Multiple statistical approaches for the discrimination of mangrove species of Rhizophoraceae using transformed field and laboratory hyperspectral data

This study aims at discriminating eight mangrove species of Rhizophoraceae family of Indian east coast using field and laboratory spectra in spectral range (350–2500 nm). Parametric and non-parametric statistical analyses were applied on spectral data in four spectral modes: (i) reflectance (ii) continuum removed, (iii) additive inverse and (iv) continuum removed additive inverse. We introduced continuum removal of inverse spectra to utilize the advantage of continuum removal in reflectance region. Non-parametric test gave better separability than parametric test. Principal component analysis and stepwise discriminant analysis were applied for feature reduction and to identify optimal wavelengths for species discrimination. To quantify the separability, Jeffries–Matusita distance measure was derived. Green (550 nm), red edge (680–720 nm) and water absorption region (1470 and 1850 nm) were found to be optimal wavelengths for species discrimination. The continuum removal of additive inverse spectra gave better separability than the continuum removed spectra.     

Spectroradiometric requirements for the reflective module of the airborne spectrometer ARES

The Airborne Reflective/Emissive Spectrometer is specified as a whisk-broom imaging spectrometer for remote sensing of land surfaces covering the wavelength regions 0.47-2.45 /spl mu/m and 8-12 /spl mu/m with 160 spectral bands. The instrument is being built by Integrated Spectronics, financed by the German Aerospace Agency (DLR) and the GeoResearch Centre Potsdam (GFZ) and will be available to the scientific community from end 2005 on. The spectroradiometric design is based on scientific requirements derived from three main application scenarios comprising vegetation, soil, and mineral sciences. Two of these are described in this letter. Measured or modeled reflectance spectra are input to a simulation model that calculates at-sensor radiance spectra, resamples them with the channel-specific response functions, adds different amounts of noise in the radiance domain, and performs a retrieval to get the corresponding noisy surface reflectance spectra. The retrieval results as a function of the sensor noise level are compared with the accuracy requirements imposed by the different application fields taking into account the technical boundary conditions. The final specifications account for the most demanding requirements of the three application fields: a spectral sampling distance of 13-14 nm in the 470-1800 nm region, and 12 nm in the 2000-2450-nm region. The required noise-equivalent radiances are 5, 3, and 2 nW/spl middot/cm/sup -2//spl middot/sr/sup -1//spl middot/nm/sup -1/ for the spectral regions 470-1000, 1000-1800, and 2000-2450 nm, respectively.     

利用青海湖水面辐射校正场对FY—1C气象卫星热红外传感器进行绝对辐射定标

利用青海湖水面场FY-1C气象卫星热红外窗区通道进行绝对辐射定标,由CE312野外热红外辐射计在水面测得辐亮度经大气订正传递到卫星入瞳处,考虑到大气吸收削弱影响,同时卫星观测路径大气产生热发散,这两部分对卫星信号的贡献由探空廓线和卫星观测几何输入MODTRAN37计算出来,同时进行CE312野外辐射计与卫星通道光谱响应匹配计算,最终得到卫星入瞳处的表观亮度.这个辐亮度与卫星通道的计数值比较得到该通道绝对定标系数.结果表明利用辐射校正场辐射定标与星上定标相差5%左右,相当于3K的亮温差.     

Is There a Physical Linkage Between Surface Emissive and Reflective Variables Over Non-Vegetated Surfaces?

For a satellite sensor with only one or two thermal infrared channels, it is difficult to retrieve the surface emissivity from the received emissive signal. Empirical linear relationship between surface emissivity and red reflectance are already established for deriving emissivity, but the inner physical mechanism remains unclear. The optical constants of various minerals that cover the spectral range from 0.44 to 13.5 μm in conjunction with modern radiative transfer models were used to produce corresponding surface reflectance and emissivity spectra. Compared to the commonly used empirical linear relationship, a more accurate multiple linear relationship between Landsat TM5 emissivity and optical reflectances was derived using the simulated data, which indicated the necessity of replacing the empirical relationship with the new one for improving surface emissivity estimate in the single channel algorithm. The significant multiple linear relationship between broadband emissivity (BBE, 8–13.5 μm) and MODIS spectral albedos was also derived using the same data. This paper demonstrates that there is a physical linkage between surface emissive and reflective variables, and provides a theoretical perspective on estimating surface emissivity for sensors with only one or two thermal infrared channels.     

Atmospheric Correction Over Coastal Turbid Waters of Bay of Bengal Using OCEANSAT-I Ocean Colour Monitor (OCM) Data

Space-borne ocean-colour remote sensor-detected radiance is heavily contaminated by solar radiation backscattered by the atmospheric air molecules and aerosols. Hence, the first step in ocean-colour data processing is the removal of this atmospheric contribution from the sensor-detected radiance to enable detection of optically active oceanic constituents e.g. chlorophyll-a, suspended sediment etc. In standard atmospheric correction procedure for OCEANSAT-1 Ocean Colour Monitor (OCM) data, NIR bands centered at 765 and 865 nm wavelengths were used for aerosol characterization. Due to high absorption by water molecules, ocean surface in these two wavelengths acts as dark background, therefore, sensor detected radiance can be assumed to have major contribution from atmospheric scattering. For coastal turbid waters this assumption of dark surface fails due to the presence of highly scattering sediments which causes sufficient water-leaving radiance in NIR bands and lead to over-estimation of aerosol radiance resulting in negative water leaving radiance for λ < 700 nm. In the present study, for the turbid coastal waters in the northern Bay of Bengal, the concept of spatial homogeneity of aerosol and water leaving reflectance has been applied to perform atmospheric correction of OCAEANSAT-1 OCM data. The results of the turbid water atmospheric correction have also been validated using in-situ measured water-leaving radiance. Comparison of satellite derived water-leaving radiance for five coastal stations with in-situ measured radiance spectra, indicates an improvement over the standard atmospheric correction algorithm giving physically realistic and positive values. Root Mean Square Error (RMSE) between the in-situ measured and satellite derived water leaving radiance for wavelengths 412 nm, 443 nm, 490 nm, 512 nm and 555 nm was found to be 1.11, 0.718, 0.575, 0.611 and 0.651%, respectively, using standard atmospheric correction procedure. By the use of spatial homogeneity concept, this error was reduced to 0.125, 0.173, 0.176, 0.225, and 0.290 and the correlation coefficient arrived at 0.945, which is an improvement over the standard atmospheric correction procedure.     

Hyper-spectral data based investigations for snow wetness mapping

Spatial information on snow wetness content (SWC) is important for hydrology, climatology applications. Limited work is available on estimation of SWC using optical sensors. In present work, spectral signature characteristics of snow (~145 samples) acquired in winters of three years, using field spectral-radiometer (350–2500 nm) were correlated with synchronized SWC measurements. Correlation is found stronger in Near-Infra-Red (NIR) and Short-Wave-Infrared (SWIR) regions than Visible (VIS). Spectral peak width at 905 and 1240 nm is found negatively correlated with SWC, while positively correlated at 1025 nm. Asymmetry tends towards right as SWC increases and has stable positive correlations as compared to other characteristics. Sensitivity of widely used snow-related indices to SWC is also analyzed. Based on analysis, new ratio method at selected wavelengths is proposed to discriminate dry and wet snow zones using air/ground borne sensors. Proposed methodology is evaluated on air-borne hyper-spectral (AVIRIS-NG) data and 88% overall accuracy with kappa coefficient 77.6 observed after validation with reference observations.     

Spectral characteristics of sensors onboard IRS-1D and P6 satellites: Estimation and their influence on surface reflectance and NDVI

This paper reports the results of a modeling study carried out with two objectives, (1) to estimate and compare effective spectral characteristics (central wavelength, bandwidth and bandpass exo-atmospheric solar irradiance E_o) of various spectral channels of LISS-III, WiFS, LISS-III*, LISS-IV and AWiFS onboard Indian Remote Sensing Satellites IRS-ID and P6 using moment method based on the laboratory measurements of sensor spectral response, and (2) to quantify the influence of varying sensor spectral response on reflectance and Normalized Difference Vegetation Index (NDVI) measurements using surface reflectance spectra corresponding to different leaf area index conditions of crop target obtained through field experiment. Significant deviation of 4 to 14 nm in central wavelength and 1.6 to 14.07 nm in spectral width was observed for the corresponding channel of IRS sensors. Coefficient of variation of the order of 0.1 to 1.11% was noticed in E_o among various IRS sensors, which could induce a difference of 0.72 to 3.35% in the estimation of top of atmosphere reflectance for crop target. The variation in spectral response of IRS sensors implied a relative difference of the order of 0.91 to 3.38% in surface reflectance and NDVI measurements. Polynomial approximations are also provided for spectral correction that can be utilized for normalizing the artifacts introduced due to differences in spectral characteristics among IRS sensors.     

An Improved Band Selection Technique for Hyperspectral Data Using Factor Analysis

This paper discusses a statistical and band transformation based approach to select bands for hyperspectral image analysis. Hyperspectral images contain large number of spectral bands with redundant information about the spectral classes in the image scene. It is necessary to reduce the high dimensionality of the data for the processing of hyperspectral data. We report a feature selection technique that removes correlated spectral bands using band decorrelation technique and obtains maximum variance image bands based on factor analysis. Factor analysis method of band selection technique is also validated against existing methods of band selection. The study is carried out for the agriculturally rich area of Musiri region of South India that has varied landcover types. Evaluation of the band selection procedure is done using signature separability measures such as Euclidean distance, Divergence, Transformed divergence and Jeffries Matusita distance. Results indicated that selected bands exhibited maximum separability and also occurred predominantly at wavelength 700 nm, 850, 1000 nm, 1200 nm, 1648 nm and 2200 nm.     

Modelling of radiation transfer in Earth’s hazy atmosphere and calculation of path radiances in IRS LISS-I Bands

A model (named ATMRAD) of radiation transfer in the Earth’s hazy atmosphere based on simplistic solution of radiation transfer equation and tropical model atmosphere is developed to estimate path radiance for correction of remote sensing multispectral data. The atmosphere is assumed to be cloud free and plane-parallel stratified with various quatities specified for 46 layers in its 100 km assumed thickness and model altitude profiles of haze aerosols specified for 23 km and 5 km visibility ranges. Altitude profiles of haze for any intermediate visibility ranges are deduced from linear mixing of the two profiles for 23 and 5 km visibilities. Deirmendjian size distribution function is assumed for aerosols. The radiation transfer calculations have been performed for 126 wavelengths in 0.25 to 3.0 μm range by considering Rayleigh and Mie Scatterings by air molecules and aerosols; and absorption by atmospheric gases. The model calculations have been carried out for path radiances in bands of IRS LISS-I sensor for different visibility ranges and solar illumination angles.     

Contribution of Landsat 8 data for the estimation of land surface temperature in Batna city,Eastern Algeria

In this study, we presented a mono-window (MW) algorithm for land surface temperature retrieval from Landsat 8 TIRS. MW needs spectral radiance and emissivity of thermal infrared bands as input for deriving LST. The spectral radiance was estimated using band 10, and the surface emissivity value was derived with the help of NDVI and vegetation proportion parameters for which OLI bands 5 and 4 were used. The results in comparison with MODIS (MOD11A1) products indicated that the proposed algorithm is capable of retrieving accurate LST values, with a correlation coefficient of 0.850. The industrial area, public facilities and military area show higher surface temperature (more than 37 °C) in comparison with adjoining areas, while the green spaces in urban areas (34 °C) and forests (29 °C) were the cooler part of the city. These successful results obtained in the study could be used as an efficient method for the environmental impact assessment.     

联合统计信息与散射模型的GF-5 AHSI可见光影像薄云校正

针对高分五号可见短波红外高光谱相机AHSI (visible-shortwave infrared Advanced HyperSpectral Imager)可见光波段存在的薄云干扰,本文提出了一种联合统计信息与散射模型的校正方法。利用AHSI影像邻近波段间地表与云雾辐射的统计差异,实现对不同场景下相对薄云辐射RCR (Relative Cloud Radiance)的准确估计。基于此,根据不同波段的散射特性,分别利用分级暗目标统计和散射模型约束策略,获取可见光波段的绝对云辐射强度,最终实现影像校正。通过设置模拟与真实实验对方法的有效性和鲁棒性进行目视和定量检验。模拟实验中,可见光波段内的薄云干扰均可被有效地去除,校正结果与真实地表十分一致;此外,RMSE (RootMean-Square Error),MAE (Mean Absolute Error)和SA (Spectral Angle) 3个评价指标的值分别为1.9891,1.6822和0.4901,远小于对比方法。真实实验中,不同场景内的薄云可被有效抑制,在较为准确恢复降质地表信息的同时保持晴空区光谱特征;Q指数,SSIM (Structural Similarity Index)和UQI (Universal Quality Index)的计算结果优于对比方法。综上,本文提出方法可用于不同场景下高分五号AHSI影像可见光波段的薄云校正,可得到目视效果良好且光谱保真度高的校正结果。     

Generation of Vegetation Fraction and Surface Albedo Products Over India from Ocean Colour Monitor (OCM) Data Onboard Oceansat-2

The use of Local Area Coverage (LAC) data from Ocean Color Monitor (OCM) sensor of Oceansat-2 with its high radiometric resolution (12 bits/pixel) and 2-day repeat cycle for rapid monitoring of vegetation growth and estimating surface albedo for the Indian region is demonstrated in this study. For the vegetation monitoring, normalized difference vegetation index (NDVI) and vegetation fraction (VF) products were estimated by maximum value composite approach fortnightly and were resampled to 1 km. The surface albedo products were realized by converting narrow-band eight-band spectral reflectance OCM data to a) visible (300–700 nm) and b) broad band (300–3,000 nm) data. For validation, the derived products were compared with respective MODIS global products and found to be in good agreement.     

海洋水色CCD成像仪光谱非线性校正

个理想的光学遥感器,在响应波段内光谱响应度是完全平直的,遥感器的输出与入射的辐射成完全的线性关系;而实际的遥感器并非如此,它的光谱响应度在响应波段内并不是平直的,这就使遥感器的输出不仅与入射的辐射有关,还与遥感器的光谱响应度有关,实际测量时,由于目标光谱形状的差异会导致输入与输出之间的非线性关系。利用仪器定标时采用的光源光谱形状与实际测量时目标光谱形状对仪器的输出数据进行校正会有效地减小由于遥感器的光谱响应非均匀性引入的测量误差。文中利用测得的海洋水色ＣＣＤ成像仪的各波段光谱响应度,计算了已归一化为具有相同的带内平均光谱辐亮度的不同温度黑体输出,由输出结果可以看到,不同的光谱形状对海洋水色ＣＣＤ成像仪的输出影响较大,为了减小由此引入的测量误差,文中对实验室定标用积分球光源和海洋水色ＣＣＤ成像仪在轨工作时的典型输入辐亮度的光谱特性进行了等效黑体拟合,并以此为依据计算了海洋水色ＣＣＤ成像仪在轨工作时的光谱非线性修正因子。     

珠海一号高光谱传感器波段平均太阳辐照度计算

传感器波段平均太阳辐照度是计算表观反射率等辐射参量的必要参数,是开展定量遥感分析的重要基础。通过中国气象局发布的零大气条件下太阳光谱辐照度数据和遥感数据运营商提供的珠海一号高光谱传感器光谱响应函数资料,计算了珠海一号三颗高光谱卫星上三组传感器的波段平均太阳辐照度。以中国资源卫星应用中心发布的高分一号资料为参考,确定结果具有较高的准确度,可为珠海一号高光谱数据的推广和使用提供参考。     

Vicarious Cross-Calibration of the China Environment Satellite Using Nearly Simultaneously Observations of Landsat-7 ETM+ Sensor

To understand the absolute radiometric calibration accuracy of the HJ-A CCD-1 sensors, image from these sensors were compared to nearly simultaneously image from Landsat-7 ETM+ sensors. Although the HJ-A CCD-1 sensor has almost the same wavelength of each central band and band width as Landsat-7 ETM+ sensor, there is slightly difference in spectral response function (SRF). The impacts of SRF difference effects would produce ~2 % uncertainty in predicting reflectance of HJ-A CCD-1 sensor using Landsat-7 ETM+ sensor. The reflectance observed by satellite at top-of-atmosphere generally depends on its’ geometric conditions. The results reveal that the impacts of geometrical conditions would impact on the vicarious cross-calibration accuracy, which should be removed. The performances of cross-calibration are calibrated and validated by four image pairs collected from Yellow River Delta, China, and Qingdao City, China, at four independent times. The results indicate that the HJ-A CCD-1 sensors can be cross calibrated to the Landsat-7 ETM+ sensors to within an accuracy of 3.99 % (denoted by Relative Root Mean Square Error) of each other in all bands except band 4, which has a 6.33 % difference.     

A satellite cross-calibration experiment

Recently, the Advanced Earth Observing Satellite 2 (ADEOS-2) was launched (December 14, 2002) successfully, and the Global Imager (GLI) onboard the ADEOS-2 satellite became operational in April 2003. In a first calibration checkup, the radiometric performance of GLI was compared relatively to that of other sensors on different satellites with different calibration backgrounds. As a calibration site, a large snowfield near Barrow, Alaska, was used, where space sensors in polar orbits view the same ground target on the same day with small differences in the local crossing times. This is why GLI, the Moderate Resolution Imaging Spectroradiometer (Terra, Aqua), the Sea-viewing Wide Field-of-view Sensor, the Advanced Very High Resolution Radiometer (N16, N17), the Medium Resolution Imaging Spectrometer, and the Advanced Along Track Scanning Radiometer datasets were selected for the following clear-sky condition days: April 14 and 26, 2003. At the same time, ground-truth experiments (e.g., measurements of ground reflectance, bidirectional reflectance distribution function, aerosol optical thickness) were carried out. Thereinafter, top-of-atmosphere (TOA) radiance/reflectance was forward calculated by means of radiative transfer code for each sensor, each band, and each day. Finally, the vicariously retrieved TOA signal was compared to TOA sensor Level 1B data. As a result, GLI's performance is encouraging at that time of the mission. GLI and the other seven sensors deliver similar sensor output in the range of about 5% to 7% around the expected vicariously calculated TOA signal.     

Assessing Metal-Induced Changes in the Visible and Near-Infrared Spectral Reflectance of Leaves: A Pot Study with Sunflower (<Emphasis Type="Italic">Helianthus annuus</Emphasis> L.)

The aim of this study was to monitor changes in leaf spectral reflectance due to phytoaccumulation of trace elements (Cd, Pb, and As) in sunflower mutant (M5 mutant line 38/R4-R6/15-35-190-04-M5) grown in spiked and in situ metal-contaminated potted soils. Reflectance spectra (350–2500 nm) of leaves were collected using portable ASD spectroradiometer, and respective leaves sample were analyzed for total metal contents. The spectral changes were quite noticeable and showed increased visible and decreased NIR reflectance for sunflower grown in soil spiked with 900 mg As kg^?1, and in in situ metal-contaminated soils. These changes also involved a blue-shift feature of red-edge position in the first derivatives spectra, studied vegetation indices and continuum removed absorption features at 495, 680, 970, 1165, 1435, 1780, and 1925 nm wavelength. Correlograms of leaf-metal concentration and reflectance values show highest degrees of overall correlation for visible, near-infrared, and water-sensitive wavelengths. Partial least square and multiple linear regression statistical models (cross-validated), respectively, based on Savitzky–Golay filter first-order derivative spectra and combination of spectral feature such as vegetation indices and band depths yielded good prediction of leaf-metal concentrations.     

Studies on Textural and Compositional Characteristics of Sand and Clay Mixtures Using Hyperspectral Radiometry

This paper examines the hyperspectral signatures (in the Visible Near Infrared (VNIR)-Shortwave Infrared (SWIR) regions) of soil samples with varying colour and minerals. 36 samples of sands (from river and beach) with differing clay contents were examined using a hyperspectral radiometer operating in the 350–2,500 nm range, and the spectral curves were obtained. Analysis of the spectra indicates that there is an overall increase in the reflectance in the VNIR-SWIR region with an increase in the content of kaolinite clay in the sand samples. As regards the red and black clays and sand mixtures, the overall reflectance increases with decreasing clay content. Several spectral parameters such as depth of absorption at 1,400 nm and 1,900 nm regions, radius of curvature of the absorption troughs, slope at a particular wavelength region and the peak reflectance values were derived. There exists a correlation between certain of these spectral parameters (depth, slope, position, peak reflectance, area under the curve and radius of the curve) and the compositional and textural parameters of the soils. Based on these well-defined relations, it is inferred that hyperspectral radiometry in the VNIR and SWIR regions can be used to identify the type of clay and estimate the clay content in a given soil and thus define its geotechnical category.     

Short-wavelength Spectral Properties of the Gravity Field from a Range of Regional Data Sets

The GRACE (gravity recovery and climate experiment) and GOCE (gravity field and steady-state ocean circulation explorer) dedicated gravity satellite missions are expected to deliver the long-wavelength scales of the Earth’s gravity field with extreme precision. For many applications in Earth sciences, future research activities will have to focus on a similar precision on shorter scales not recovered by satellite missions. Here, we investigate the signal power of gravity anomalies at such short scales. We derive an average degree variance and power spectral density model for topography-reduced gravity anomalies (residual terrain model anomalies and de-trended refined Bouguer anomalies), which is valid for wavelengths between 0.7 and 100  km. The model is based on the analysis of gravity anomalies from 13 test regions in various geographical areas and geophysical settings, using various power spectrum computation approaches. The power of the derived average topography-reduced model is considerably lower than the Tscherning–Rapp free air anomaly model. The signal power of the individual test regions deviates from the obtained average model by less than a factor of 4 in terms of square-root power spectral amplitudes. Despite the topographic reduction, the highest signal power is found in mountainous areas and the lowest signal power in flat terrain. For the derived average power spectral model, a validation procedure is developed based on least-squares prediction tests. The validation shows that the model leads to a good prediction quality and realistic error measures. Therefore, for least-squares prediction, the model could replace the use of autocovariance functions derived from local or regional data.