Optical Dataset formed after cross matching : https://www.kaggle.com/bhanvimenghani/optical-csv
It was in the middle of the eighteenth century that Kant and Wright first suggested that the Milky Way represents a finite-sized disk-like system of stars. As an extension of their philosophical argument about the nature of the Galaxy, Kant went on to suggest that if the Milky Way is limited in extent, perhaps the diffuse and very faint "elliptical nebulae" seen in the night sky might actually be extremely distant disk-like systems, similar to our own. He called these objects island universes.
As a first step in understanding any new collection of objects, it is necessary to classify them according to their intrinsic characteristics. Hubble played a major role in this classification. Hubble in his book "The realm of nebulae", Hubble proposed that galaxies be grouped into three primary categories based on their overall appearance. This morphological classification scheme, known as the Hubble sequence, divides galaxies into ellipticals (E's), spirals, and irregulars (Irr's).The spirals are further subdivided into two parallel sequences, the normal spirals (S's), and the barred spirals (SB's). A transitional class of galaxies between ellipticals and spirals is known as lenticulars. It can be either normal (S0's) or barred (SB0's). Hubble then arranged this morphological sequence in the form of a tuning-fork diagram.
Hubble originally thought (incorrectly) that the tuning-fork diagram could be interpreted as an evolutionary sequence for galaxies. As a result, he referred to galaxies toward the left of the diagram as early types and to those toward the right as late types, terminology that is still in widespread use today. Within the category of ellipticals, Hubble made divisions based on the observed ellipticity of the galaxy, defined by
Where α and β are the apparent major and minor axes of the ellipse, respectively, projected onto the plane of the sky. Galaxies with ellipticities greater than ∈=0.7 have never been observed, implying that no E galaxies with intrinsic ellipticities greater than 0.7 appear to exist. The apparent ellipticity may not correspond well to an actual ellipticity since the orientation of the spheroid to our line of sight plays a crucial role in our observations.
This is a spheroidal galaxy that has lengths a=b and c<a.
A prolate spheroidal galaxy has axis lengths b=c and a>b.
Hubble subdivided the spiral sequences into Sa, Sab, Sb, Sbc, Sc, and SBa, SBab,
SBb, SBbc, SBc. The galaxies with the most prominent bulges the largest bulge-to-disk luminosity ratios,(L_bulge⁄(L_disk~0.3)) the most tightly wound spiral arms. and the smoothest distribution of stars in the arms are classified as Sa's (or SBa's), while Sc's (or SBc's) have smaller bulge-to-disk ratios (L_bulge⁄L_disk ~0.005) Hubble split the remaining category of irregulars into Irr I if there was at least some hint of an organized structure, such as spiral arms, and Irr II for the most extremely disorganized structures. Hubble split the remaining category of irregulars into Irr I if there was at least some hint of an organized structure, such as spiral arms, and Irr II for the most extremely disorganized structures.
As a further refinement to the system, the lenticular galaxies are also sometimes subdivided according to the amount of dust absorption in their disks. S01 galaxies have no dust their disks, while S03 galaxies have significant amounts of dust, and similarly for SB01 through SB03.In order to make finer distinctions between normal and barred spirals, de Vaucouleurs had also suggested referring to normal spirals as SA rather than simply S. Intermediate types with weak bars are then characterized as SAB, and strongly barred galaxies are SB. The overall picture of classification looks like this
As we all know the galaxies are basically of three types -- Elliptical, Spiral and Irregular.
- They have smooth light distributions. The Hubble classifications of these galaxies contain an integer 'n' that describes the elongation of the galaxy image. The classification is determined by the semi major and semi minor axes of the galaxy's isophotes (a line in a diagram connecting points where the intensity of the light is same). It is denoted by En, where
n = 10(1- b/a)
They are given number from 0 to 7. An E0 galaxy looks like a circle whereas an E7 galaxy looks long and elongated.
In a 3D ellipsoidal shape with semi axes a, b, c if:
a=b>c it is called oblate, a>b=c it is called prolate, a>b>c it is called tri-axial.
-They have no current star formation and are usually found in galaxy clusters and compact groups.
-They have old stars and contains little gas and dust.
M89 -- E0 Galaxy
NGC 4621 -- E5 Galaxy
-
Components of S galaxies are spiral arms, bulge, bar, spheroid
-
They have light profiles with a characteristic spiral spin.
-
They contain both new and old stars. New stars are formed, especially in the spiral arms.
-
They seem to avoid dense groups and are very rare in clusters.
NGC 2008
NGC 1300
NGC 428 Barred Spiral Galaxy
=================================================================================================================================================================================================================================================
-
It doesn't have a distinct regular shape like spiral or elliptical galaxies and do not fall into any classes of Hubble sequence.
-
They are usually found in groups or clusters where the collision or near-misses between the galaxies is common.
-
One type of irregular galaxies is 'Starburst Galaxy'. They shine very brightly as the new stars are born in a short period of time.
NGC 1427A
-
They have central bulge but have no spiral arms. If the central bulge is not very bright it will be difficult to differentiate between lenticular galaxy and E0 galaxy.
-
Few lenticular galaxies have bars like spiral galaxies; they are called barred lenticular galaxies and denoted as SB0. Other normal galaxies are denoted as S0.
NGC 936 -- SB0 Galaxy
Application:
Most galaxies can be broadly separated into two morphological types -- spiral and elliptical .Galaxy Zoo was the first attempt to analyze the distribution of a large number of spiral and elliptical galaxies in the local universe. Using the power of crowd sourcing, it provided morphological classifications of nearly 900,000 galaxies. A more recent catalog of galaxy morphology is the catalog of ∼3,000,000 Sloan Digital Sky Survey (SDSS) galaxies classified automatically using machine learning. While the catalog is large, it is limited in the sense that the vast majority of the galaxies in that catalog do not have spectroscopic data.
Photometric redshifts
An area of astrophysics that has greatly increased in popularity in the last few years is the estimation of redshifts from photometric data. This is because, although the distances are less accurate than those obtained with spectra, the sheer number of objects with photometric measurements can often make up for the reduction in individual accuracy by suppressing the statistical noise of an ensemble calculation.
o test machine learning and variable selection algorithms for computing photometric redshift, optimize it for a specific population of galaxies, and mainly apply these algorithms to provide a large catalog of galaxy morphology and photometric redshift. The catalog is similar to the early Galaxy Zoo 1 catalog, but because it was classified automatically it provides a much higher number of galaxies. Of the ∼ 3 - 106 galaxies in the catalog, ∼ 1.5 - 106 are galaxies with 98% agreement rate with the Galaxy Zoo 1 debiased "super clean" accuracy. It is limited in the sense that, like Galaxy Zoo, it represents the galaxies in the catalog, and not necessarily a complete and unbiased sample of SDSS galaxies. The catalog is publicly available at:
https://figshare.com/articles/Morphology_and_photometric_%20redshift_catalog/4833593
Galaxies
At low redshifts, the calculation of photometric redshifts for normal galaxies is quite straightforward due to the break in the typical galaxy spectrum at 4000A. Thus, as a galaxy is redshifted with increasing distance, the color (measured as a difference in magnitudes) changes relatively smoothly. As a result, both template and empirical photo-z approaches obtain similar results, a root-mean-square deviation of ~ 0.02 in redshift, which is close to the best possible result given the intrinsic spread in the properties
Quasars/AGN
Historically, the calculation of photometric redshifts for quasars and other AGN has been even more difficult than for galaxies, because the spectra are dominated by bright but narrow emission lines, which in broad photometric pass bands can dominate the color. The color-redshift relation of quasars is thus subject to several effects, including degeneracy, one emission line appearing like another at a different redshift, an emission line disappearing between survey filters, and reddening. In addition, the filter sets of surveys are generally designed for normal galaxies and not quasars. The assignment of these quasar photo-zs is thus a complex problem that is amenable to data mining in a similar manner to the classification of AGN.
Other Galaxy Classifications
Many of the physical properties, and thus classification, of a galaxy are determined by its stellar population. The spectrum of a galaxy is therefore another method for classification and can sometimes produce a clearer link to the underlying physics than the morphology. Spectral classification is important because it is possible for a range of morphological types to have the same spectral type, and vice versa, because spectral types are driven by different underlying physical processes.
Galaxies come in a range of different sizes and shapes, or more collectively, morphology. The most well-known system for the morphological classification of galaxies is the Hubble Sequence of elliptical, spiral, barred spiral, and irregular, along with various subclasses. This system correlates to many physical properties known to be important in the formation and evolution of galaxies.
Recently, the popular Galaxy Zoo project has taken an alternative approach to morphological classification, employing crowd sourcing: an application was made available online in which members of the general public were able to view images from the SDSS and assign classifications according to an outlined scheme. The project was very successful, and in a period of six months over 100,000 people provided over 40 million classifications for a sample of 893,212 galaxies, mostly to a limiting depth of r = 17.77 mag. The classifications included categories not previously assigned in astronomical data mining studies, such as edge-on or the handedness of spiral arms, and the project has produced multiple scientific results. The approach represents a complementary one to automated algorithms, because, although humans can see things an algorithm will miss and will be subject to different systematic errors, the runtime is hugely longer: a trained ANN will produce the same 40 million classifications in a few minutes, rather than six months.
- Statistical inference and visualization with very-large-N datasets represents a scientific and technological framework needed to cope with this data flood. Considerable work is being conducted by computer scientists and applied mathematicians in other applied fields so that independent development by astro-statisticians might not be necessary to achieve certain goals.
Aside from the computational challenges with large numbers of data vectors and a large dimensionality, this poses some highly non-trivial statistical problems. The problems are driven not just by the size of the data sets, but mainly (in the statistical context) by the heterogeneity and intrinsic complexity of the data.
- Multivariate analysis with measurement errors and censoring
Astrophysicists often devote as much effort to precise determination of their errors as they devote to the measurements of the quantities of interest. The instruments are carefully calibrated to reduce systematic uncertainties, and background levels and random fluctuations are carefully evaluated to determine random errors.
- Probability and Bayesian computation
Bayes Theorem and Bayes factors are becoming increasingly well known in astronomical research. Part of the problem is conceptual astronomers need training in how to construct likelihoods for familiar parametric situations.
- Links between astrophysical theory and wavelets
Wavelet analysis suffers a profound limitation in astronomy
Theoretical challenges:
- Surveys of Galaxy Redshifts:
Application of the redshift-distance relation (Hubble's law) allows the analysis of the largescale distribution of galaxies. Comparison of the observed redshifts with those expected on the basis of other distance estimates allows mapping of the gravitational field and the underlying density distribution. Estimation of the many inherent selection biases and instrumental limitations is critical in understanding how our view of the universe is affected by our observational perspective and by the way information is received by current technologies.
- Analysis of Patterns in Galaxy Clustering
The current state of play with regard to correlation function analysis, both in terms of computational and sampling problems and with regard to the fundamental limitations of such an approach in giving a complete statistical description of the pattern. Various techniques have been devised to attempt to quantify the topology of the clustering pattern with varying degrees of success.
- Assessment of Sub clustering in Clusters of Galaxies
The lack of adequate cluster samples to provide an unbiased look at the problem, the lack of supplemental data in most clusters to confirm or reject possible small-scale structures.
-
REIONIZATION SOURCES
-
LYMAN ALPHA EMITTERS
-
DROPOUT GALAXIES
-
Luminosity Function
Reference: The Most Distant Galaxies: Theoretical Challenges
-(http://dx.doi.org/10.1063/1.3518844)
-
Challenges and Opportunities of Data Rich Astronomy
-
Star-Galaxy Image Classification
A classical problem in the analysis of astronomical panoramic imagery.The characteristic resolution of an astronomical image is given through a combination of the instrumental resolution.
-Dynamical, Real-Time Classification of Astronomical Transient Events in Synoptic Sky Surveys
A need for dynamical, real-time scientific measurement systems, consisting of discovery instruments or sensors, a real-time computational analysis and decision engine, and optimized follow-up instruments which can be deployed selectively in (or in near) real-time, where measurements feed back into the analysis immediately.
Reference: Some Pattern Recognition Challenges in Data-Intensive Astronomy