL), The proposed method applies the Ensemble Kalman Filter (ENKF) method to improve the estimation accuracy. Results indicate that the time series LAI estimated by this approach is superior to the MODIS LAI. Furthermore, because the model does not require the historical observation of every pixel, it is applicable over a wider range of uses.

The ground measurements of Leaf Area Index (LAI) are usually used to validate the LAI estimated using remote sensing observations. The main problem encountered in the validation is the scale mismatch between the sampling area of ground measurements and coarse-resolution image pixel, especially when those sampling area are not ideal homogenous. This study introduced a new approach of upscaling ground measurements to the coarse-resolution scale for the validation of LAI estimations. This upscaling method was based on the Taylor Series expansion Model (TSM). The high-resolution images were used to provide auxiliary information at the sub-pixel scale of coarse-resolution image pixel. The possible error associated with this method is derived from the neglection of the third- and higher-order TSM terms and the uncertainty of the empirical model. The upscaled ground measurements with upscaling error smaller than half of eigenaccuracy could be used for validation of LAI estimations with coarse resolution.

The Leaf Area Index (LAI) values derived from remote sensing images with different resolutions can be different, because of the heterogeneity of land surface. To validate the coarse resolution LAI product, the in situ LAI measurements and the high resolution LAI map need to be upscaled to the coarse resolution(such as 1 km). As the principle of fractal geometry has been applied to describe the irregular distribution of vegetation, the measured LAI can be upscaled using Fractal Dimension (FD) which is the key parameter of fractal geometry. In this paper two FD computing algorithms, Triangular Prism Surface Area method (TPSA) and Differential Box Counting algorithm (DBC), were compared in two aspects. One aspect was the ability of describing heterogeneity of LAI in different landscapes; another was the spatial distribution of FD and upscaled LAI. The log-log curves of computing algorithms showed that TPSA performs better to capture the difference between different landscapes than DBC algorithm does. Also TPSA can detect the scale inflexion of LAI spatial distribution but the DBC can't. The pattern of FD and upscaled LAI map using TPSA is similar with the real pattern of vegetation, while the FD and upscaled LAI map using DBC algorithm cannot capture the real pattern of vegetation.

Forest canopy height is an important variable to the modeling of energy over regional and global scales. This paper first examined the relationship between field-surveyed canopy height and LiDAR-derived canopy height, regression between them had an RMSE and R2value of 0.94 m and 0.64. To extrapolate the LiDAR height to a continuous area, we compared the ability of four sources of optical remote sensing data (MODIS BRFs, MODIS NBAR, MISR and SPOT data) in predicting the LiDAR measured canopy height. Multivariate linear regression and single variable nonlinear regression models were developed, and the best model accurately predicted the LiDAR height using MODIS BRFs data (RMSE=1.2 m, R2= 0.67). This model was applied to the whole study area and finally the canopy height map of the study area was generated.

Surface albedo is an important parameter in modeling climate processes as it determines the energy budget of the earth's surface. Operational surface albedo products are available from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor. However, due to the factors associated with weather, sensors and algorithms, albedo products from satellite observations often have many gaps. In this paper, we reformed the semi-empirical kernel-driven model to express the relation between the bidirectional reflectance distribution function (BRDF) and the Leaf Area Index (LAI). Within a small region, the soil properties under plant canopies are similar, therefore the reflectance is unanimous. When the land cover types are identical and the LAI values are same, the canopy top reflectance should be same. Thus, within a small region, the pixels of same properties have the same reflectance. For one pixel, if the MODIS albedo product has no retrieval but MODIS LAI product has high quality retrieval, the LAI information can be used to filled the MODIS albedo data gaps. A comparison indicates that the filled data conform well with the in-situ measurements albedo.

Spatial resolution is one of the central issues in land cover mapping from remote sensing imagery, and sub-pixel land cover mapping is difficult and challenging in this domain. Soft classification can provide more information than hard classification. However, the spatial location of land cover compositions within each pixel is unknown. To solve this problem, super-resolution mapping methods have been developed in recent years. In this paper, a landscape structure based approach for super-resolution land cover mapping is introduced to generate super-resolution land cover maps from remote sensing data. The method was used to map simulated target images and a real landscape and the results indicate that this approach has the ability to reconstruct complicated landscape with linear features, and large or small patches relative to the pixel size.

Aphid (Hemiptera: Aphididae) appears in wheat planting area of China almost every year and have had significant economic impacts on wheat yield. As a result, large amounts of insecticides are used to control aphid populations, which may cause environmental pollution. Therefore, remote sensing as a repeatable and rapid method is necessary for monitoring aphid damage level. The study analyzed the hyperspectral characteristics of wheat infested by several aphid damage levels and selected out the sensitive bands to aphid damage levels, and the aphid damage hyperspectral index (ADHI) was developed based on the most sensitive bands to aphid damage levels in the visible, near-infrared and short-wave infrared regions. The results indicated that ADHI exhibited a high correlation with aphid damage levels (R2=0.839), so it had the potential to detect wheat damage caused by aphid.

The new data assimilation algorithm is developed to estimate LAI from time-series MODIS reflectance data (MOD09A1). The canopy radiative transfer model (ACRM) is coupled with an empirical LAI dynamic model, and the ensemble Kalman smoother (EnKS) is used to estimate the parameters of the coupled model from MOD09A1 data. The preliminary analysis using MODIS surface reflectance data at some AmeriFlux network sites was performed to validate this method. The results show that the algorithm is helpful to produce the temporally continuous LAI estimation of cropland efficiently. By comparing with the field measured LAI, the retrieved LAI has been significantly improved and shown more smooth in time series than the MODIS LAI product.

Land surface temperature (LST) and land surface emissivity (LSE) are two key parameters in numerous environmental studies. In this paper, a stepwise refining temperature and emissivity separation (SRTES) algorithm is proposed based on the analysis of the relationship between surface self-emission and atmospheric downward spectral radiance in a narrow spectral region. The SRTES algorithm utilizes the residue of atmospheric downward spectral radiance in the calculated surface self-emission as a criterion and adopts a stepwise refining method to determine both the emissivity at the location of an atmospheric emission line in a narrow spectral region and the surface temperature. Three methods have been used to evaluate the SRTES algorithm. First, numerical experiments are conducted to evaluate if the SRTES algorithm can accurately retrieve the “true” LST and LSE from the simulated data. When a noise equivalent spectral error of$2.5\ e^{-9}\ \hbox{W/cm}^{2}/\hbox{sr}/\hbox{cm}^{-1}$is added into the simulated data, the retrieved temperature bias$(T_{\rm bias})$is 0.04$\pm$0.04 K, and the root-mean-square error (rmse) of the retrieved emissivity is below 0.002 except in the extremities of the 714–1250$\hbox{cm}^{-1}$spectral region. Second,in situmeasurements are used to validate the SRTES algorithm. The average rmse of the retrieved emissivity of ten samples is about 0.01 in the 750–1050$\hbox{cm}^{-1}$spectral region and is 0.02 in the 1051–1250$\hbox{cm}^{-1}$spectral region, but the rmse is larger when the sample emissivity is relatively low. Third, our new algorithm is compared with the iterative spectrally smooth temperature and emissivity separation (ISSTES) algorithm using both a simulated data set andin situmeasurements. The comparison demonstrates that the SRTES algorithm performs better than the ISSTES algorithms, and it can overcome some of the common drawbacks in the existing hyperspectral TES algorithms for the accurate retrieval of both temperature and emissivity.

The Effective Plant Area Index (PAIe) of forest canopy is an important parameter in the canopy reflectance modeling and validation. PAIe can be transformed to leaf area index (LAI) with clumping index. But it is still very difficult to obtain its ground truth value by in situ measurement. In this study, we measured PAIe of 3 typical wood land sites in China by means of three indirect optical techniques: plant canopy analyzer LAI-2000, TRAC, and Digital Fisheye Camera. The 5 measured stands include Qinghai spruce, peach, poplar, willow and silver chain. In this paper, the forest canopy PAIe measured by those three instruments is compared firstly. Then our new approach is how to use the three measured data to get the better PAIe estimation with less overall error. The method of minimizing overall error is adopted to produce PAIe of the sample plots in our study sites, which has less overall error comparing with the PAIe measured by each individual instrument. This approach is also validated by using computer simulated wide-angle viewing pictures when true LAI/PAIe values are given.

LAI is the more important parameter of vegetation canopy, so LAI inversion from remote sensing observations is the hot study field, especially for the high spatial and high temporal resolution remote sensing data. Beijing-1 microsatellite is an applied earth observing microsatellite of China, which can also give us the good data of short cycle time and wider coverage. So it is necessary to generate the quantitative product of BJ-1 remote sensing data. In this paper, the main object is to study on the method of the leaf area index inversion for producing BJ-1 LAI product. The neuronal network method is used to get the relationship between LAI and reflectance in green, red and NIR band. Based on the BJ-1 LAI inversion, the second object of this paper is to generate of high spatial and high temporal resolution LAI product. A method is proposed to get high spatial and temporal resolution LAI product by fusing the time-series MODIS LAI product(1 km, 8-day product)and BJ-1 LAI. Through this study, we can get the LAI products of BJ-1, which is with the high spatial resolution and high time resolution. This product will provide more information of vegetation for BJ-1 microsatellite data applications.

Satellite remote sensing enables derivation of LAI globally at available spatial resolution and temporal frequency, and several LAI products have been produced. However, there are problems for the current global or regional LAI products, which restrict the application of these products. On the one hand, there are gaps between the large number of parameters of physical models and the small amount of data obtained by single sensor, which may cause the decrease in accuracy that LAI products should have. On the other hand, there are gaps between instantaneous observation of remote sensing and parameters which have change rules. Custom methods to retrieve LAI from remote sensing data are just involving the transient observations, discarding the information about process. To resolve these problems, we develop a methodology to retrieve LAI by involving diverse data from multiple sensor and LAI change rule. The methodology can take full advantage of the different band and angle information of time series MODIS and MISR data to improve the accuracy of the retrieved LAI over the MODIS LAI product compared to the field measured LAI data. And the retrieved LAI is also temporally continuous.

It is an urgent need for natural disaster monitoring to generate biophysical variables data with high accuracy timely from remotely sensed data. A real-time inversion method to estimate leaf area index (LAI) using MODIS time series reflectance data (MOD09A1) is developed in this paper. A seasonal autoregressive integrated moving average (SARIMA) model is used to derive LAI climatology. A dynamic model is then constructed based on the climatology from the SARIMA model to evolve LAI in time, and used to provide the short-range forecast of LAI. Predictions from the model are used with the ensemble Kalman filter (EnKF) techniques to recursively update biophysical variables as new observations arrive. The validation results show that the real-time inversion method is able to produce a relatively smooth LAI product efficiently, and the accuracy is significantly improved over the MODIS LAI product.

Multiple leaf area index (LAI) products have been generated from remote-sensing data. Among them, the Moderate-Resolution Imaging Spectroradiometer (MODIS) LAI product (MOD15A2) is now routinely derived from data acquired by MODIS sensors onboard Terra and Aqua satellite platforms. However, the MODIS LAI product is not spatially and temporally continuous and is inaccurate in many areas for some vegetation types. In this paper, a new algorithm is developed to estimate LAI from time-series MODIS reflectance data (MOD09A1). A radiative-transfer model is coupled with a double-logistic LAI temporal-profile model, and the shuffled complex evolution optimization method, developed at the University of Arizona, is used to estimate the parameters of the coupled model from the temporal signature in a given time window. Preliminary analysis using MODIS surface-reflectance data at flux sites was performed to validate this method. The results show that the new algorithm is able to construct a temporally continuous LAI product efficiently, and the accuracy has been significantly improved over the MODIS LAI product as compared to field-measured LAI data.

In this paper, an artificial neural network approach to estimate LAI from the combination of CYCLOPES and MODIS products over the 2001 to 2003 period is described in detail. Reflectances in RED, NIR and SWIR band and LAI with good quality were chosen according to the Quality Control information and the temporal consistency between the two LAI products. Four different reflectance and LAI combinations from both sensors were used as the input and output variables of the ANNs with different land cover types for training. The prediction abilities of the trained ANNs were validated using the datasets which were not used in the training process. It is observed that the ANNs can be well trained and have promising prediction abilities. The time series LAI derived from the trained ANNs is charactered by better temporal consistency compared with the original MODIS LAI product.

The bidirectional reflectance distribution function (BRDF) of land surfaces specifies the behavior of surface directional reflectance as a function of illumination and viewing angles. The Moderate resolution Imaging Spectroradiometer (MODIS) and The Multiangle Imaging Spectroradiometer (MISR) provide their BRDF model parameters products respectively. Since the BRDF model parameters products are inverted from the observations with limited viewing directions for a given pixel, it is necessary to evaluate whether they can effectively characterize the directional reflectance in other viewing directions. In this study, we choose BRDF products data on four land cover types to learn their bidirectional reflectance signatures and to analyze the representation of MISR and MODIS BRDF model parameters products. The results show that: the MISR BRDF product shows hot spot more clearly than MODIS product, and both BRDF models' representative ability for extrapolating the reflectance at those directions without viewing dada tends to weaken when the viewing zenith angle increases.

Surface albedo is one of the key radiation parameters required for modeling of the Earth's energy budget, and many ecological and climate models usually require high quality consistent surface albedo as inputs and validation sources. The Moderate Resolution Imaging Spectroradiometer (MODIS) has been providing surface albedo products periodically. However, due to influence of weather, sensors and algorithms, MODIS albedo products often have many gaps and low quality pixels. This paper proposed an improved spatio-temporal smoothing algorithm utilizing multilevel procedure that combines spatial interpolation and tempora smoothing to get better albedo products based on multiyear observations and high quality neighboring pixels. Compared with field measurements, the improved albedo products in China based on MOD43B3 show better correlation with the measured surface albedo with the overall root mean squared errors around 0.0767. The generated albedo products may be applied in land surface models.

An anisotropic flat index (AFX) is among the operational BRDF and albedo products offered in both the V004 and V005 reprocessed versions. We examine this BRDF shape indicator with 20 ground multiangular data sets as well as MODIS satellite samples, and find that the AFX routinely captures the BRDF shape. An AFX1.0 corresponds to a bowl shaped anisotropy pattern. For green vegetation, the BRDF shape is related to canopy architecture. Therefore, the behavior of the AFX provides an opportunity to infer canopy structure of the surface cover that produces the anisotropic effect.

The accuracy of Land Use/Land Cover data derived from remote sensing images is critical for many applications. Classification error is caused by the interaction of numerous factors, including landscape characteristics, sensor resolution, preprocessing algorithms, and classification procedures. The purpose of this paper is to extend the evaluation of the Collection 4 MODIS land cover product (MOD12Q1) to Shandong Province, east China, and to analyze the distribution of errors from a landscape pattern perspective. Logistic regression was employed to assess the impact of landscape characteristics on classification accuracy. A correct classification probability map based on logistic model was given for describing MOD12 land cover product's error distribution. The results indicate that classification accuracy increases as land cover patch size increases and as heterogeneity decreases.

This study analyzes the accuracy of the Particle Filter (PF) assimilation algorithm to retrieve Leaf Area Index (LAI) from remotely sensed observations using the crop growth model CERES_Maize as a dynamic system, the radiative transfer model SAIL as the observation equation, and MOD09 for external observations. Nonlinearity of the crop growth and radiative models makes the posterior probability of retrieved LAI non-Gaussian. The advantage of PF is its ability to estimate accurately the non-Gaussian posterior probability of retrieved LAI by the particles system. We retrieve LAI by the bootstrap particle filter algorithm whenever a remotely sensed observation was available. By comparing our filtered results to measured LAI at the Yushu area of Jilin province, China, we found that this algorithm greatly improved LAI retrieval. The crop growth model's constraint information and accurate estimation of posterior probability contributed to the improvement in retrieved LAI. We validated the accuracy of maize yield estimation by field measurements.

A new algorithm is developed to estimate LAI from time-series MODIS reflectance data (MOD09A1) based on coupled radiative transfer model and process model. The radiative transfer model is coupled with an empirical LAI dynamic model to simulate the time series reflectances. An optimization method is used to adjust the values of the parameters of the coupled model to seek a model trajectory that best fits a set of observations in a given time window. The preliminary analysis using MODIS surface reflectance data at some fluxnet sites was performed to validate this method. The results show that the algorithm is able to produce spatially and temporally continuous LAI product efficiently, and the accuracy of the retrieved LAI has been significantly improved over the MODIS LAI product compared to the field measured LAI data.

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.

Leaf area index (LAI) is an important parameter for estimating growth status of crops. An important method for retrieving LAI from remotely sensed observations is by the canopy reflectance model inversion. But many model inversion methods didn't take into account the relationship between estimated LAIs in different crop growth stages. In this research, we consider the crop growth model which describe how LAI change with crop growth stages. The main method for this research is assimilating multiple crop's canopy reflectance observed in different growth stages into the cost function for LAI retrieval. Data assimilation algorithm we used in this study is the variation algorithm. The model inversion results have showed that time series observations of canopy reflectance can decrease uncertainty of estimated LAI in model inversion.

Leaf area index (LAI) is an important parameter for describing vegetation canopy structure in the terrestrial ecosystem on the global, continental and regional scales. In this paper, a simple data assimilation method for improving estimation of MODIS LAI time-series data products based on a new 2-D LMS (two-dimensional least mean square) adaptive filter was proposed. Firstly, The new 2-D LMS adaptive filter algorithm is introduced and analyzed. Secondly, A simple data assimilation method for improving estimation of MODIS LAI time-series data products based on the new 2-D LMS adaptive filter and quality control data of MODIS LAI is proposed. Finally, the experiments are performed based on the simple data assimilation method using MODIS LAI data products from 2000 to 2005 of southwestern China.

In this paper, a methodology for improving the Leaf Area Index (LAI) product of the vegetation canopy and the preliminary retrieval results by integrating Moderate Resolution Imaging Spectroradiometer (MODIS) and Multi-angle Imaging SpectroRadiometer (MISR) data is presented. We attempt to improve the estimation of LAI through a physical inversion algorithm with a canopy reflectance model. Taking Konza Prairie experiment as an example, the results suggest that this method can utilize effectively the MISR and MODIS observing information and the prior knowledge which can be obtained from the ground measuring and the sensor products.

The delineation of management zones is an economical and effective measure for the variable-rate application in precision agriculture. The methods of empirical and unsupervised classification have been used by many researchers in the delineation of management zones, but these methods are only built upon the information of attributes in every spatial cell, and the spatial relationships and their spatial interaction between cells are not considered. AS a result, there are many isolated cells or patches in the zoned map, this is not advantageous for the operation of the variable-rate application. Based on the traditional k-means cluster (K-M) and the spatial autocorrelation, a new method, spatial contiguous k-means clustering algorithm (SC-KM), was developed in this study. According to the spatial variability of wheat growth under within-field level extracted from OMIS image of the key growth stage, management zones were delineated by using K-M and SC-KM methods. Two evaluation indices were employed to evaluate the zoned results of the above mentioned two methods .The results showed that the sum of the weighted variance of the corresponding within-zones based on the two methods appeared no significant difference, and that the SC-KM method could remove lots of isolated cells or patches and improved the continuity of the corresponding management zone map, compared with the K-M method. The zoned result based on the SC-KM method can be used as the variable management unit for precision agriculture and can be used to advise the sampling of subsequent soil or crop.

In this paper, extracting urban areas from Landsat TM images is studied. We proposed a new method for urban and rural residential areas extraction. A classification example based on barren index(BI) is given in this article. We classified the image with several methods and compared the classification results. The result indicates this proposed method based on BI has obvious advantages over conventional multi-band spectral data classification. Classification map created by this method suffers less from a lack of spatial coherency and has a high classification accuracy.

A hybrid inversion scheme for estimating surface variables of vegetation is presented under Bayesian Network theory, and then is used to estimate chlorophyll content of winter wheat leaves and Leaf Area Index (LAI) of canopy. Results using data simulated by coupled models----PROSAIL and those with additional Gaussian white noise show that both LAI and Cab can be estimated with an appreciated accuracy under the proposed scheme, except that there are about 10% of total points falling into failure inversion. Then an uncertain data handling method is employed to solve the failure problem, as a result the failure points are removed successfully though the RMSE of estimated the two variables is larger slightly. The presented hybrid inversion scheme is a knowledge inferring mechanism in principle, so the updated information content in the inversion process is quantitatively calculated thanks to the concept of entropy introduced from thermodynamics..

The reflective spectrum of winter-wheat shows different features with different observed scale, which is important to accurate application of remote sensing. But in many cases, people didn't consider that and the errors were resulted. In this article, firstly the definition of three-scale: material, endmember and pixel was further explained based the previous definition, and the ground measuring data and Omis aerial image of winter wheat of Shunyi, a city in Beijing, explained the difference of three-scale spectrum. To understand the relationship between three-scale spectrums, two methods were used: physical model and statistical method. At the canopy scale, the wheat can be thought as the mixture of leaf and soil. This mixture is nonlinear, the SAIL model, which is based on radiation transfer formula, can be used to simulate the canopy spectrum with the leaf spectrum as input. Mixed pixels and atmosphere effect are the main factors for the difference between canopy and pixel spectrum. But at aerial scale, the wheat canopy is homogeneous and we found for the Omis image, the spectrum of wheat in the same region with changing window is changing in very small extend.

With the development of remote sensing and the technology of computer, computer simulation models are paid more and more attention to research Bidirectional Reflectance Distribution Function (BRDF) of the Earth surface, which can describe vegetations with much more detailed structures and simulate the interaction between light and vegetations on the Earth more reality. As known to all, according to the principles of the BRDF models, physical models of vegetation in the field of remote sensing can be divided into three categories: Geometry Optical models (GO), Radiance Transfer models (RT) and computer simulation models. To understand the process and mechanism of the interaction between light and vegetations better, some computer simulation models are provided, for example, DIANA which is based on the method of Radiosity, RAYTRAN, based on the method of ray tracing and Monte Carlo and so on. In this paper, a model based on Radiosity method is used to simulate the bidirectional reflectance factor (BRF) of summer corn field. It includes three parts: modeling the 3D scene; calculating the radiostiy of the components in the scene; and then accounting BRF of the scene. In the paper, at first 3D structures of corns are reconstructed by an extended L-system based on the measurement in the site of Luancheng Hebei Province, China in 2000. Then the Radiosity- Graphic combined model is applied to simulate the light of the scene of corn, and BRF of the scene of corn will be computed and compared with the measured BRF in the station of Luancheng. Simulated and measured results are fitted very well. Then the hemispherical BRF of the scene are calculated. At last, after analyzing the process of simulation and the results, several advices to advance the Radiosity model are put forward.

The construction of standardized spectral database for typical objects of land surface aims to push development and application of quantitative remote sensing. The database is composed of four main parts including spectral database, model database, prior knowledgebase, and image database. The spectral database contains collection of existed spectral data, with quality evaluation, and collection of new spectral data together with corresponding parameters based on the new standard and criterion. The image database mainly contains a set of typical image data, which are synchronous to the ground spectral measurements obtained by different sensors. The prior knowledgebase widely includes such as data of LULC in different scales, DEM data, phenology data etc. The model database mainly contains models for spectral simulation, and includes also some general models for processing images. The model database is a key part of the spectral database. Though model simulation, database can provide characteristic spectral curves in different phases, scales and states. Furthermore, this database also includes demonstration of typical applications, which can make user to master application method of database quickly and exploit the application field of spectral database according to the detailed requirements. The construction of spectral database mainly provide some functions according to different users as follows: user management, upload and download of spectral data, management and transfer of prior knowledge, query of spectral data and corresponding parameters, model computing, query of remote sensing image, application demonstration, system management, user communication and so on. The core product of database is to provide measured and simulated spectral data of typical ground objects

Recently, there were many strong sandstorms occurred in North China, which became a severely economical, social and environmental problem, especially for Beijing City. The bare soil is an important source of sandstorm, and the croplands naked in winter in 55 counties around Beijing City became the main source of the sandstorm. It is very important to monitor the seasonal bare soil areas using remote sensing data. The multitemporal TM images of Beijing area in May 16, 1997, April 7, 2003, May 25, 2003, October 24, 2003 and January 28, 2004 were selected. Firstly, the smoothing filter-based intensity modulation (SFIM) completed the image fusion. Secondly, the multistage classification approach was adopted. The multisource data, such as DEM, NDVI, bare soil index (BI), shadow index (SI), were integrated and applied in classification, and the seasonal bare soil areas, such as wasteland, cropland, were successfully extracted. Thirdly, the seasonal and inter-annual dynamic changes of the bare soil areas were also analyzed.

The paper focuses on interpreting the different spatial relationships between NDVI and Ts, a triangular or a trapezoid, and analyzing transformation condition and the physical connotation and ecological meaning of the vegetation index-surface temperature feature space. Further, using the Temperature-Vegetation Dryness Index (TVDI), we explain the existent meaning of a triangular shape after NDVI arrives at saturated state (NDVI=1) by analyzing the relationship between NDVI, LAI and evapotranspiration. The specific relations between NDVI and Ts will help us validate and update land surface models well.

The objective in this study was to develop proper vegetation indices for prediction of soil irrigation demanding under vegetation covering conditions. The traditional method for the winter wheat yellow rust field survey is time consuming. It was discussed of the selection method of characteristic spectral bands and the establishing of inversion model to monitor winter wheat yellow rust using hyperspectral data in this study. The correlation coefficients between selected vegetation index and disease incidence (DI) at infected stages. Inversion models between DI and vegetation index such as normalized difference vegetation index (NDVI), ratio vegetation index (RVI), transformed vegetation index (TVI) were used to monitor yellow rust. The multi-temporal hyperspectral airborne images were acquired from winter booting stage to milking stage, and the yellow rust disease of winter wheat was analyzed using hyperspectral images. Compared with healthy wheat, spectral reflectance of disease wheat was higher in 560-670 nm bands but lower in near infrared bands and the absorption depth of chlorophyll in red band and reflectance peak in green band are relatively reduced. A novel spectral index for yellow rust indices was presented, and the degree and area of yellow rust disease were successfully remotely sensed from the multi-temporal hyperspectral data based on spectral index.

Remote sensing offers the potential to determine rapidly the physiological condition of crop over large areas. Canopy reflectance data selected at key growth stages of winter wheat were analyzed. It indicated that winter wheat growth phonological changes can be evaluated by red edge variables (REV) characteristics. Regression equations between chlorophyll concentration, nitrogen concentration, soluble sugar were established, prediction of foliar soluble sugar and chlorophyll concentration by red edge position, foliar nitrogen concentration by the amplitude of red edge, starch concentration by the near infrared plateau, leaf area index by the area of red edge peak are successful and feasible. For this purpose red edge variables can play a vital role in providing time-specific and time-critical information for precision farming, due to their capabilities in measuring canopy spectrum variability.

NDVI is the most widely applied vegetation index, it can indicate the information of growth status, LAI, et al. However, NDVI is also affected by the atmosphere conditions, soil background, view conditions and so on. In this paper, we analyses the effects of chlorophyll, surface soil moisture, the ratio of sky light, view zenith and Sun zenith on NDVI under different LAI using the leaf radiative transfer model PROSPECT and canopy reflectance model SAIL, and find the sensitive factors and ranges.

Uncertainty is an inherent property of Remotely Sensed data. Under the architecture of Bayesian network, which can integrate the quantitative and qualitative knowledge into a comprehensive probabilistic knowledge representation and inference environment, this paper presents a model for data mining from spectra library. Using the filed measured data to drive the model, we obtain the crop structure variables such as Leaf Area Index (LAI) information, e.g. its probability distribution, which can be looked as the priori knowledge of the parameter during the process of inversion.

Computer simulation is one of the vegetation canopy reflectance modeling method, which can realistically simulate the radiation interaction process within vegetation canopies and their dependence on various canopy structural parameters. In this paper, we mainly use the computer simulation methods, combined computer graphics with radiosity equation to simulate the radiative transfer process of the scene in both visible and near-infrared regions. There are file components in our 3D model: computer graphics technique to generate 3D objects; radiosity equation to describe the radiation exchange among the 3D objects; graphics based methods to compute view factors; radiosity computation and 3D image display of the objects; lastly introduced our 3D simulation database built up to put all information in it, such as simulated vegetation canopy's 3D structure files and the spectral data information of the vegetation and scene, in order to convenience users to query and retrieve correlating spectral information

This paper presents a general method and some preliminary results on validating MODIS albedo products, and how albedo changing in the winter-wheat growing period. The available albedo observations are with different scales, including ground measurements and MODIS Albedo products. The study results show that the winter wheat's albedo of both the satellite observation scale and ground measurements scale have the same trend at every growth stage

One of the problems in normalized difference vegetation index (NDVI) based agricultural applications is the time lag between precipitation and NDVI. In this study, the CERES-Wheat model under decision support system for agrotechnology transfer shell was used to simulate the time lag between precipitation and NDVI under minted conditions in the Guanzhong Plain, China. The results showed that there were about 14 to 37 days' time lags between precipitation and NDVI, and the time lags depended on the growing stages of winter wheat. The time lag was about 14 days at the crop's tasselling stage, while it was about 37 days for the crop's reviving stage. The results also indicated that there were year to year variations of the time lags and the time lag should be considered for NDVI based agricultural applications.

With the developing of satellite technology, more and more sensors, based on various resolutions and functions, are launched to the sky to perform their missions. Enormous data are sent back to the Earth stations. In deriving surface parameters using these remotely sensed data, the transportability of algorithms from one resolution to another often cause the scale effect because of the surface heterogeneity on the Earth which can induce the change of reflectance. The problem, that the change of reflectance data affected by discontinuity as part of surface heterogeneity impacts the retrieval of vegetation leaf area index (LAI), is addressed in This work. Two cases, inducing the scaling issue in deriving surface parameters of interest, are considered here. One is the discontinuity between contrasting cover types within a mixed scene, the other is the nonlinear relationship of NDVI and LAI. Therefore, it is necessary to apply the correction based on NDVI-LAI relationships to modify scaling problem. In the processing, considering the field of wheat, firstly a series of 3D scenes with wheat and soil mixed are made based on the field measurement; secondly, computer simulation model $the method of radiosity, which can calculate the balance of light energy in the simulated scenes, is used to model BRF (bi-directional reflectance factor) of these scenes. If the NDVI-LAI relationship from homogeneous scenes can be taken as standard, the relationship from heterogeneous scenes will be modified according to contextural parameter. Some conclusions are drawn from the investigation: (1) different distributional contextures of vegetation even with the same LAI affect the reflectance heavily; (2) we compare the reflectance simulated by the method of radiosity with the mean reflectance calculated using the area-weighted linear relationship of reflectance from components, and find that the accuracy of the mean reflectance can be accepted so that the linear equation to calculate the reflectance of mixed pixels is reasonable to relate images with high and low resolution; (3) using contextural parameter for quantifying the scale effect can get promising results.

In our former research, we raised the concept of histo-variogram to describe the spatial heterogeneity among the Earth objects compactly based on the analysis of the characteristics of other spatial analyzing methods such as variogram information entropy. Just like all the other nature phenomenon, the spatial heterogeneity is also scale dependent. That is, the heterogeneity of one kind of Earth scene at a certain scale may become homogeneous at the other scale and vice versa. In current research, we want to find: (1) if there exists one scale scope in which the heterogeneity keeps relative stabilization; (2) if the heterogeneity of a specific Earth scene has regional coherence. This is very important for model selection or parameter adjustment during inversion of quantitative remote sensing. We use the concept of "total fractal dimension" that arose in our former research to measure the heterogeneity of the Earth scene. The data source we used is stochastic sampled sub-regions from LUCC of Beijing district with different scales. The preparatory result shows that the heterogeneity of different Earth scenes has its specific scale scope and heterogeneity of most Earth scenes have regional coherence.

In the process of measurement, a number of factors will affect the quality of data. Therefore, data must be verified and assessed before their applications. This work discussed some methods for quality assessment of crop spectral data, which include methods of analysis of spectral characteristics, statistical test and spectral simulation. The method of spectral analysis compares measured spectra with the reference spectrum, and analyzes the location of wave crest, wave trough and the shape, intensity of spectral reflectance curves. The method of statistical test consists of shape similarity test and intensity test. The shape similarity test analyzes the correlation between measured spectral data and the reference spectral datum over the special wavelength range and assesses quality of the measured data by the correlation coefficients. The intensity test calculates the mean value and standard deviation of spectral data, and forms a spectral zone around the mean value. We consider it as the abnormal one if one spectral curve goes beyond the spectrum zone. The method of spectral simulation mainly compares measured spectra with simulated spectra by combined PROSPECT-SAIL model. Results show these methods of quality assessment are feasible.

Chlorophyll and carotenoid, relating to the physiological function of leaves, are two pigments which can absorb the light energy during the process of plant photosynthesis. Among the pigments, the chlorophyll plays an important role in the photosynthesis, and its content, as a predictor of the nutritional status of vegetation, is one of the main factors to evaluate the environment and growth conditions for the winter wheat. This work, based on the reflectance spectra of wheat in Xiao Tangshan County, in China, took PLS regression as the quantitative inversion method to have established the hyperspectral inversion model between chlorophyll content and the reflectance spectra of wheat. Through analysis, it indicated that the chlorophyll content of wheat was highly relative to the reflectance of hyperspectral from 350 nm to 1060 nm. The correlation coefficient between the prediction value and the measured value is as high as 0.9, and the RMSEP is lower than 0.4. The research provided an effective method to estimate the chlorophyll content using the quantitative inversion technology of hyperspectral RS.

Traditional vegetation indices are usually constructed by using red and near-infrared band reflectance data under single solar incidence-observation geometry. However, because of the anisotropy of the Earth surface's reflectance, vegetation indices acquired from different solar-incidence observation geometries exhibit lots of variances. Meanwhile, most of those indices only utilize vegetation's spectral information, and anisotropic reflectance of vegetation is considered as a disturbing factor rather than a source of vegetation's structural information. In this paper, kernel-based vegetation index (KVI) is constructed based on the semi-empirical kernel-based BRDF model parameters. Validation results of ground measured bidirectional reflection data of different vegetation types show: kernel-based vegetation index has better linear relationship with corresponding vegetation's leaf area index (LAI) than widely used normalized difference vegetation index does. The results of upscaling KVI from local scale to global large scale suggest the effect of scale needs to be considered when using it to different scale data sets. This study suggests that KVI provides a new method of better using multi-spectrum and multi-angle reflectance data, and has certain potential for multi-angular remote sensing applications

Kernels are always pre-determined in current kernel-driven model applications, but they seem to have some disadvantages in the requirement for more accurate remote sensing because one kernel combination is used for the inversion of all land cover types. In this paper, we use 28 different multi-angular data sets, which represent major types of land cover, to find the relations of different kernel selections with land cover types. The kernel combinations in the models we compare are volume kernels of Ross-Thick, Ross-Thin and geometric optical kernels of Li-Transit, Li-SparseR and Li-Dense. The airborne multi-angle TIR/VNIR image system (AMTIS) data set, which was obtained in Shunyi county of Beijing, China in April 2002, was used for the inversion. The inversion results of pre-determined kernel selections and class-based kernel selections are compared.

With the development of remote sensing, new models are available continuously. In order to extend the practicability of the model library, the authors introduce component object model (COM) technique. COM is a software architecture that allows the components made by different software vendors to be combined into a variety of applications. Remote sensing model library is composed of three parts, common objects, model objects and accessorial objects. Common objects include input/output procedure, solar angle calculating procedure in bi-directional reflectance, and metadata about all the models. Model objects include the models contributing to quantitative remote sensing applications, which comprise system models, simulant models and application models. Accessorial objects include prior knowledge, measurement data and image data. In the article, the executable project is validated with an instance in the end. The spectrum of typical land surface observed in nadir viewing direction is simulated in pixel scale. In this process, crop model, PROSPECT model and SAIL model are used to calculate the spectrum character of the pixel. Crop model is used to simulate leaf area index, and PROSPECT model is to simulate reflectance and transmission of leaves. The final result is calculated with SAIL model. We simulated the spectrum of winter wheat in given growing season. In the work, each model and each common object are designed to be components. The model library based on COM technique adapts to the progress and is propitious to be expanded and modified.

Eddy covariance and Bowen ratio measurements were carried out over different surfaces, and used to characterize the daily variation of energy fluxes and energy balance under different weather condition. These will be used to identify the phenomena to be addressed in future modeling works. Results show that (1) The diurnal variation characteristics of the fluxes were different over different surfaces, especially the latent heat flux, (2) The energy imbalance persisted in different surfaces with an average about 20%; the energy balance closure was better in the afternoon than in the morning averagely, possibly suggesting the underestimation of storage terms, which are usually larger in the morning.

This paper describes the atmospheric correction algorithm for the Airborne Multi-angle TIR/VNIR Imaging System (AMTIS) in VIS/NIR bands. The method is named as a BRDF loop based algorithm. Under the assumption of that the ground cover is non-Lambertian, MODTRAN4 is used to calculate the atmosphere parameters, and the parameters used to decouple the multi-bouncing effect such as direct-hemisphere reflectance and hemisphere-direct reflectance are estimated in the loop correction process. The BRDF model used to calculate albedo is the kernel-driven model. To accelerate the atmospheric correction, all the atmospheric parameters used in this algorithm are pre-computed used MODTRAN4.1 and saved in the look-up table. A prior knowledge is used in the inversion of kernel-driven model when the multi-angle sampling is not enough. The atmospheric correction result is validated using the synchronous ground measurements. From the comparison with the general Lambert assumption based method, it is found that our BRDF loop based method can partly remove the atmospheric smoothing effect.

The Airborne Multi-angle TIR/VNIR Imaging System (AMTIS) samples the surface at a number of view angles and offers the potential of retrieval of atmospheric aerosol properties, land surface bidirectional reflectance etc. This paper presents the retrieval algorithm of aerosol optical depth and single scattering albedo from visual and near-infrared bands of AMTIS based on a simplified path radiance model. Atmospheric parameters such as molecular scattering and absorption are calculated using MODTRAN4. Under the assumption that the aerosol optical depth above the sensor is not affected by the surface, the aerosol optical depth above the sensor (4.2 km) is also calculated using MODTRAN4. The path radiance is divided into two components: the singly scattered radiance and multiple-scattering radiance. The AMTIS images acquired on April 11, 2001 in the Shunyi experiment are used in the retrieval. The algorithm performs best over dark surfaces, such as water. The retrieved aerosol optical depth is close to the result from the synchronous Sun photometer data with an error about 0.05/spl sim/0.1. Retrieved single scattering albedo is very close to that of the continental aerosol model of 6S.

atmospheric profile is available. Under the assumption that the surface emissivity is isotropic, two atmospheric parameters are needed to remove the atmospheric effect: the water vapor content (W) and the effective mean atmosphere temperature (Ta). Ta can be estimated from the muti-angular observations of the pixel with “minimum standard deviation” of brightness temperatures. After Ta is known, W can be retrieved from the pixels with multiangular observations under the assumption of isotropic emissivity and horizontal uniform atmosphere. Then the atmospheric effect can be removed after the two atmospheric parameters are known. Through the sensitivity analysis of the algorithm to emissivity and transmittance, it can be found that W is very sensitive to the error of emissivity. However, it can also be estimated using the pixel with known surface temperature and emissivity measured synchronously.

This paper mainly proposed using multi-temporal spectral data with given weight to estimate yield using contribution ratios of different stages to improve yield estimation, then use stepwise regression to build models for yield estimation. The contribution ratio is calculated by principal component analysis respectively. Result show that this methodology reflects the crop growth status, physiological characters at different stages, and improves the accuracy obviously.

The Spectrum Knowledge Base System (SKBS) for Typical Object in China, built up by taking advantage of J2EE technology, is capable of providing the functionalities in spectrum analysis, query and comparison. More importantly, the spectrum scale effect, especially the scale extension, can be achieved in SKBS, which is based on the model-driven theory with the support of the prior knowledge.

In this paper, we proposed the construction of a spectral knowledge library, which is composed of observed data library, image library, prior knowledge library and remote sensing model library. We stated the status and functions of remote sensing model library in the whole library, and discussed its structure, system architecture and development technique under network environment We also discussed remote sensing model library's driving mechanism under the support of spectral knowledge library and meta data.

One problem in quantitative remote sensing inversion is the correlations between variables. Path analysis is a statistical technique that differentiates between correlation and causation, features multiple linear regressions, and generates path coefficients. In this paper, path analysis was applied to study the correlations and causations of two cases in remote sensing reversion. One is the retrieval of land surface temperature, and another is to explain the results of land surface moisture estimation. We found that path analysis can be used to study the direct effect and the indirect effects of a variable in remote sensing inversion, and gave a better explanation of the results of multiple linear regression analysis.

At XiaoTangShan Precision Agriculture Experiment Base, suburban of Beijing city, we obtained the hyperspectral data of wheat canopy by field measurement and synchronal relative foliage water content (RFWC) at lab, according to the different growth stage of wheat, in the spring of 2002. In this paper, we extracted spectral feature parameters firstly. Then, the correlation analysis between the spectral features parameters and RFWC was made. two spectral feature parameters, which have good relativity to foliage water content, to make linear regression models between RFWC and spectral feature parameter according to the growth stages of wheat.

Albedo is a key parameter in monitoring the energy exchanges between the solar radiations and the land surfaces. The MODIS team generates the albedo products every 16 days. The products need to be validated by ground truths under different environmental conditions. In this study, we developed a 3-step validation procedure. The Ambrals (Algorithm for Modeling Bidirectional Reflectance Anisotropies of the Land Surface) model inversion was used to retrieve the albedo from the measured BRDF data over the winter wheat fields at the point/plot scale. And then, as our second step, the albedo values from the Airborne Multiangular Thermal-infrared Imaging System (AMTIS) over the same target area were estimated and validated using the ground point measurements. Finally, the retrieved albedo from airborne data were aggregated and compared with the MODIS albedo products. Our validation procedure has demonstrated a practical method to validate that albedo from spacebrone remotely sensed data (e.g., MODIS). The validation results show that the MODIS albedo products are reasonably good. Albedo is a key parameter in monitoring the energy exchanges of land surfaces. The hemispherical albedo is traditionally observed by albedometer at local meteorological stations, where the observing targets are usually grassland in a specific environment. Because some applications require albedo over a large area, retrieving regional and global albedo products from remote sensing observations can be more productive. The MODIS albedo products are from the multi-angular remote sensing (MARS) observations of every 16-days accumulation. The production needs to be validated by ground truths. One of the main problems in the validation is that the field-measured albedo is different in scale from the albedo retrieval using remote sensing data. The albedometer field measurement is over a small area, less than 1m 2 , while the spatial resolution of the MODIS albedo product is about 1 km. Another problem is associated with the different wavebands between the albedometer and the MODIS sensors. As a possible solution, we created a 3-steps validation procedure. As the first step, we used the BRDF data measured in the field to retrieve the albedo by Ambrals model inversion. The observing target was winter wheat. The retrieved albedo is comparable with that one measured by albedometer since both measurements are in the same observing scale. The effect of the wavebands difference was also corrected at this step. In the second step, we retrieved the albedo from the airborne MARS observation data of the same target. The spatial resolution is 1.36m at nadir. The retrieved albedo from airborne AMTIS BRDF data can be validated by using our field measurement. Finally, the retrieved albedo from airborne data was compared with the MODIS albedo product. Scaling-up needs to be considered in the comparison. In this work, the field measurements and airborne data came from the large satellite-airborne-ground synchronous experiment in the April of 2001. The experimental region is in the Shunyi county, 50km northeast of the Beijing City, China.

LAI (Leaf Area Index) is an important parameter for plant growth status detecting. The traditional method to measure leaves' area in field is time and labor consuming. In some remote sensing applications, LAI might be needed more intensively such as daily and weekly. Crop simulation models can be effective tools for simulating LAI, crop and soil water statuses, crop growth and development parameters at the daily step. Ground LAI measurements for winter wheat were carried out in several fields of Shunyi county, Beijing, China at several days' interval during the crop growth season. The CERES-wheat model is run under local soil, weather and management conditions of a field site to simulate daily LAI values. The simulated LAI corresponded comparatively well with the measured ones at the early stage. The results suggested that the technique might be promising for estimating LAI, this make it possible to develop an approach to assimilate LAI of winter wheat between ground measurements and CERES-wheat simulation.

In the field of remote sensing, it is important to understand interaction between light and vegetation. The interrelation of them has been addressed in many works, and many different radiant models of vegetation have been proposed, such as: geometrical optical models, turbid medium models, hybrid models and computer simulation models. With developing of quantitative remote sensing research, computer simulation models, for example, Monte Carlo simulation model and Radiosity show their importance in analyzing the experimental data. In order to continue calculating the reflectivity from the vegetation by using a computer simulation model, it is essential to build the 3D structure of the vegetation. Therefore, many 3D structure data and optical parameters about the real winter wheat were measured firstly, i.e. height of stem, positions and sizes of the leaves, distributions on the field of wheat. Because these data are numerous and discrete, it is very difficult to simulate the virtual scene with them directly. To cope with it, we arranged all data and parameters in several layers based on the object oriented technique. Moreover, in order to simplify and deduce the structural variables that will be applied to build the 3D visual winter wheat model, we analyzed experimental data statistically in the process of realistic structural model. Several geometric and logical relations about structural variables were developed subsequently, and some variables varying with season were summarized to get the simple regulation with the purpose of simulating growing process of the winter wheat. The extended Lindenmayer system (L-system) method is then used to simulate the virtual scene of winter wheat by giving a few structural variables simplified before. Once the simulation is correct, scattering and reflectance from the 3D structural scene can be calculated using the Monte Carlo simulation model or Radiosity and so on. Our results show that (a) our lighting simulation system efficiently provides the required information at the desired level of accuracy, and (b) the plant growth model is extremely well calibrated against real plants. Furthermore, the method and the relations developed in this paper can be used in other subjects, such as computer graphics.

We propose an iterative multistage inversion (IMI) algorithm to retrieve the land surface component temperatures for nonisothermal vegetation canopy. Our algorithm is based on a thermal emission model that can simulate the directional effects from the nonisothermal surfaces. Our IMI algorithm just inverts the most uncertain and most sensitive parameters at each step using the most sensitive observation samples, and then adjusts the initial values based on the retrieval results. This inversion process is repeated until convergence condition is satisfied. Compared with the inversion method that try to invert all of the parameters at the same time, the IMI algorithm tends to give more accurate mean values for the parameters and is more stable when the noise level is relative low.

Leaf area index (LAI) is an important parameter for describing vegetation canopy structure in the terrestrial ecosystem. LAI is closely related to plant transpiration, sunlight intercept, photosynthesis and Net Primary Productivity. Multiangular remote sensing is capable of providing more three-dimension information of vegetation, and it is powerful in solving the problem of the same object with different spectrum or vice versa. As a result, multiangular remote sensing and Bidirectional Reflectance Distribution Function (BRDF) model based inversion may be more suitable for Leaf Area index (LAI) retrieval over row crop canopies. However, it's still difficult to get LAI without enough a priori knowledge due to the underdetermined problems in inversion. We use the multispectral information to get the a priori estimation of LAI, and then perform BRDF model inversion. Different from the general one channel based BRDF model inversion methods, our new methods use the muiltiangular and multispectral data sets together to increase the available information in inversion, i.e., it is a synthetic method. From the inversion results we found that the new synthetic method is more effective in LAI inversion.

The problem of quantitative remote sensing inversion is ill-posed in essence. In order to turn the inversion problem from uncertainty into certainty, we must try to make full use of all the Information we have. Furthermore, reducing the number of parameters to be inversed is another good way. For thermal inversion, separation of land surface and atmospheric contribution (SAS) to the remote sensing signal is a necessity to both atmosphere research and land surface research. Taking advantage of the spatial information generated from the visible and near infrared spectral bands, we presented an algorithm to get the relative contribution to remote sensing signal (radiance) from land surface and atmosphere. The algorithm is based on the difference of the spatial thermal pattern between land surface and atmosphere. As a result, regressive image and difference image is generated. Regressive image contains the contribution from land surface and homogenous part of atmosphere. It has similar spatial pattern with land surface. And difference image contains the contribution of inhomogenous part of atmosphere. It reflects the spatial pattern of atmosphere. The test with MAS (MODIS Airborne Simulator) data shows that SAS algorithm has the ability to remove the thermal abnormal from the original thermal image. The real test proves that most of chippy cirrus cloud has been removed from the observed radiance and little effect remains in the regressive image. From the study, we can draw a conclusion that visible and near infrared image is helpful to thermal inversion. It also reveals that spatial information is one kind of good information source that should be used in remote sensing inversion.

Remote sensing inversion problem is always ill-posed. However, regularization aims at turning the ill-posed problems into certainty. In this paper, taking the linear kernel-driven model as an example, we put forward the maximum entropy algorithm based on information theory to determine the Tikhonov regularization parameter. Then, analyse and compare it with other mature methods. Result shows that the maximum entropy algorithm has advantages when the variance of the observations' noise is small or uncertainty of the prior knowledge is not too large.

Multi-angular remote sensing became a hot topic after the non-Lambert characteristic of the Earth's surface had been accepted popularly. A large amount of multi-angular remote sensing data has been obtained with the launch of multi-angle remote sensing sensors. Therefore, modeling of the bi-directional reflectance distribution function (BRDF) for the Earth objects is one of the main subjects at present. A large satellite-airborne-ground synchronous remote sensing experiment was carried out during March 29 to May 10, 2001 at Shunyi, China. The main observation target in this experiment is focused on winter wheat. To describe the BRDF of winter wheat in its early growing stages, we propose a geometric-optical model that is suitable for sparse vegetation, and then try to retrieve the structure parameters based on this model using the field measurements. The purport of the model and its inversion is to inspect the ravages of drought on the wheat just as it is turning green. The winter wheat in our measurement field is sparse and disperses without clear row structures in its turning-green stage. Typical row structure based models and uniform structure based models are not suitable. Our model is developed based on the Li-Strahler geometrical-optical model proposed in 1985. Each cluster of wheat is treated as a hemi-ellipsoid in this model. All of the leaves in the cluster are assumed to cover the hemi-ellipsoid randomly. Leaf area index and leaf angle distribution are two important parameters that are related to the surface area of the hemi-ellipsoid and the leaf distribution on this surface respectively. Leaf angle distribution is also related to the shape of the hemi-ellipsoid. Due to the large uncertainty of the number of hemi-ellipsoids in a unit area, we retrieve this parameter based on our model using the most sensitive samples first, and then treat it as a priori knowledge in the later inversion. The next stage is studying how to use multi-angle remote sensing data to invert vegetation structure parameters.

Predictions of climate change typically use a GCM linked to a land surface model. Land surface models, e.g. Biosphere-Atmosphere Transfer Scheme (BATS), estimate the albedo of trees over snow roughly with the parameters of roughness length, z/sub 0/, and snow depth, d. Based on their work, we further consider the difference in directional-to-hemisphere albedo for different solar zenith angle (SZA), and leaf area index (LAI) dependence. In order to keep the basic feature of the BATS model and to add these two new features, we simplified the geometric optical and radiative transfer (GORT) hybrid model of Li, et al. [1995] to reach this purpose. This model can be rather simple to retrieve the albedo of remote sensing pixel. It can be a strong tool to understand the climate system.

Local statistic-based fusion algorithms are discussed, which can be applied to fuse high-resolution VNIR images and a single low-resolution TIR image. These algorithms are based the experiments that structural information of observed objects In visible and near-infared spectral range (VNIR) is essentially correlated with the environmental information (especially moisture information) in thermal infrared spectral range (TIR). The local is performed at three scales. This algorithm can be useful method to merge the TM image and VNIR images.

The remotely sensed images can be understood and analyzed better through comparing the remote sensing spectrum and the spectrum with known land surface structure and physical chemistry parameters. It is unavoidable to compute land surface spectrum based on prior knowledge and remote sensing physical models when we cannot obtain the measured spectrum on suitable time and pixel size. The remote sensing knowledge base integrates data, image, and physical models for quantitative remote sensing. Here in our remote sensing spectral knowledge base, the digital elevation model, land cover and land cover maps and soil map are restored in the spatial database. The climatic changes and spatial patterns of vegetation are restored in the expert system. And component spectrum of vegetation, soil, water, snow, minerals, man-made object are measured and also stored in the database. It is apparent that they themselves have data systems and the systems usually are different with input and output parameter systems of remote sensing physical models. In order to solve the discrepancies, metadata of data and model are put forward. About design of the knowledge base, defining and organizing metadata is key task. Accordingly to data, image and model in the knowledge base, the metadata also consists of metadata about data, metadata about image, and metadata about model. Metadata describes formation, quality and meaning of data, image and model. Data exchanges between database, image base and model base are found on the metadata, and is coordinated by an expert system based on rule, so data extraction and re-organizing will obtain flexible as large as possible. The data-engine extracts data from databases and transfers the data between database and model base, and the model-agent selects suitable models to extend the structure and physical chemistry parameters, and extract land surface facts, finally compute the land surface spectrum on multi-temporal & spatial scales

The sources of the inverse error of remote sensing physical models are analyzed and divided into two groups. From the point of view of controlling these errors, a synthetic inverse scenario is put forward. A case study using simulated data shows that this scenario is better than ordinary methods in robustness and global convergency.

According to the Li-Strahler model, a BRDF model of row structure with gap, in which three components and six paths are taken into account, is established through analysis of the process of radiation transfer. The model is verified by the data obtained in the Satellite-Airborne-Ground Synchronous Quantitative Remote Sensing Experiment implemented from March 29 to May 10 in 2001, in Shunyi, Beijing, China. As is shown, the model can demonstrate effectively the reflection of crops with row structure.

The main problem in the validation is that the field-measured albedo is different in scale from the albedo estimated using remote sensing data. The albedo-meter based field measurement is over a small area. On the contrary, the spatial resolution of the MODIS albedo production is about 1 km. Another problem is the different wavebands or albedo-meter and MODIS sensor. As a solution, we suggest to use the field multiangular measurements data that is captured in a small field of view (FOV) to retrieve the albedo by Ambrals model inversion. As a result, the retrieved albedo is comparable with that measured by albedo-meter since they have the same scale. At the same time, the effect of wavebands difference is also corrected. In the second step, we extend this method to the airborne MARS observations of the same target. Scaling-up can be considered based on the two inverted albedos. The retrieved albedo from airborne data can be validated using field measurement too. Finally, the airborne retrieved albedo can be used to validate the MODIS albedo production. In this framework, the field measured data sets come from a large satellite-airborne-ground synchronous experiment that was taken in Shunyi, which is in the north of Beijing, in April of 2001.

We take the kernel-driven model as an example, focus on the information content definition and calculation of multi-angular remote sensing (MARS) data. We study four methods to measure the information content of MARS data: Fisher statistic, information entropy, determinant and sum of the diagonal elements of the information matrix, how to use the Fisher statistic theory and information entropy to measure the information content of MARS data, to calculate the information content of the dataset on the three unknowns for different subsets of the data. The analyses show that information entropy is a good tool for measuring information content of MARS data.

We propose a thermal bidirectional gap model to describe the thermal directional emission from row crop canopies. An important concept of overlap index is used in this model to express the correlation between the gaps in the Sun and view directions. Detailed directional thermal emissions, row structure, LAI, component temperatures were measured in the experiment taken in Shunyi China, 2001. These data are used to validate our model. As an illustration, we compared our bidirectional gap model with the model that doesn't consider gaps (Kimes model) and the model only consider gaps in view direction. It is found that our model gives out the closest results to the field measurements.

A priori knowledge can come from field measurements of bidirectional reflectance factors for various surface cover types. How to express and use this kind of knowledge is very important currently. 73 sets of field observations are used to explore the possible expression of a priori knowledge in linear kernel-driven BRDF models in this paper.

We present an algorithm to retrieve land surface temperatures from wideband thermal infrared measurements using the model of Li et al. (1999). Forward simulation and inversion demonstrates the method to be stable in the presence of observation noise. Results from inversions performed using field measurements show that errors are generally less than the uncertainty in the observations.