Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Spectral clustering”

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 397 records · Page 22
All sourcesNASA NTRSDOE OSTI
NASA NTRS2004

An Artificial Intelligence Classification Tool and Its Application to Gamma-Ray Bursts

Despite being the most energetic phenomenon in the known universe, the astrophysics of gamma-ray bursts (GRBs) has still proven difficult to understand. It has only been within the past five years that the GRB distance scale has been firmly established, on the basis of a few dozen bursts with x-ray, optical, and radio afterglows. The afterglows indicate source redshifts of z=1 to z=5, total energy outputs of roughly 10(exp 52) ergs, and energy confined to the far x-ray to near gamma-ray regime of the electromagnetic spectrum. The multi-wavelength afterglow observations have thus far provided more insight on the nature of the GRB mechanism than the GRB observations; far more papers have been written about the few observed gamma-ray burst afterglows in the past few years than about the thousands of detected gamma-ray bursts. One reason the GRB central engine is still so poorly understood is that GRBs have complex, overlapping characteristics that do not appear to be produced by one homogeneous process. At least two subclasses have been found on the basis of duration, spectral hardness, and fluence (time integrated flux); Class 1 bursts are softer, longer, and brighter than Class 2 bursts (with two second durations indicating a rough division). A third GRB subclass, overlapping the other two, has been identified using statistical clustering techniques; Class 3 bursts are intermediate between Class 1 and Class 2 bursts in brightness and duration, but are softer than Class 1 bursts. We are developing a tool to aid scientists in the study of GRB properties. In the process of developing this tool, we are building a large gamma-ray burst classification database. We are also scientifically analyzing some GRB data as we develop the tool. Tool development thus proceeds in tandem with the dataset for which it is being designed. The tool invokes a modified KDD (Knowledge Discovery in Databases) process, which is described as follows.

Hakkila, Jon↗
NASA NTRS2006

Star Formation Rates in Cooling Flow Clusters: A UV Pilot Study with Archival XMM-Newton Optical Monitor Data

We have analyzed XMM-Newton Optical Monitor (OM) UV (180-400 nm) data for a sample of 33 galaxies. 30 are cluster member galaxies, and nine of these are central cluster galaxies (CCGs) in cooling flow clusters having mass deposition rates which span a range of 8 - 525 Solar Mass/yr. By comparing the ratio of UV to 2MASS J band fluxes, we find a significant UV excess in many, but not all, cooling flow CCGs, a finding consistent with the outcome of previous studies based on optical imaging data (McNamara & O'Connell 1989; Cardiel, Gorgas, & Aragon-Salamanca 1998; Crawford et al. 1999). This UV excess is a direct indication of the presence of young massive stars, and therefore recent star formation, in these galaxies. Using the Starburst99 spectral energy distribution (SED) model of continuous star formation over a 900 Myr period, we derive star formation rates of 0.2 - 219 solar Mass/yr for the cooling flow sample. For 2/3 of this sample it is possible to equate Chandra/XMM cooling flow mass deposition rates with UV inferred star formation rates, for a combination of starburst lifetime and IMF slope. This is a pilot study of the well populated XMM UV cluster archive and a more extensive follow up study is currently underway.

Hicks, A. K.↗
NASA NTRS1973

Adaptive on-line classification of multi-spectral scanner data

A possible solution to the analysis of the massive amounts of multi-spectral scanner data from the Earth Resource Technical Satellite (ERTS) program is proposed. This solution is offered as an adaptive on-line classification scheme. The classifier is described as well as its controller which is based on ground truth data. Cluster analysis is presented as an alternative approach to the ground truth data. Adaptive feature selection is discussed and possible mini-computer implementations are offered.

Fromm, F. R.↗
NASA NTRS1984

Anaysis of the quality of image data required by the LANDSAT-4 Thematic Mapper and Multispectral Scanner

The spatial, geometric, and radiometric qualities of LANDSAT 4 thematic mapper (TM) and multispectral scanner (MSS) data were evaluated by interpreting, through visual and computer means, film and digital products for selected agricultural and forest cover types in California. Multispectral analyses employing Bayesian maximum likelihood, discrete relaxation, and unsupervised clustering algorithms were used to compare the usefulness of TM and MSS data for discriminating individual cover types. Some of the significant results are as follows: (1) for maximizing the interpretability of agricultural and forest resources, TM color composites should contain spectral bands in the visible, near-reflectance infrared, and middle-reflectance infrared regions, namely TM 4 and TM % and must contain TM 4 in all cases even at the expense of excluding TM 5; (2) using enlarged TM film products, planimetric accuracy of mapped poins was within 91 meters (RMSE east) and 117 meters (RMSE north); (3) using TM digital products, planimetric accuracy of mapped points was within 12.0 meters (RMSE east) and 13.7 meters (RMSE north); and (4) applying a contextual classification algorithm to TM data provided classification accuracies competitive with Bayesian maximum likelihood.

Colwell, R. N.↗
NASA NTRS1991

Laboratory studies of the newly discovered infrared band at 4705.2 cm-1 (2.1253 micrometers) in the spectrum of Io: the tentative identification of CO2

We discuss over 120 laboratory experiments pertaining to the identification of the new absorption band discovered by Trafton et al. (1991) at 4705.2 cm-1 (2.1253 micrometers) in the spectrum of Io. It is shown that this band is not due to overtones or combinations of the fundamental bands associated with the molecules (or their chemical complexes) already identified on Io, namely, SO2, H2S, and H2O. Thus, this band is due to a new, previously unidentified, component of Io. Experiments also demonstrate that the band is not due to molecular H2 frozen in SO2 frosts. Since the frequency of this band is very close to the first overtone of the nu 3 asymmetric stretching mode of CO2, we have investigated the spectral behavior of CO2 under a variety of conditions appropriate for Io. The profile of the Io band is not consistent with the rotational envelope expected for single, freely rotating, gaseous CO2 under Io-like conditions. It was found that pure, solid CO2 and CO2 intimately mixed in a matrix of solid SO2 and H2S produce bands with similar widths (5-10 cm-1), but that these bands consistently fall at frequencies about 10-20 cm-1 (approximately 0.007 micrometer) lower than the Io band. CO2 in SO2 : H2S ices also produces several additional bands that are not in the Io spectra. The spectral fit improves, however, as the CO2 concentration in SO2 increases, suggesting that CO2-CO2 interactions might be involved. A series of Ar : CO2 and Kr : CO2 matrix isolation experiments, as well as laboratory work done elsewhere, show that CO2 clustering shifts the band position to higher frequencies and provides a better fit to the Io band. Various laboratory experiments have shown that gaseous CO2 molecules have a propensity to cluster between 80 and 100 K, temperatures similar to those found on the colder regions of Io. We thus tentatively identify the newly discovered Io band at 4705.2 cm-1 (2.1253 micrometers) with CO2 multimers or "clusters" on Io. Whether these clusters are buried within an SO2 frost, reside on the surface, or are in a residual, steady-state "atmospheric aerosol" population over local coldtraps is not entirely clear, although we presently favor the latter possibility. The size of these clusters is not well defined, but evidence suggests groups of more than four molecules are required. The absorption strength of the 2 nu 3 CO2 cluster overtone determined in the laboratory, in conjunction with the observed strength of the Io band, suggests that the disk-integrated abundance of CO2 is less than 1% that of the SO2. Studies of the sublimation behavior of CO2 indicate that it probably resides predominantly in the cooler areas (< 100 K) of Io. The relative constancy of the Io feature over a variety of orbital phases suggests that the polar regions may contain much of the material. Some consequences of the physical properties of CO2 under conditions pertinent to Io are discussed. The presence of CO2 clusters on Io could be verified by the detection of any one of several other infrared bands associated with the CO2 molecule, of which the strongest are the nu 3 12CO2 asymmetric stretch fundamental near 2350 cm-1 (4.25 micrometers) and the nu 2 bending mode fundamental near 660 cm-1 (15.1 micrometers). Weaker bands that may also be detectable include the nu 3 13CO2 asymmetric stretch fundamental near 2280 cm-1 (4.39 micrometers), the 2 nu 2 + nu 3 combination/overtone band near 3600 cm-1 (2.78 micrometers), and the nu 1 + nu 3 combination band near 3705 cm-1 (2.70 micrometers).

NASA Center ARC↗
NASA NTRS2017

Creating Aerosol Types Form CHemistry (CATCH): a New Algorithm to Extend the Link Between Remote Sensing and Models

Current remote sensing methods can identify aerosol types within an atmospheric column, presenting an opportunity to incrementally bridge the gap between remote sensing and models. Here a new algorithm was designed for Creating Aerosol Types from CHemistry (CATCH). CATCH-derived aerosol types—dusty mix, maritime, urban, smoke, and fresh smoke—are based on first-generation airborne High Spectral Resolution Lidar (HSRL-1) retrievals during the Ship-Aircraft Bio-Optical Research (SABOR) campaign, July/August 2014. CATCH is designed to derive aerosol types from model output of chemical composition. CATCH-derived aerosol types are determined by multivariate clustering of model-calculated variables that have been trained using retrievals of aerosol types from HSRL-1. CATCH-derived aerosol types (with the exception of smoke) compare well with HSRL-1 retrievals during SABOR with an average difference in aerosol optical depth (AOD) <0.03. Data analysis shows that episodic free tropospheric transport of smoke is underpredicted by the Goddard Earth Observing System- with Chemistry (GEOS-Chem) model. Spatial distributions of CATCH-derived aerosol types for the North American model domain during July/August 2014 show that aerosol type-specific AOD values occurred over representative locations: urban over areas with large population, maritime over oceans, smoke, and fresh smoke over typical biomass burning regions. This study demonstrates that model-generated information on aerosol chemical composition can be translated into aerosol types analogous to those retrieved from remote sensing methods. In the future, spaceborne HSRL-1 and CATCH can be used to gain insight into chemical composition of aerosol types, reducing uncertainties in estimates of aerosol radiative forcing.

K W Dawson↗
NASA NTRS2005

Star Formation Rates in Cooling Flow Clusters: A UV Pilot Study with Archival XMM-Newton Optical Monitor Data

We have analyzed XMM-Newton Optical Monitor (OM) UV (180-400 nm) data for a sample of 33 galaxies. 30 are cluster member galaxies, and nine of these are central cluster galaxies (CCGs) in cooling flow clusters having mass deposition rates which span a range of 8 - 525 solar mass per year. By comparing the ratio of UV to 2MASS J band fluxes, we find a significant UV excess in many, but not all, cooling flow CCGs, a finding consistent with the outcome of previous studies based on optical imaging data (McNamara & O Connell 1989; Cardiel, Gorgas, & Aragon-Salamanca 1998; Crawford et al. 1999). This UV excess is a direct indication of the presence of young massive stars, and therefore recent star formation, in these galaxies. Using the Starburst99 spectral energy distribution (SED) model of continuous star formation over a 900 Myr period, we derive star formation rates of 0.2 - 219 solar mass per year for the cooling flow sample. For 2/3 of this sample it is possible to equate Chandra/XMM cooling flow mass deposition rates with UV inferred star formation rates, for a combination of starburst lifetime and IMF slope. This is a pilot study of the well populated XMM UV cluster archive and a more extensive follow up study is currently underway.

Hicks, A. K.↗
NASA NTRS1994

Surface Brightness Profiles and Energetics of Intracluster Gas in Cool Galaxy Clusters and ROSAT Observations of Bright, Early-Type Galaxies

Preliminary results on the elliptical galaxy NGC 1407 were published in the proceedings of the first ROSAT symposium. NGC 1407 is embedded in diffuse X-ray-emitting gas which is extensive enough that it is likely to be related to the surrounding group of galaxies, rather than just NGC 1407. Spectral data for NGC 1407 (AO2) and IC 1459 (AO3) are also included in a complete sample of elliptical galaxies I compiled in collaboration with David Davis. This allowed us to construct the first complete X-ray sample of optically-selected elliptical galaxies. The complete sample allows us to apply Malmquist bias corrections to the observed correlation between X-ray and optical luminosities. I continue to work on the implications of this first complete X-ray sample of elliptical galaxies. Paul Eskridge Dave Davis and I also analyzed three long ROSAT PSPC observations of the small (but not dwarf) elliptical galaxy M32. We found the X-ray spectra and variability to be consistent with either a Low Mass X-Ray Binary (LMXRB) or a putative 'micro"-AGN.

White, Raymond E., III↗
NASA NTRS1974

Detection of new celestial objects at far-infrared wavelengths

During a high-altitude balloon flight, sources of far-infrared radiation have been detected with apparent fluxes greater than or about equal to 3 pW/sq cm in the spectral band from 50 to 500 microns. While the sources are not uniquely identifiable with any well-known class of celestial objects, two associations should be noted. Many of the objects lie close to local stars; there is also a tendency to cluster around the ecliptic plane. Alternatively, a new class of celestial objects may have been observed.

Friedlander, M. W.↗
NASA NTRS1981

1548C27 - An interesting new cometary nebula

The object 1548C27, a faint cometary nebula of classical form, discovered by an examination of early Near Infrared Photographic Sky Survey (NIPSS) data (1979) is presented. Direct imaging and polarimetric, photometric, and spectroscopic observations are reported. Early survey test photographs show that the object lies at R.A. 19h40m48s, Decl. +23 deg 17 arcmin 09 arc sec (1950) in the Vulpecula constellation in the immediate vicinity of the complex H II region and galactic cluster NGC 6820/6823. From the photographs, the nebula was estimated to be 15 m visual and of color class one. The object was observed spectroscopically in the region 5700-6800 A using an Image Dissector Scanner with a 1.8 m reflector, and the spectral scan, obtained on November 4, 1978, is presented. New information on cometary nebulae may further illuminate the evolutionary importance of the objects.

Craine, E. R.↗
NASA NTRS1991

Spectral observations of the extreme ultraviolet background

The design of a grazing incidence spectrometer for measuring the diffusive extreme-ultraviolet astronomical background radiation is outlined, and the data obtained on the diffusive background between 80 and 650 A is discussed. The strongest emission line detected is solar backscattered He I 584 A radiation; solar 304 A radiation resonantly scattered by geocoronal He II is observed in both first and second order; and each of three remaining spectral features is found to be consistent with those expected from hot interstellar gas. Of these features, the strongest is the feature at 635 A identified as 630 A O v emission. A feature centered at about 99 A is fitted by a cluster of Fe XVIII and Fe XIX lines from gas at log T = 6.6-6.8, and a feature discovered near 178 A is found to be consistent with Fe X and Fe XI emission from gas at log T = 6.

Labov, Simon E.↗
NASA NTRS2007

A Catalog of Galaxy Clusters Observed by XMM-Newton

Images and the radial profiles of the temperature, abundance, and brightness for 70 clusters of galaxies observed by XMM-Newton are presented along with a detailed discussion of the data reduction and analysis methods, including background modeling, which were used in the processing. Proper consideration of the various background components is vital to extend the reliable determination of cluster parameters to the largest possible cluster radii. The various components of the background including the quiescent particle background, cosmic diffuse emission, soft proton contamination, and solar wind charge exchange emission are discussed along with suggested means of their identification, filtering, and/or their modeling and subtraction. Every component is spectrally variable, sometimes significantly so, and all components except the cosmic background are temporally variable as well. The distributions of the events over the FOV vary between the components, and some distributions vary with energy. The scientific results from observations of low surface brightness objects and the diffuse background itself can be strongly affected by these background components and therefore great care should be taken in their consideration.

Snowden, S. L.↗
NASA NTRS2020

Atacama Cosmology Telescope: Component-separated maps of CMB temperature and the thermal Sunyaev-Zel'dovich effect

Optimal analyses of many signals in the cosmic microwave background (CMB) require map-level extraction of individual components in the microwave sky, rather than measurements at the power spectrum level alone. To date, nearly all map-level component separation in CMB analyses has been performed exclusively using satellite data. In this paper, we implement a component separation method based on the internal linear combination (ILC) approach which we have designed to optimally account for the anisotropic noise (in the 2D Fourier domain) often found in ground-based CMB experiments. Using this method, we combine multi-frequency data from the Planck satellite and the Atacama Cosmology Telescope Polarimeter (ACTPol) to construct the first wide area (≈2100 sq. deg.), arcminute-resolution component-separated maps of the CMB temperature anisotropy and the thermal Sunyaev-Zel'dovich (tSZ) effect sourced by the inverse-Compton scattering of CMB photons off hot, ionized gas. Our ILC pipeline allows for explicit deprojection of various contaminating signals, including a modified blackbody approximation of the cosmic infrared background (CIB) spectral energy distribution. The cleaned CMB maps will be a useful resource for CMB lensing reconstruction, kinematic SZ cross-correlations, and primordial non-Gaussianity studies. The tSZ maps will be used to study the pressure pro les of galaxies, groups, and clusters through cross-correlations with halo catalogs, with dust contamination controlled via CIB deprojection. The data products described in this paper are available on LAMBDA.

Mathew S. Madhavacheril↗
NASA NTRSDate not supplied

NASA/GEWEX SRB Surface Radiation Budget Fluxes Through the Prism of Weather States

The NASA/GEWEX Surface Radiation Budget Version 4 (SRB Rel4) product is now publicly available, covering the period from July 1983 through June 2017. This supersedes the SRB Rel3 (Stackhouse et al., 2011) product which has been used in the community for a wide variety of applications, including climate model validation, agriculture, solar energy, and architecture. SRB Rel4 uses the newly recalibrated and processed ISCCP HXS product as its primary input for cloud and radiance data, replacing ISCCP DX with a ninefold increase in pixel count (10km instead of 30km). This version retains a 1°x1° resolution but benefits from a much larger number of samples per grid box than the SRB Rel3. ISCCP also provides an atmospheric temperature and moisture dataset known as nnHIRS which we use here, along with Seaflux and Landflux surface and near-surface meteorological parameters. Rel4 incorporates several important algorithm improvements. These include recalculated shortwave (SW) atmospheric transmissivities and reflectivities yielding a somewhat less transmissive atmosphere. Both shortwave and longwave (LW) now also include variable aerosol composition and radiative properties, allowing for the use of a detailed aerosol history from the Max Planck Institute Aerosol Climatology (MACv1). LW and SW algorithms now produce pristine sky fluxes, allowing the aerosol flux effects to be quantified. For SW, ocean albedo and snow/ice albedo are improved from Release 3. Total solar irradiance is now variable, and reduced to an average of 1361 Wm-2. The radiative treatment of ice cloud is improved. For the LW, a climatological monthly varying spectral surface emissivity is added. Here we evaluate the SRB Rel4 top of atmosphere (TOA) and surface fluxes in the context of weather states and oceanic regions. Weather states (Tselioudis et al., 2013, Tselioudis et al., 2021) are defined by K-means clustering 2-D histograms of satellite-retrieved cloud optical depths and cloud top pressures. The most recent (Tselioudis et al., 2021) weather states derived from the International Satellite Cloud Climatology Project (ISCCP) H-series data results in eight cloud weather states and one clear sky weather state, for a total of nine weather states. We examine the statistics of SRB Rel4 fluxes for each weather state and a variety of relevant land and ocean regions. We compare to similar analyses performed on CERES EBAF and SYN1Deg flux products, and validate against Baseline Surface Radiometer Network (BSRN) land stations and Pacific Marine Environmental Laboratory (PMEL) ocean buoy data. It is demonstrated that the weather state concept is valuable when examining the comparative strengths and weaknesses of satellite radiative flux algorithms.

Atmospheric Radiation↗
NASA NTRS2004

From the Cluster Temperature Function to the Mass Function at Low Z

This XMM project consisted of three observations of the nearby, hot galaxy cluster Triangulum Australis, one of the cluster center and two offsets. The goal was to measure the radial gas temperature profile out to large radii and derive the total gravitating mass within the radius of average mass overdensity 500. The central pointing also provides data for a detailed two-dimensional gas temperature map of this interesting cluster. We have analyzed all three observations. The derivation of the temperature map using the central pointing is complete, and the paper is soon to be submitted. During the course of this study and of the analysis of archival XMM cluster observations, it became apparent that the commonly used XMM background flare screening techniques are often not accurate enough for studies of the cluster outer regions. The information on the cluster's total masses is contained at large off-center distances, and it is precisely the temperatures for those low-brightness regions that are most affected by the detector background anomalies. In particular, our two offset observations of the Triangulum have been contaminated by the background flares ("bad cosmic weather") to a degree where they could not be used for accurate spectral analysis. This forced us to expand the scope of our project. We needed to devise a more accurate method of screening and modeling the background flares, and to evaluate the uncertainty of the XMM background modeling. To do this, we have analyzed a large number of archival EPIC blank-field and closed-cover observations. As a result, we have derived stricter background screening criteria. It also turned out that mild flares affecting EPIC-pn can be modeled with an adequate accuracy. Such modeling has been used to derive our Triangulum temperature map. The results of our XMM background analysis, including the modeling recipes, are presented in a paper which is in final preparation and will be submitted soon. It will be useful not only for our future analysis but for other XMM cluster observations as well.

Mushotzky, Richard↗
NASA NTRS1986

Stellar populations; Proceedings of the Meeting, Baltimore, MD, May 20-22, 1986

Various papers covering recent findings on stellar populations are presented. The topics addressed include: the bulges of the Galaxy and M31, kinematics of the local subdwarfs and the collapse of the Galaxy, the red stellar population of the local group dwarf galaxies, globular clusters in the local group galaxies, the initial mass function, and star formation histories of galactic disks. Also discussed are: yield and abundance constraints on galactic chemical evolution, population synthesis in early type galaxies, the evolution of stellar populations in galaxies, a theoretical perspective on spectral evolution of galaxies, and the structure and kinematics of galaxy populations.

Norman, Colin A.↗
NASA NTRS2005

Plasma sheet turbulence observed by Cluster II

Cluster fluxgate magnetometer (FGM) and ion spectrometer (CIS) data are employed to analyze magnetic field fluctuations within the plasma sheet during passages through the magnetotail region in the summers of 2001 and 2002 and, in particular, to look for characteristics of magnetohydrodynamic (MHD) turbulence. Power spectral indices determined from power spectral density functions are on average larger than Kolmogorov's theoretical value for fluid turbulence as well as Kraichnan's theoretical value for MHD plasma turbulence. Probability distribution functions of the magnetic fluctuations show a scaling law over a large range of temporal scales with non-Gaussian distributions at small dissipative scales and inertial scales and more Gaussian distribution at large driving scales. Furthermore, a multifractal analysis of the magnetic field components shows scaling behavior in the inertial range of the fluctuations from about 20 s to 13 min for moments through the fifth order. Both the scaling behavior of the probability distribution functions and the multifractal structure function suggest that intermittent turbulence is present within the plasma sheet. The unique multispacecraft aspect and fortuitous spacecraft spacing allow us to examine the turbulent eddy scale sizes. Dynamic autocorrelation and cross correlation analysis of the magnetic field components allow us to determine that eddy scale sizes fit within the plasma sheet. These results suggest that magnetic field turbulence is occurring within the plasma sheet resulting in turbulent energy dissipation.

Weygand, James M.↗
NASA NTRS2007

Global Simulation of Electromagnetic Ion Cyclotron Waves

It is well known that the effects of electromagnetic ion cyclotron (EMIC) waves on ring current (RC) ion and radiation belt (RB) electron dynamics strongly depend on such particle/wave characteristics as the phase-space distribution function, frequency, wave-normal angle, wave energy, and the form of wave spectral energy density. The consequence is that accurate modeling of EMIC waves and RC particles requires robust inclusion of the interdependent dynamics of wave growth/damping, wave propagation, and particles. Such a self-consistent model is being progressively developed by Khazanov et al. [2002 - 2007]. This model is based on a system of coupled kinetic equations for the RC and EMIC wave power spectral density along with the ray tracing equations. We will discuss the recent progress in understanding EMIC waves formation mechanisms in the inner magnetosphere. This problem remains unsettled in spite of many years of experimental and theoretical studies. Modern satellite observations by CRRES, Polar and Cluster still do not reveal the whole picture experimentally since they do not stay long enough in the generation region to give a full account of all the spatio-temporal structure of EMIC waves. The complete self-consistent theory taking into account all factors significant for EMIC waves generation remains to be developed. Several mechanisms are discussed with respect to formation of EMIC waves, among them are nonlinear modification of the ionospheric reflection by precipitating energetic protons, modulation of ion-cyclotron instability by long-period (Pc3/4) pulsations, reflection of waves from layers of heavy-ion gyroresonances, and nonlinearities of wave generation process. We show that each of these mechanisms have their attractive features and explains certain part experimental data but any of them, if taken alone, meets some difficulties when compared to observations. We conclude that development of a refined nonlinear theory and further correlated analysis of modern satellite and ground-based data is needed to solve this very intriguing problem.

Khazanov, G. V.↗