Spectral Mixture Analysis Techniques Research Articles

Evaluation of EnMAP hyperspectral satellite data for epithermal alteration mapping at Cuprite-Goldfield, southwestern Nevada, USA

The Environmental Mapping and Analysis Program (EnMAP) is a German hyperspectral satellite mission launched into orbit on April 1, 2022. The EnMAP records 246 hyperspectral channels in the 0.42-2.45 μm wavelength range, with a spatial resolution of 30 m. This wavelength range is useful for environmental and geological mapping. This study evaluates the application of the EnMAP data to the mapping of epithermal alteration systems at Cuprite and Goldfield in southwestern Nevada, USA. The Cuprite has served since 1980s as a test site of the U.S. National Aeronautics and Space Administration (NASA) for the evaluation of airborne and satellite remote sensing sensors. The surface reflectance Cuprite-Goldfield EnMAP scene analyzed in this study was recorded on April 9, 2023. The EnMAP data were analyzed using the spectral mixture analysis technique. The results obtained from the EnMAP moderate spatial resolution satellite hyperspectral data show accurate detection of the alteration zones in the Cuprite-Goldfield study area. The study indicates the potential of the EnMAP data for various geological and environmental applications.

Applying hyperspectral remote sensing methods to ship detection based on airborne and ground experiments

ABSTRACT Recent advances in hyperspectral remote sensing using hundreds of spectral channels make it possible to derive oceanic variables and to understand the spatial distribution of diverse oceanic phenomena. In this study, hyperspectral measurements were obtained using airborne and ground experiments and hyperspectral data processing techniques were applied for the detection of vessels. The observed hyperspectral data were analysed by using the spectral characteristics of endmembers and a representative spectral mixture analysis technique. The pixels in the image were classified into endmembers by using spectral matching algorithms and the marine library spectra from preliminary experiments. The estimated lengths and widths of the detected ships based on a series of hyperspectral data processing methods show root mean square error (RMSE) of 2.28% and 8.72% with respect to actual ship sizes, respectively. The investigation of the errors revealed that undulating surface waves, surface water penetration of the deck, and the submerged portion of the ship below the water surface potentially contributes to unclear ship boundaries. In this study, an optimal method for the objective determination of the threshold of abundance fractions based on pixels with multiple endmember contributions is proposed. The validation results based on the proposed hyperspectral method are in good agreement with the actual sizes of the ships.

Spectral Unmixing Using a Sparse Multiple-Endmember Spectral Mixture Model

In remote sensing data exploitation, the spectral mixture analysis technique is generally used to detect the land cover materials and their corresponding proportions present in the observed scene. Traditionally, a fixed endmember spectral signature for each land cover material is used to perform the unmixing task. In the literature, some scholars have proposed performing the unmixing by taking the spectral variability into consideration. Among these spectral-variability-based unmixing approaches, multiple-endmember spectral mixture analysis (MESMA) is probably the most widely used method. However, when the number of land cover materials is large, the computational load of the MESMA method could be very heavy. In this paper, a sparse multiple-endmember spectral mixture model (SMESMM) is proposed to handle this problem. This model treats the spectral mixture procedure as a linear block sparse inverse problem. The SMESMM is first solved using a block sparse algorithm to obtain an initial block sparse solution. Then, MESMA is used to resolve the mixed pixel using the selected land cover materials, which correspond to the nonzero blocks in the solution obtained in the first step. The block sparse solution obtained in the first step can help to determine how many and which land cover materials are involved in the considered mixed pixel. This can largely decrease the number of possible candidate models for the MESMA method when the number of land cover materials is large. Experimental results on simulated and real hyperspectral data demonstrate the efficacy of the proposed method.

Hyperspectral Unmixing With Endmember Variability via Alternating Angle Minimization

In hyperspectral unmixing applications, one typically assumes that a single spectrum exists for every endmember. In many scenarios, this is not the case, and one requires a set or a distribution of spectra to represent an endmember or class. This inherent spectral variability can pose severe difficulties in classical unmixing approaches. In this paper, we present a new algorithm for dealing with endmember variability in spectral unmixing, based on the geometrical interpretation of the resulting unmixing problem, and an alternating optimization approach. This alternating-angle-minimization algorithm uses sets of spectra to represent the variability present in each class and attempts to identify the subset of endmembers which produce the smallest reconstruction error. The algorithm is analogous to the popular multiple endmember spectral mixture analysis technique but has a much more favorable computational complexity while producing similar results. We illustrate the algorithm on several artificial and real data sets and compare with several other recent techniques for dealing with endmember variability.

Automated Subpixel Surface Water Mapping from Heterogeneous Urban Environments Using Landsat 8 OLI Imagery

Water bodies are a fundamental element of urban ecosystems, and water mapping is critical for urban and landscape planning and management. Remote sensing has increasingly been used for water mapping in rural areas; however, when applied to urban areas, this spatially- explicit approach is a challenging task due to the fact that the water bodies are often of a small size and spectral confusion is common between water and the complex features in the urban environment. Water indexes are the most common method of water extraction at the pixel level. More recently, spectral mixture analysis (SMA) has been widely employed in analyzing the urban environment at the subpixel level. The objective of this study is to develop an automatic subpixel water mapping method (ASWM) which can achieve a high accuracy in urban areas. Specifically, we first apply a water index for the automatic extraction of mixed land-water pixels, and the pure water pixels that are generated in this process are exported as the final result. Secondly, the SMA technique is applied to the mixed land-water pixels for water abundance estimation. As for obtaining the most representative endmembers, we propose an adaptive iterative endmember selection method based on the spatial similarity of adjacent ground surfaces. One classical water index method (the modified normalized difference water index (MNDWI)), a pixel-level target detection method (constrained energy minimization (CEM)), and two widely used SMA methods (fully constrained least squares (FCLS) and multiple endmember spectral mixture analysis (MESMA)) were chosen for the water mapping comparison in the experiments. The results indicate that the proposed ASWM was able to detect water pixels more efficiency than other unsupervised water extraction methods, and the water fractions estimated by the proposed ASWM method correspond closely to the reference fractions with the slopes of 0.97, 1.02, 1.04, and 0.98 and the R-squared values of 0.9454, 0.9486, 0.9665, and 0.9607 in regression analysis corresponding to different test regions. In the quantitative accuracy assessment, the ASWM method shows the best performance in water mapping with the mean kappa coefficient of 0.862, mean producer’s accuracy of 82.8%, and mean user’s accuracy of 91.8% for test regions.

Multiscale Water Body Extraction in Urban Environments From Satellite Images

Water is a fundamental element in urban environments, and water body extraction is important for landscape and urban planning. Remote sensing has increasingly been used for water body extraction; however, in urban environments, this kind of approaches is challenging because of the significant within-class spectral variance in water areas and the presence of complex ground features. The objective of this study is to develop an automatic method that could improve water body extraction in urban environments from moderate spatial resolution satellite images. Central to our method is the combined use of multiscale extractions and spectral mixture analysis techniques in adaptive local regions. Specifically, we first calculate the NDWI image from experimental images for selecting water sample pixels. Second, on the basis of the selected water pixels, we apply an improved spectral mixture analysis technique on the experimental image to get water abundance of every pixel, and segment the abundance image to extract water bodies at the global scale. Third, in a similar manner, we iteratively conduct the water body extraction in multiscale local regions to refine the water bodies. Finally, the final result of water bodies is obtained when a stopping criterion is satisfied. We have implemented this method to produce water maps from an ALOS/AVNIR-2 image and a Terra/ASTER image covering urban areas. The experimental results illustrate that the proposed method has substantially outperformed two related methods that use the NDWI-based thresholding and the SVM classification for the entire image.

Analyzing the effect of shadow on the relationship between ground cover and vegetation indices by using spectral mixture and radiative transfer models

We present and evaluate an experimental relationship between the fraction of ground cover (FV) and multispectral vegetation indices (VI) derived from medium resolution images (Landsat 5-TM) in vertical shoot trellised vineyards. The results indicate a strong linear relationship between FV and the Normalized Difference Vegetation Index (NDVI) and Soil Adjusted Vegetation Index (SAVI), resulting in correlation coefficients greater than 0.90. These relationships were evaluated for the effect of variations in illumination angles and shadow enlargement using two analytical approaches: the Linear Spectral Mixture Analysis techniques and a radiative transfer approach with the Markov-chain canopy reflectance model, with additions to simulate the row structure. Previous to this analysis, both models were evaluated by comparing the model results with VIs in row vineyards obtained from satellite images, performing fairly well. The exploratory analysis demonstrated that the use of a single relationship based on the NDVI index could result in significant inaccuracies for larger zenith angles and row directions perpendicular to the sun azimuth at the satellite acquisition time. In contrast, the SAVI improved the linearity of the relationships and resulted in less sensitivity to changes in the sun angles and row directions.

A spatially adaptive spectral mixture analysis for mapping subpixel urban impervious surface distribution

Spectral mixture analysis (SMA) has been widely employed in analyzing urban environments, especially for estimating urban impervious surface distribution at the subpixel level. When implementing SMA, endmember selection is considered an important step, and an inappropriate endmember set could severely affect the accuracy of fractional land covers. Although different degrees of success have been achieved in existing SMA approaches, the paradox of endmember selection is still unsolved: theoretically “purest” endmembers that can be selected with relative ease do not always yield optimal results, while the selection of the most “representative” endmembers is very difficult with simple SMA. To address this problem, we propose a spatially adaptive SMA (SASMA) technique to automatically extract and synthesize the “most representative” endmembers for SMA through considering both between-class and within-class variations. In particular, we developed a classification tree method to automatically extract endmember candidates through incorporating spectral and spatial information. In addition, to mitigate the effect of within-class variation, we employed synthetic spectra of neighboring endmember candidates in a local search window as the most “representative” endmember signature. Results indicate that SASMA performs well in estimating subpixel impervious surface distribution with relatively high precision (mean absolute error of 8.50%, root mean square error of 15.25%, and R2 of 0.701) and small bias (systematic error of −0.93%).

Magnitude- and Shape-Related Feature Integration in Hyperspectral Mixture Analysis to Monitor Weeds in Citrus Orchards

Traditionally, spectral mixture analysis (SMA) fails to fully account for highly similar ground components or endmembers. The high similarity between weed and crop spectra hampers the implementation of SMA for steering weed control management practices. To address this problem, this paper presents an alternative SMA technique, referred to as Integrated Spectral Unmixing (InSU). InSU combines both magnitude (i.e., reflectance) and shape (i.e., derivative reflectance) related features in an automated waveband selection protocol. Analysis was performed on different simulated mixed pixel spectra sets compiled from in situ-measured weed canopy, Citrus canopy, and soil spectra. Compared to traditional linear SMA, InSU significantly improved weed cover fraction estimations. An average decrease in fraction abundance error (Deltaf) of 0.09 was demonstrated for a signal-to-noise ratio (SNR) of 500 : 1, while for a SNR of 50 : 1, the decrease was 0.06.

The derivation of an Arctic sea ice normal through temporal mixture analysis of satellite imagery

The use of temporal mixture analysis (TMA) for creating a long-term baseline, or environmental “normal” is described. TMA is a promising analysis method derived from the hyperspectral image processing technique of spectral mixture analysis (SMA). TMA is algebraically identical to SMA, except that it is applied to temporal spectra rather than to electromagnetic spectra. TMA has particular potential to extract climate signals from long image sequences.To demonstrate the utility of TMA, this paper documents its use to isolate nine fundamental temporal signals (“endmembers”) from a 20-year Northern Hemisphere sea ice concentration image time series. The temporal endmembers establish a baseline of temporal variability that can be treated as an environmental normal. The “fraction images” produced by the analysis highlight the regions where the temporal signals are strongest and provide new insights into the dynamics of the Arctic sea ice cover.

Impact of pixel size on mapping surface water in subsolar imagery

We observed surface water in a wetland, imaging in the subsolar or specular direction the exceptionally bright specular reflection of sunlight at a ground resolution of 0.3 m. We then simulated ground resolutions between 1.7 m and 1.2 km through aggregation of the 0.3 m pixels. Contrary to the expectations of some of our colleagues in the wetlands community, for these data, the accuracy of spectral mixture analysis (SMA) estimates of surface water increases as pixel ground footprint size increases. Our results suggest that regional to global scale assessments of flooded landscapes and wetlands that do not involve issues requiring 1 m resolution per se may be addressed with acceptable accuracy by applying SMA techniques to low resolution imagery. Our results indicate within-pixel estimates of surface water area derived from data measured by subsolar viewing sensors with large ground pixel footprints, such as satellite POLarization and Directionality of Earth Radiance (POLDER) data, may be highly accurate under strong surface wind conditions.

Remote sensing of spatial patterns of urban renewal using linear spectral mixture analysis: A case of central urban area of Shanghai (1997–2000)

It is very important to integrate remote sensing with urban geography that the spectral mixture analysis technique is applied to urban land cover evolvement and its eco-environmental effect. Urban land cover is mainly composed of complicated artificial materials, which is the key factor to limit the development of the spectral mixture analysis technique. There are two main aspects in which the technique of spectral mixture analysis is applied to urban geography: one is to calculate vegetation fraction; the other is to build quantitative model of the urban impervious surface obtained from the combination between high albedo fraction and low albedo fraction. The technique of spectral mixture analysis is firstly applied to study urban renewal pattern, scale and mode which happened in Shanghai City from 1997 to 2000.

Evaluation of soil erosion protective cover by crop residues using vegetation indices and spectral mixture analysis of multispectral and hyperspectral data

Crop residues are efficient in reducing erosion and surface water runoff on agricultural soils. Evaluating the crop residue cover fraction and its spatial distribution is important to scientists involved in the modelling of soil erosion and surface runoff, and also to authorities wishing to assess soil conservation adoption by farmers. This study focuses on the evaluation of four remote sensing techniques to estimate the cover fraction of cereal crop residues (i.e., wheat and corn) from multispectral and hyperspectral measurements. These are the Soil Adjusted Corn Residue Index (SACRI), the Crop Residue Index Multiband (CRIM), the Normalized Difference Index (NDI) and the spectral mixture analysis technique (SMA). Field campaigns that were carried out by the FLOODGEN project in Sainte-Angèle-de-Monnoir, Québec, Canada and in the Pays-de-Caux located in the Normandy region of France, allowed us to gather digital photographs, spectra and other measurements to determine the actual ground cover fraction. A linear regression analysis between results derived from Landsat-5 TM simulated field spectra and the actual ground cover fractions showed best results for the CRIM on the Ste-Angèle-de-Monnoir study site ( R 2 = 0.96), and equally good results for the Pays-de-Caux study site ( R 2 = 0.94). Results were not as good when SMA was applied to the same Landsat-5 TM simulated field spectra with R 2 values of 0.70 and 0.68 for both sites, respectively. However, results improved significantly when SMA was applied to the hyperspectral data in which case the R 2 values increased to 0.92 for the Sainte-Angèle-de-Monnoir site and 0.89 for the Pays-de-Caux study site. Results obtained with the NDI and SACRI from both simulated TM and hyperspectral field spectra were not conclusive.

Practical limits on hyperspectral vegetation discrimination in arid and semiarid environments

Hyperspectral remote sensing is a promising tool for the analysis of vegetation and soils in remote sensing imagery. The purpose of this study is to ascertain how well hyperspectral remote sensing data can retrieve vegetation cover, vegetation type, and soil type in areas of low vegetation cover. We use multiple endmember spectral mixture analysis (MESMA), high-quality field spectra, and AVIRIS data to determine how well full-range spectral mixture analysis (SMA) techniques can retrieve vegetation and soil information. Using simulated AVIRIS-derived reflectance spectra, we find that, in areas of low vegetation cover, MESMA is not able to provide reliable retrievals of vegetation type when covers are less than at least 30%. Overestimations of vegetation are likely, but vegetation cover in many circumstances can be estimated reliably. Soil type retrievals are more than 90% reliable in discriminating dark-armored desert soils from blown sands. This simulation comprises a best-case scenario in which many typical problems with remote sensing in areas of low cover or desert areas are minimized. Our results have broad implications for the applicability of full-range SMA techniques in analysis of data from current and planned hyperspectral sensors. Several phenomena contribute to the unreliability of vegetation retrievals. Spectrally indeterminate vegetation types, characterized by low spectral contrast, are difficult to model correctly even at relatively high covers. Combinations of soil and vegetation spectra have the potential of generating mixtures that resemble an unmixed spectrum from different material, further confounding vegetation cover and soil type retrievals. Intraspecies spectral variability and nonlinear mixing produce uncertainties in spectral endmembers much larger than that only due to instrumental noise modeled here. Having established limits on linear spectral unmixing in areas of low cover through spectral simulations, we evaluate AVIRIS-derived reflectance data from the Mojave Desert, California. We show that MESMA is capable of mapping soil surface types even when vegetation type cannot be reasonable retrieved.