Search NASA⌕ Search

SEARCH · Search NASA

Results for “image based modeling”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,747 records · Page 97

Physics-based reward driven image analysis in microscopy

The rise of electron microscopy has expanded our ability to acquire nanometer and atomically resolved images of complex materials. The resulting vast datasets are typically analyzed by human operators, an intrinsically challenging process due to the multiple possible analysis steps and the corresponding need to build and optimize complex analysis workflows. We present a methodology based on the concept of a Reward Function coupled with Bayesian Optimization, to optimize image analysis workflows dynamically. The Reward Function is engineered to closely align with the experimental objectives and broader context and is quantifiable upon completion of the analysis. Here, cross-section, high-angle annular dark field (HAADF) images of ion-irradiated (Y, Dy)Ba 2 Cu 3 O 7–δ thin-films were used as a model system. The reward functions were formed based on the expected materials density and atomic spacings and used to drive multi-objective optimization of the classical Laplacian-of-Gaussian (LoG) method. These results can be benchmarked against the DCNN segmentation. This optimized LoG* compares favorably against DCNN in the presence of the additional noise. We further extend the reward function approach towards the identification of partially-disordered regions, creating a physics-driven reward function and action space of high-dimensional clustering. We pose that with correct definition, the reward function approach allows real-time optimization of complex analysis workflows at much higher speeds and lower computational costs than classical DCNN-based inference, ensuring the attainment of results that are both precise and aligned with the human-defined objectives.

47 OTHER INSTRUMENTATION↗

Source-encoded waveform inversion in the Northern Hemisphere

SUMMARY We use source-encoded waveform inversion to image Earth’s Northern Hemisphere. The encoding method is based on measurements of Laplace coefficients of stationary wavefields. By assigning to each event a unique frequency, we compute Fréchet derivatives for all events simultaneously based on one ‘super’ forward and one ‘super’ adjoint simulation for a small fraction of the computational cost of classical waveform inversion with the same data set. No cross-talk noise is introduced in the process, and the method does not require all events to be recorded by all stations. Starting from global model GLAD_M25, we performed 100 conjugate gradient iterations using a data set consisting of 786 earthquakes recorded by 9846 stations. Synthetic inversion tests show that we achieve good convergence based on this data set, and we see a consistent misfit reduction during the inversion. The new model, named SE100, has much higher spatial resolution than GLAD_M25, revealing details of the Yellowstone and Iceland hotspots, subduction beneath the Western United States and the upper mantle structure beneath the Arctic Ocean.

Cui, Congyue↗

Mapping the dark matter halo of early-type galaxy NGC 2974 through orbit-based models with combined stellar and cold gas kinematics

ABSTRACT We present an orbit-based method of combining stellar and cold gas kinematics to constrain the dark matter profile of early-type galaxies. We apply this method to early-type galaxy NGC 2974, using Pan-STARRS imaging and SAURON stellar kinematics to model the stellar orbits, and introducing H i kinematics from VLA observation as a tracer of the gravitational potential. The introduction of the cold gas kinematics shows a significant effect on the confidence limits of especially the dark halo properties: we exclude more than $95{{\ \rm per\ cent}}$ of models within the 1σ confidence level of Schwarzschild modelling with only stellar kinematics, and reduce the relative uncertainty of the dark matter fraction significantly to $10{{\ \rm per\ cent}}$ within 5Re. Adopting a generalized Navarro–Frenk–White (NFW) dark matter profile, we measure a shallow cuspy inner slope of $0.6^{+0.2}_{-0.3}$ when including the cold gas kinematics in our model. We cannot constrain the inner slope with the stellar kinematics alone.

Yang, Meng↗

The Atomic Structural Dynamics of γ-Al 2 O 3 Supported Ir–Pt Nanocluster Catalysts Prepared from a Bimetallic Molecular Precursor: A Study Using Aberration-Corrected Electron Microscopy and X-ray Absorption Spectroscopy

This study describes a prototypical, bimetallic heterogeneous catalyst: compositionally well-defined Ir–Pt nanoclusters with sizes in the range of 1–2 nm supported on γ-Al 2 O 3 . Deposition of the molecular bimetallic cluster [Ir 3 Pt 3 (μ-CO) 3 (CO) 3 (η-C 5 Me 5 ) 3 ] on γ-Al 2 O 3 , and its subsequent reduction with hydrogen, provides highly dispersed supported bimetallic Ir–Pt nanoparticles. Using spherical aberration-corrected scanning transmission electron microscopy (Cs-STEM) and theoretical modeling of synchrotron-based X-ray absorption spectroscopy (XAS) measurements, our studies provide unambiguous structural assignments for this model catalytic system. The atomic resolution Cs-STEM images reveal strong and specific lattice-directed strains in the clusters that follow local bonding configurations of the γ-Al 2 O 3 support. Combined nanobeam diffraction (NBD) and high-resolution transmission electron microscopy (HRTEM) data suggest the polycrystalline γ-Al 2 O 3 support material predominantly exposes (001) and (011) surface planes (ones commensurate with the zone axis orientations frequently exhibited by the bimetallic clusters). Here, the data reveal that the supported bimetallic clusters exhibit complex patterns of structural dynamics, ones evidencing perturbations of an underlying oblate/hemispherical cuboctahedral cluster–core geometry with cores that are enriched in Ir (a result consistent with models based on surface energetics, which favor an ambient cluster termination by Pt) due to the dynamical responses of the M–M bonding to the specifics of the adsorbate and metal–support interactions. Taken together, the data demonstrate that strong temperature-dependent charge-transfer effects occur that are likely mediated variably by the cluster–support, cluster–adsorbate, and intermetallic bonding interactions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

TH-E-BRF-04: Characterizing the Response of Texture-Based CT Image Features for Quantification of Radiation-Induced Normal Lung Damage

Purpose: Radiation induced lung damage (RILD) is an important dose-limiting toxicity for patients treated with radiation therapy. Scoring systems for RILD are subjective and limit our ability to find robust predictors of toxicity. We investigate the dose and time-related response for texture-based lung CT image features that serve as potential quantitative measures of RILD. Methods: Pre- and post-RT diagnostic imaging studies were collected for retrospective analysis of 21 patients treated with photon or proton radiotherapy for NSCLC. Total lung and selected isodose contours (0–5, 5–15, 15–25Gy, etc.) were deformably registered from the treatment planning scan to the pre-RT and available follow-up CT studies for each patient. A CT image analysis framework was utilized to extract 3698 unique texture-based features (including co-occurrence and run length matrices) for each region of interest defined by the isodose contours and the total lung volume. Linear mixed models were fit to determine the relationship between feature change (relative to pre-RT), planned dose and time post-RT. Results: Seventy-three follow-up CT scans from 21 patients (median: 3 scans/patient) were analyzed to describe CT image feature change. At the p=0.05 level, dose affected feature change in 2706 (73.1%) of the available features. Similarly, time affected feature change in 408 (11.0%) of the available features. Both dose and time were significant predictors of feature change in a total of 231 (6.2%) of the extracted image features. Conclusion: Characterizing the dose and time-related response of a large number of texture-based CT image features is the first step toward identifying objective measures of lung toxicity necessary for assessment and prediction of RILD. There is evidence that numerous features are sensitive to both the radiation dose and time after RT. Beyond characterizing feature response, further investigation is warranted to determine the utility of these features as surrogates of clinically significant lung injury.

60 APPLIED LIFE SCIENCES↗

Investigation of the role of thermal boundary layer processes in initiating convection under the NASA SPACE Field Program

The current NWS ground based network is not sufficient to capture the dynamic or thermodynamic structure leading to the initiation and organization of air mass moist convective events. Under this investigation we intend to use boundary layer mesoscale models (McNider and Pielke, 1981) to examine the dynamic triggering of convection due to topography and surface thermal contrasts. VAS and MAN's estimates of moisture will be coupled with the dynamic solution to provide an estimate of the total convective potential. Visible GOES images will be used to specify incoming insolation which may lead to surface thermal contrasts and JR skin temperatures will be used to estimate surface moisture (via the surface thermal inertia) (Weizel and Chang, 1988) which can also induce surface thermal contrasts. We will use the SPACE-COHMEX data base to evaluate the ability of the joint mesoscale model satellite products to show skill in predicting the development of air mass convection. We will develop images of model vertical velocity and satellite thermodynamic measures to derive images of predicted convective potential. We will then after suitable geographic registration carry out a pixel by pixel correlation between the model/satellite convective potential and the 'truth' which are the visible images. During the first half of the first year of this investigation we have concentrated on two aspects of the project. The first has been in generating vertical velocity fields from the model for COHMEX case days. We have taken June 19 as the first case and have run the mesoscale model at several different grid resolutions. We are currently developing the composite model/satellite convective image. The second aspect has been the attempted calibration of the surface energy budget to provide the proper horizontal thermal contrasts for convective initiation. We have made extensive progress on this aspect using the FIFE data as a test data set. The calibration technique looks very promising.

Mcnider, Richard T.↗

Gaussian Mixture Models for Temporal Depth Fusion

Sensing the 3D environment of a moving robot is essential for collision avoidance. Most 3D sensors produce dense depth maps, which are subject to imperfections due to various environmental factors. Temporal fusion of depth maps is crucial to overcome those. Temporal fusion is traditionally done in 3D space with voxel data structures, but it can be approached by temporal fusion in image space, with potential benefits in reduced memory and computational cost for applications like reactive collision avoidance for micro air vehicles. In this paper, we present an efficient Gaussian Mixture Models based depth map fusion approach, introducing an online update scheme for dense representations. The environment is modeled from an ego-centric point of view, where each pixel is represented by a mixture of Gaussian inverse-depth models. Consecutive frames are related to each other by transformations obtained from visual odometry. This approach achieves better accuracy than alternative image space depth map fusion techniques at lower computational cost.

Matthies, Larry↗

CNR considerations for rapid real-time MRI tumor tracking in radiotherapy hybrid devices: Effects of B{sub 0} field strength

Purpose: This work examines the subject of contrast-to-noise ratio (CNR), specifically between tumor and tissue background, and its dependence on the MRI field strength, B{sub 0}. This examination is motivated by the recent interest and developments in MRI/radiotherapy hybrids where real-time imaging can be used to guide treatment beams. The ability to distinguish a tumor from background tissue is of primary importance in this field, and this work seeks to elucidate the complex relationship between the CNR and B{sub 0} that is too often assumed to be purely linear. Methods: Experimentally based models of B{sub 0}-dependant relaxation for various tumor and normal tissues from the literature were used in conjunction with signal equations for MR sequences suitable for rapid real-time imaging to develop field-dependent predictions for CNR. These CNR models were developed for liver, lung, breast, glioma, and kidney tumors for spoiled gradient-echo, balanced steady-state free precession (bSSFP), and single-shot half-Fourier fast spin echo sequences. Results: Due to the pattern in which the relaxation properties of tissues are found to vary over B{sub 0} field (specifically the T{sub 1} time), there was always an improved CNR at lower fields compared to linear dependency. Further, in some tumor sites, the CNR at lower fields was found to be comparable to, or sometimes higher than those at higher fields (i.e., bSSFP CNR for glioma, kidney, and liver tumors). Conclusions: In terms of CNR, lower B{sub 0} fields have been shown to perform as well or better than higher fields for some tumor sites due to superior T{sub 1} contrast. In other sites this effect was less pronounced, reversing the CNR advantage. This complex relationship between CNR and B{sub 0} reveals both low and high magnetic fields as viable options for tumor tracking in MRI/radiotherapy hybrids.

60 APPLIED LIFE SCIENCES↗

Domain-Adapted Convolutional Networks for Satellite Image Classification: A Large-Scale Interactive Learning Workflow

Satellite imagery often exhibits large spatial extent areas that encompass object classes with considerable variability. This often limits large-scale model generalization with machine learning algorithms. Notably, acquisition conditions, including dates, sensor position, lighting condition, and sensor types, often translate into class distribution shifts introducing complex nonlinear factors and hamper the potential impact of machine learning classifiers. Here, this article investigates the challenge of exploiting satellite images using convolutional neural networks (CNN) for settlement classification where the class distribution shifts are significant. We present a large-scale human settlement mapping workflow based-off multiple modules to adapt a pretrained CNN to address the negative impact of distribution shift on classification performance. To extend a locally trained classifier onto large spatial extents areas we introduce several submodules: First, a human-in-the-loop element for relabeling of misclassified target domain samples to generate representative examples for model adaptation; second, an efficient hashing module to minimize redundancy and noisy samples from the mass-selected examples; and third, a novel relevance ranking module to minimize the dominance of source example on the target domain. The workflow presents a novel and practical approach to achieve large-scale domain adaptation with binary classifiers that are based-off CNN features. Experimental evaluations are conducted on areas of interest that encompass various image characteristics, including multisensors, multitemporal, and multiangular conditions. Domain adaptation is assessed on source–target pairs through the transfer loss and transfer ratio metrics to illustrate the utility of the workflow.

47 OTHER INSTRUMENTATION↗

SU-G-JeP3-06: Lower KV Image Dose Are Expected From a Limited-Angle Intra-Fractional Verification (LIVE) System for SBRT Treatments

Purpose: In order to track the tumor movement for patient positioning verification during arc treatment delivery or in between 3D/IMRT beams for stereotactic body radiation therapy (SBRT), the limited-angle kV projections acquisition simultaneously during arc treatment delivery or in-between static treatment beams as the gantry moves to the next beam angle was proposed. The purpose of this study is to estimate additional imaging dose resulting from multiple tomosynthesis acquisitions in-between static treatment beams and to compare with that of a conventional kV-CBCT acquisition. Methods: kV imaging system integrated into Varian TrueBeam accelerators was modeled using EGSnrc Monte Carlo user code, BEAMnrc and DOSXYZnrc code was used in dose calculations. The simulated realistic kV beams from the Varian TrueBeam OBI 1.5 system were used to calculate dose to patient based on CT images. Organ doses were analyzed using DVHs. The imaging dose to patient resulting from realistic multiple tomosynthesis acquisitions with each 25–30 degree kV source rotation between 6 treatment beam gantry angles was studied. Results: For a typical lung SBRT treatment delivery much lower (20–50%) kV imaging doses from the sum of realistic six tomosynthesis acquisitions with each 25–30 degree x-ray source rotation between six treatment beam gantry angles were observed compared to that from a single CBCT image acquisition. Conclusion: This work indicates that the kV imaging in this proposed Limited-angle Intra-fractional Verification (LIVE) System for SBRT Treatments has a negligible imaging dose increase. It is worth to note that the MV imaging dose caused by MV projection acquisition in-between static beams in LIVE can be minimized by restricting the imaging to the target region and reducing the number of projections acquired. For arc treatments, MV imaging acquisition in LIVE does not add additional imaging dose as the MV images are acquired from treatment beams directly during the treatment.

60 APPLIED LIFE SCIENCES↗

THE SIZE EVOLUTION OF PASSIVE GALAXIES: OBSERVATIONS FROM THE WIDE-FIELD CAMERA 3 EARLY RELEASE SCIENCE PROGRAM

We present the size evolution of passively evolving galaxies at z {approx} 2 identified in Wide-Field Camera 3 imaging from the Early Release Science program. Our sample was constructed using an analog to the passive BzK galaxy selection criterion, which isolates galaxies with little or no ongoing star formation at z {approx}> 1.5. We identify 30 galaxies in {approx}40 arcmin{sup 2} to H < 25 mag. By fitting the 10-band Hubble Space Telescope photometry from 0.22 {mu}m {approx}< {lambda}{sub obs} {approx}< 1.6 {mu}m with stellar population synthesis models, we simultaneously determine photometric redshift, stellar mass, and a bevy of other population parameters. Based on the six galaxies with published spectroscopic redshifts, we estimate a typical redshift uncertainty of {approx}0.033(1 + z). We determine effective radii from Sersic profile fits to the H-band image using an empirical point-spread function. By supplementing our data with published samples, we propose a mass-dependent size evolution model for passively evolving galaxies, where the most massive galaxies (M{sub *} {approx} 10{sup 11} M{sub Sun }) undergo the strongest evolution from z {approx} 2 to the present. Parameterizing the size evolution as (1 + z){sup -{alpha}}, we find a tentative scaling of {alpha} Almost-Equal-To (- 0.6 {+-} 0.7) + (0.9 {+-} 0.4)log (M{sub *}/10{sup 9} M{sub Sun }), where the relatively large uncertainties reflect the poor sampling in stellar mass due to the low numbers of high-redshift systems. We discuss the implications of this result for the redshift evolution of the M{sub *}-R{sub e} relation for red galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗

MO-FG-202-06: Improving the Performance of Gamma Analysis QA with Radiomics- Based Image Analysis

Purpose: The use of gamma analysis for IMRT quality assurance has well-known limitations. Traditionally, a simple thresholding technique is used to evaluated passing criteria. However, like any image the gamma distribution is rich in information which thresholding mostly discards. We therefore propose a novel method of analyzing gamma images that uses quantitative image features borrowed from radiomics, with the goal of improving error detection. Methods: 368 gamma images were generated from 184 clinical IMRT beams. For each beam the dose to a phantom was measured with EPID dosimetry and compared to the TPS dose calculated with and without normally distributed (2mm sigma) errors in MLC positions. The magnitude of 17 intensity histogram and size-zone radiomic features were derived from each image. The features that differed most significantly between image sets were determined with ROC analysis. A linear machine-learning model was trained on these features to classify images as with or without errors on 180 gamma images.The model was then applied to an independent validation set of 188 additional gamma distributions, half with and half without errors. Results: The most significant features for detecting errors were histogram kurtosis (p=0.007) and three size-zone metrics (p<1e-6 for each). The sizezone metrics detected clusters of high gamma-value pixels under mispositioned MLCs. The model applied to the validation set had an AUC of 0.8, compared to 0.56 for traditional gamma analysis with the decision threshold restricted to 98% or less. Conclusion: A radiomics-based image analysis method was developed that is more effective in detecting error than traditional gamma analysis. Though the pilot study here considers only MLC position errors, radiomics-based methods for other error types are being developed, which may provide better error detection and useful information on the source of detected errors. This work was partially supported by a grant from the Agency for Healthcare Research and Quality, grant number R18 HS022244-01.

60 APPLIED LIFE SCIENCES↗

Shock Formation of Slow Magnetosonic Waves in Coronal Plumes

We investigate the height of shock formation in coronal plumes for slow magnetosonic waves. The models take into account plume geometric spreading, heat conduction, and radiative damping. The wave parameters as well as the spreading functions of the plumes and the base magnetic field strengths are given by empirical constraints mostly from Solar and Heliospheric Observatory (SOHO)/ Ultraviolet Coronograph Spectrometer (UVCS), Extreme Ultraviolet Imaging Telescope (EIT), Michelson Doppler Imager (MDI), and Large Angle Spectrometric Coronagraph (LASCO). Our models show that shock formation occurs at relatively low coronal heights, typically within 1.2 RsuN, depending on the model parameters. The shock formation is calculated using the well-established wave breaking criterion given by the intersection of C+ characteristics in the space-time plane. Our models show that shock heating by slow magnetosonic waves is expected to be relevant at most heights in solar coronal plumes, although such waves are probably not the main energy supply mechanism.

Cuntz, Manfred↗

GALAXY HALO TRUNCATION AND GIANT ARC SURFACE BRIGHTNESS RECONSTRUCTION IN THE CLUSTER MACSJ1206.2-0847

In this work, we analyze the mass distribution of MACSJ1206.2-0847, particularly focusing on the halo properties of its cluster members. The cluster appears relaxed in its X-ray emission, but has a significant amount of intracluster light that is not centrally concentrated, suggesting that galaxy-scale interactions are still ongoing despite the overall relaxed state. The cluster lenses 12 background galaxies into multiple images and one galaxy at z = 1.033 into a giant arc and its counterimage. The multiple image positions and the surface brightness (SFB) distribution of the arc, which is bent around several cluster members, are sensitive to the cluster galaxy halo properties. We model the cluster mass distribution with a Navarro-Frenk-White profile and the galaxy halos with two parameters for the mass normalization and the extent of a reference halo assuming scalings with their observed near-infrared light. We match the multiple image positions at an rms level of 0.''85 and can reconstruct the SFB distribution of the arc in several filters to a remarkable accuracy based on this cluster model. The length scale where the enclosed galaxy halo mass is best constrained is about 5 effective radii-a scale in between those accessible to dynamical and field strong-lensing mass estimates on the one hand and galaxy-galaxy weak-lensing results on the other hand. The velocity dispersion and halo size of a galaxy with m{sub 160W,AB} = 19.2 and M{sub B,Vega} = -20.7 are {sigma} = 150 km s{sup -1} and r Almost-Equal-To 26 {+-} 6 kpc, respectively, indicating that the halos of the cluster galaxies are tidally stripped. We also reconstruct the unlensed source, which is smaller by a factor of {approx}5.8 in area, demonstrating the increase in morphological information due to lensing. We conclude that this galaxy likely has star-forming spiral arms with a red (older) central component.

79 ASTRONOMY AND ASTROPHYSICS↗

Damage predictions of aluminum thin-walled structures subjected to explosive loads.

Predicting failure of thin-walled structures from explosive loading is a very complex task. The problem can be divided into two parts; the detonation of the explosive to produce the loading on the structure, and secondly the structural response. First, the factors that affect the explosive loading include: size, shape, stand-off, confinement, and chemistry of the explosive. The goal of the first part of the analysis is predicting the pressure on the structure based on these factors. The hydrodynamic code CTH is used to conduct these calculations. Secondly, the response of a structure from the explosive loading is predicted using a detailed finite element model within the explicit analysis code Presto. Material response, to failure, must be established in the analysis to model the failure of this class of structures; validation of this behavior is also required to allow these analyses to be predictive for their intended use. The presentation will detail the validation tests used to support this program. Validation tests using explosively loaded aluminum thin flat plates were used to study all the aspects mentioned above. Experimental measurements of the pressures generated by the explosive and the resulting plate deformations provided data for comparison against analytical predictions. These included pressure-time histories and digital image correlation of the full field plate deflections. The issues studied in the structural analysis were mesh sensitivity, strain based failure metrics, and the coupling methodologies between the blast and structural models. These models have been successfully validated using these tests, thereby increasing confidence of the results obtained in the prediction of failure thresholds of complex structures, including aircraft.

36 MATERIALS SCIENCE↗

Upgrading and testing the 3D reconstruction of gamma-ray air showers as observed with an array of Cherenkov telescopes

Stereoscopic arrays of Imaging Cherenkov Telescopes allow to reconstruct gamma-ray-induced showers in 3 dimensions. An analysis method based on a simple 3D-model of electromagnetic showers and implemented in the framework of the H.E.S.S. experiment was recently improved by an additional quality criterion which reduces the background contamination by a factor of about 2 in the case of extended sources, while hardly affecting gamma-ray selection efficiency. Moreover, the dramatic flares of PKS 2155-304 in July 2006, which provided H.E.S.S. data with an almost pure gamma-ray sample, offered the unique opportunity of a precision test of the 3D-reconstruction method as well as of the H.E.S.S. simulations used in its calibration. An agreement at a few percent level is found between data and simulations for the distributions of all 3D shower parameters.

79 ASTRONOMY AND ASTROPHYSICS↗

How Much Chemistry Does a Deep Neural Network Need to Know to Make Accurate Predictions?

In the last few years, we have seen the rise of deep learning applications in a broad range of computational chemistry research problems. Using human-engineered chemical features, such as molecular descriptors and fingerprints, deep learning models have shown similar, if not better performance that most traditional machine learning algorithms. Recently, we reported on the development of Chemception, a deep convolutional neural network (CNN) architecture for general-purpose small molecule property prediction. On average, Chemception matched the performance of expert-developed QSAR/QSPR models trained on chemical features (molecular fingerprints), despite that it was trained on just 2D images of molecular drawings with minimal chemical information. Here, we investigate the effects of systematically removing and adding basic chemical information to the image channels of the 2D images used to train Chemception. By augmenting our images with only 3 additional basic chemical information, we demonstrate the improvement of Chemception performance – that it is now more accurate than contemporary deep learning models trained on ECFP fingerprints for the prediction of toxicity, activity and solvation free energy, as well as physics-based free energy simulation methods for computing solvation properties. By altering the chemical information content in the image channels, and examining the resulting performance of Chemception, we also identify to two different “learning patterns” in toxicity/activity as compared to solvation free energy, and it parallels the fundamental differences in contemporary chemistry research for predicting toxicity/activity and solvation free energy.

Artificial Intelligence, Cheminformatics, Computat↗

Temporal changes in endmember abundances, liquid water and water vapor over vegetation at Jasper Ridge

Imaging spectrometry offers a new way of deriving ecological information about vegetation communities from remote sensing. Applications include derivation of canopy chemistry, measurement of column atmospheric water vapor and liquid water, improved detectability of materials, more accurate estimation of green vegetation cover and discrimination of spectrally distinct green leaf, non-photosynthetic vegetation (NPV: litter, wood, bark, etc.) and shade spectra associated with different vegetation communities. Much of our emphasis has been on interpreting Airborne Visible/Infrared Imaging Spectrometry (AVIRIS) data spectral mixtures. Two approaches have been used, simple models, where the data are treated as a mixture of 3 to 4 laboratory/field measured spectra, known as reference endmembers (EM's), applied uniformly to the whole image, to more complex models where both the number of EM's and the types of EM's vary on a per-pixel basis. Where simple models are applied, materials, such as NPV, which are spectrally similar to soils, can be discriminated on the basis of residual spectra. One key aspect is that the data are calibrated to reflectance and modeled as mixtures of reference EM's, permitting temporal comparison of EM fractions, independent of scene location or data type. In previous studies the calibration was performed using a modified-empirical line calibration, assuming a uniform atmosphere across the scene. In this study, a Modtran-based calibration approach was used to map liquid water and atmospheric water vapor and retrieve surface reflectance from three AVIRIS scenes acquired in 1992 over the Jasper Ridge Biological Preserve. The data were acquired on June 2nd, September 4th and October 6th. Reflectance images were analyzed as spectral mixtures of reference EM's using a simple 4 EM model. Atmospheric water vapor derived from Modtran was compared to elevation, and community type. Liquid water was compare to the abundance of NPV, Shade and Green Vegetation (VG) for select sites to determine whether a relationship existed, and under what conditions the relationship broke down. Temporal trends in endmember fractions, liquid water and atmospheric water vapor were investigated also. The combination of spectral mixture analysis and the Modtran based atmospheric/liquid water models was used to develop a unique vegetation community description.

Roberts, Dar A.↗