# Pea galaxy

Galaxy Zoo Green Peas
Three Hubble Space Telescope pictures of Green Peas

A Pea galaxy, also referred to as a Pea or Green Pea, might be a type of Luminous Blue Compact Galaxy which is undergoing very high rates of star formation.[1] Pea galaxies are so-named because of their small size and greenish appearance in the images taken by the Sloan Digital Sky Survey (SDSS).

Pea Galaxies were first discovered in 2007 by the volunteer users within the forum section of the online astronomy project Galaxy Zoo (GZ).[2]

## Description

The Pea galaxies, also known as Green Peas (GPs), are compact oxygen-rich emission line galaxies that were discovered at redshifts between z = 0.112 and 0.360.[1] These low-mass galaxies have an upper size limit generally no bigger than 16,300 light-years (5,000 pc) across, and typically they reside in environments less than two-thirds the density of normal galaxy environments.[1] An average GP has a redshift of z = 0.258, a mass of ~3,200 million M (~3,200 million solar masses), a star formation rate of ~10 M/yr (~10 solar masses a year), an [O III] equivalent width of 69.4 nm and a low metallicity.[1][3] A GP is purely star-forming, rather than having an Active Galactic Nucleus. They have a strong emission line at the [OIII] wavelength of 500.7 nm. [OIII], O++ or doubly ionized oxygen, is a forbidden line of the visual spectrum and is only possible at very low densities.[1][4]

Comparing a GP to the Milky Way can be useful when trying to visualize these star-forming rates. An average GP has a mass of ~3,200 million M (~3,200 million solar masses).[1] The Milky Way is a spiral galaxy and has a mass of ~1,125,000, million M (~1,125,000 million solar masses).[5] So the Milky Way has the mass of ~390 GPs. Research has shown that the Milky Way converts ~2 M/yr (~2 solar masses per year) worth of interstellar gas into stars.[6] An average GP converts ~10 M/yr (~10 solar masses) of interstellar gas into stars, which is ~5 times the rate of the Milky Way. If the relative masses of GPs and the Milky Way are taken into account, the average GP converts interstellar gas ~1,950 times more efficiently.

GPs exist at a time when the Universe was three-quarters of its present age and so are clues as to how galaxy formation took place in the earlier Universe.[7] With the publication of Amorin's GTC paper in February 2012, it is now thought that GPs might be old galaxies having formed most of their stellar mass several billion years ago. Old stars have been spectroscopically confirmed in one of the three galaxies in the study by the presence of Magnesium.[8]

"These galaxies would have been normal in the early Universe, but we just don’t see such active galaxies today", said Schawinski. "Understanding the Green Peas may tell us something about how stars were formed in the early Universe and how galaxies evolve".[4]

## History of discovery

The 80 Green Peas

Galaxy Zoo (GZ) is a project online since July 2007 which seeks to classify up to one million galaxies.[9] In July 2007, a few days after the start of GZ, a discussion was started on GZ's Internet forum by Hanny Van Arkel called "Give peas a chance" in which various green objects were posted. This thread started humorously, as the name is a play on words of the title of the John Lennon song "Give Peace a Chance", but by December 2007, it had become clear that some of these unusual objects were a distinct group of galaxies. These "Pea galaxies" appear in the SDSS as unresolved green images. This is because the Peas have a very bright, or powerful, emission line in their spectra for highly-ionized oxygen, which in SDSS color composites increases the luminosity, or brightness, of the "r" color band with respect to the two other color bands "g" and "i". The "r" color band shows as green in SDSS images.[1][10] Enthusiasts, calling themselves the "Peas Corps" (another humorous play on the Peace Corps), collected over a hundred of these Peas, which were eventually placed together into a dedicated thread started by Carolin Cardamone in July 2008. The collection, once refined, provided values that could be used in a systematic computer search of the GZ database of one million objects, which eventually resulted in a sample of 251 Pea galaxies, also known as Green Peas (GPs).

In November 2009, authors C. Cardamone, K. Schawinski, M. Sarzi, S. Bamford, N. Bennert, C. Urry, C. Lintott, W. Keel and 9 others published a paper in the MNRAS titled "Galaxy Zoo Green Peas: Discovery of A Class of Compact Extremely Star-Forming Galaxies".[1] Within this paper, 10 Galaxy Zoo volunteers are acknowledged as having made a particularly significant contribution. They are: Elisabeth Baeten, Gemma Coughlin, Dan Goldstein, Brian Legg, Mark McCallum, Christian Manteuffel, Richard Nowell, Richard Proctor, Alice Sheppard and Hanny Van Arkel. They are thanked for "giving Peas a chance". Citations for 2009MNRAS.399.1191C are available from the SAO/NASA Astrophysics Data System.[11] More details here:Cardamone 2009 Physics

It would be wrong to assume that the 80 GPs were all new discoveries. Out of the 80 original, 46 GPs have previous citations dated before November 2009 in the NASA Extragalactic Database. The original 80 GPs were part of a sample from SDSS data-release 7 (DR7), but did not include galaxies from other sources. Some of these other sources did include objects that might well have been classed as GPs if they were in the SDSS sample. One example of a paper that demonstrates this is: In April 2009, authors J. J. Salzer, A. L. Williams and C. Gronwall published a paper in the Astrophysical Journal Letters titled "A Population of Metal-Poor Galaxies with ~L* Luminosities at Intermediate Redshifts".[12] In this paper, "new spectroscopy and metallicity estimates for a sample of 15 star-forming galaxies with redshifts in the range 0.29 – 0.42" was presented. These objects were selected using the KPNO International Spectroscopic Survey (KISS).[13] Certainly 3 of these 15 when viewed as objects in SDSS are green (KISSR 1516, KISSR 2042 and KISSRx 467). Indeed, quoting from Salzer et al. 2009, section 4.1, it reads "A New Class of Galaxy? Given the large number of studies of metal abundances in galaxies with intermediate and high redshift mentioned in the Introduction, it may seem odd that systems similar to those described here have not been recognized previously."[12]

In June 2010, authors R. Amorin, E. Perez-Montero and J. Vilchez published a paper in The Astrophysical Journal Letters titled "On the oxygen and nitrogen chemical abundances and the evolution of the "green pea" galaxies".[3] In it they explore issues concerning the metallicity of 79 GPs, disputing the original findings in Cardamone et al. They conclude, "arguing that recent interaction-induced inflow of gas, possibly coupled with a selective metal-rich gas loss drive by supernova winds may explain our findings and the known galaxy properties".[3] More details here:Two papers by Amorin

In February 2011, authors Y. Izotov, N. Guseva and T. Thuan published a paper in the Astrophysical Journal titled "Green Pea Galaxies and Cohorts: Luminous Compact Emission-line Galaxies in the Sloan Digital Sky Survey".[14] They argue that GPs are not a rare class of galaxies, but rather a subset of a class known as 'Luminous Compact Galaxies' (LCGs).[14] More details here:Luminous Compact Galaxies

In May 2011, authors R. Amorin, R. Perez-Montero and J.Vilchez published a 'Conference proceeding' titled "Unveiling the Nature of the "Green Pea" galaxies". [15] In this publication, they announce that they have conducted a set of observations using the Optical System for Imaging and low Resolution Integrated Spectroscopy (OSIRIS) at the Gran Telescopio Canarias, and that there is a forthcoming paper about their research. These observations "will provide new insights on the evolutionary state of the Green Peas. In particular, we will be able to see whether the Green Peas show an extended, old stellar population underlying the young starbursts, like those typically dominant in terms of stellar mass in most Blue Compact Galaxies".[15]More details here: Two papers by Amorin

In November 2011, authors Y. Izotov, N. Guseva, K. Fricke and C. Henkel published a paper in Astronomy and Astrophysics titled 'Star-forming galaxies with hot dust emission in the SDSS discovered by the Wide-field Infrared Survey Explorer (WISE)'.[16] In this paper, they find four galaxies that have very red colours in the wavelength range 3.4 micrometres (W1) and 4.6 micrometres (W2). This implies that the dust in these galaxies is at temperatures up to 1000K. These four galaxies are GPs and more than double the number of known galaxies with these characteristics.[16]

In January 2012, authors L. Pilyugin, J. Vilchez, L. Mattsson and T. Thuan published a paper in the MNRAS titled: "Abundance determination from global emission-line SDSS spectra: exploring objects with high N/O ratios".[17] In it they compare the oxygen and nitrogen abundances derived from global emission-line SDSS spectra of galaxies using (1) the electron temperature method and (2) two recent strong line calibrations: the O/N and N/S calibrations.[17] Three sets of objects were compared: composite hydrogen-rich nebula, 281 SDSS galaxies and a sample of GPs with detectable [OIII]-4363 auroral lines.[17] Among the questions surrounding the GPs are how much nebulae influence the spectra, and therefore results, of the GPs. Through comparisons of the three objects using proven methodology and analysis of metallicity, they conclude that "the high nitrogen-to-oxygen ratios derived in some Green Pea galaxies may be caused by the fact that their SDSS spectra are spectra of composite nebulae made up of several components with different physical properties (such as metallicity). However, for the hottest Green Pea galaxies, which appear to be dwarf galaxies, this explanation does not seem to be plausible. It would work only if the HII regions in these galaxies have a dispersion of abundances much larger than that typically found in dwarf galaxies."[17]

In January 2012, author S. Hawley published a paper in the Publications of the Astronomical Society of the Pacific titled "Abundances in "Green Pea" Star-forming Galaxies".[18] In this paper, former NASA astronaut Steve Hawley compares the results from previous GP papers regarding their metallicities. Hawley compares different ways of calibrating and interpreting the various results, mainly from Cardamone et al. and Amorin et al. but some from Izotov et al., and suggests why the various discrepancies between these papers' findings might be. He also considers such details as the contribution of Wolf–Rayet stars to the gas ionization, and which sets of emission lines give the most accurate results for these galaxies. He ends by writing: "The calibrations derived from the Green Peas differ from those commonly utilized and would be useful if star-forming galaxies like the Green Peas with extremely hot ionizing sources are found to be more common."[18]

In February 2012, authors S. Chakraborti, N. Yadav, C. Cardamone and A. Ray published a paper in The Astrophysical Journal Letters titled 'Radio Detection of Green Peas: Implications for Magnetic Fields in Young Galaxies'.[19] In this paper, magnetic studies using new data from the Giant Metrewave Radio Telescope describe various observations based around the GPs. They show that the three "very young" starburst galaxies that were studied have magnetic fields larger than the Milky Way. This is at odds with the current understanding that galaxies build up their magnetic properties over time.[19] More details here:Radio detection

In April 2012, authors R. Amorin, E. Perez-Montero, J. Vilchez and P. Papaderos published a paper in the Astrophysical Journal titled "The Star Formation History and Metal Content of the 'Green Peas'. New Detailed GTC-OSIRIS spectrophotometry of Three Galaxies".[8] They give the results for the deep broad-band imaging and long-slit spectroscopy for 3 GPs that had been observed using the OSIRIS instrument, mounted on the 10.4m Gran Telescopio Canarias at the Roque de los Muchachos Observatory.[8] More details here:GTC-OSIRIS

In August 2012, authors R. Amorín, J. Vílchez, G. Hägele, V. Firpo, E. Pérez-Montero and P. Papaderos published a paper in the Astrophysical Journal Letters titled "Complex gas kinematics in compact, rapidly assembling star-forming galaxies".[20] Using the ISIS spectrograph on the William Herschel Telescope they publish results of the high-quality spectra that they took of six galaxies, five of which are GPs. After studying the Hydrogen alpha emission lines in the spectra of all six, it is shown that these emission lines are made up of multiple lines, meaning that the GPs have several chunks of gas and stars moving at large velocities relative to each other. Indeed, these lines also show that the GPs are effectively a 'turbulent mess', with parts (or clumps) moving at speeds of over 500 km/s (five hundred km/s) relative to each other.[20]

In September 2012, authors S. Parnovsky, I. Izotova and Y. Izotov published a paper in Astrophysics and Space Science titled "H alpha and UV luminosities and star formation rates in a large sample of luminous compact galaxies".[21] In it, they present a statistical study of the star formation rates (SFR) derived from the Galaxy Evolution Explorer observations in the Ultraviolet continuum and in the H alpha emission line for a sample of ~800 luminous compact galaxies (LCGs). GPs are considered as a subset of these LCGs (See Izotov et al. 2011 below). Within the larger set of LCGs, including the GPs, SFR of up to ~110 M/yr (~110 solar masses a year) are found.[21]

In April 2013, authors A. Jaskot and M. Oey published a paper in the Astrophysical Journal titled "The Origin and Optical Depth of Ionizing Radiation in the "Green Pea" Galaxies".[22] Six "extreme" GPs. are studied. Using these, they endeavour to narrow down the list of possibilities about what is producing the radiation and the substantial amounts of high-energy photons that might be escaping from the GPs.[22] Following on from this paper, observations on the Hubble Space Telescope, totalling 24 orbits, were taken in December 2013.[23] The Cosmic Origins Spectrograph and the Advanced Camera for Surveys were used on four of the "extreme" GPs. More details here:Two papers by Jaskot and Oey

In January 2014, authors Y. Izotov, N. Guseva, K. Fricke and C. Henkel published a paper in Astronomy & Astrophysics entitled "Multi-wavelength study of 14000 star-forming galaxies from the Sloan Digital Sky Survey".[24] In it, they use a variety of sources to demonstrate "that the emission emerging from young star-forming regions is the dominant dust-heating source for temperatures to several hundred degrees in the sample star-forming galaxies".[24] The first source of data is SDSS from which 14,610 spectra with strong emission lines are selected. These 14,610 spectra were then cross-identified with sources from photometric sky surveys in other wavelength ranges. Those are: 1)GALEX for the ultraviolet; 2)the 2MASS survey for the near-infrared; 3)the Wide-field Infrared Survey Explorer All-Sky Source Catalog for infrared at differing wavelengths; 4)the IRAS survey for the far-infrared and the 5)NVSS Survey at radio-wavelengths. Only a small fraction of the SDSS objects were detected in the last two surveys. Among the results is a list of twenty galaxies with the highest magnitudes which have hot dust of several hundred degrees. Of these twenty,all could be classified as GPs and/or LCGs.[24] Also among the results, the luminosity is obtained in the sample galaxies in a wide wavelength range. At the highest luminosities, the sample galaxies had luminosites approaching those of high-redshift Lyman-break galaxies.[24]

In February 2014, authors A. Jaskot and M. Oey published a conference report titled "The Origin and Optical Depth of Ionizing Photons in the Green Pea Galaxies".[25] This will appear in "Massive Young Star Clusters Near and Far: From the Milky Way to Reionization", based on the 2013 Guillermo Haro Conference. More details here:Two papers by Jaskot and Oey

## Physics from the Cardamone 2009 paper

Graph showing specific star formation rate plotted against galaxy mass, with the GPs (purple diamonds) and the Galaxy Zoo Merger Sample (black points)

At the time this paper was published, only five Green Peas (GPs) had been imaged by the Hubble Space Telescope (HST). Three of these images reveal GPs to be made up of bright clumps of star formation and low surface density features indicative of recent or ongoing galaxy mergers.[1] These three HST images were imaged as part of a study of local ultraviolet (UV-luminous) galaxies in 2005.[26] Major mergers are frequently sites of active star-formation and to the right a graph is shown that plots specific star formation rate (SFR / Galaxy Mass) against galaxy mass.[27] In this graph, the GPs are compared to the 3003 mergers from the Galaxy Zoo Merger Sample (GZMS).[28] It shows that the GPs have low masses typical of dwarf galaxies and much higher star-forming rates (SFR) when compared to the GZMS. The black, dashed line shows a constant SFR of 10 M/yr (~10 solar masses). Most GPs have a SFR between 3 and 30 M/yr (between ~3 and ~30 solar masses).

Graph showing 103 GPs plotted as Starburst galaxies (red stars), transition objects (green crosses) or A.G.N. (blue diamonds)

GPs are rare. Of the one million objects that make up GZ's image bank, only 251 GPs were found. After having to discard 148 of these 251 because of atmospheric contamination of their spectra, the 103 that were left, with the highest signal-to-noise ratio, were analyzed further using the classic emission line diagnostic by Baldwin, Phillips and Terlevich which separates starbursts and active galactic nuclei.[29] 80 were found to be starburst galaxies.[1] The graph to the left classifies 103 narrow-line GPs (all with SNR ≥ 3 in the emission lines) as 10 active galactic nuclei (blue diamonds), 13 transition objects (green crosses) and 80 starbursts (red stars). The solid line is: Kewley et al. (2001) maximal starburst contribution (labelled Ke01).[30][31] The dashed line is: Kauffmann et al. (2003) separating purely star-forming objects from AGN (labelled Ka03).[32]

Histogram showing [OIII] Eq.Wth. of 10,000 comparison galaxies (red); 215 UV-luminous Galaxies (blue); GPs (green)

GPs have a strong [OIII] emission line when compared to the rest of their spectral continuum. In a SDSS spectrum, this shows up as a large peak with [OIII] at the top.[33] The wavelength of [OIII] (500.7 nm) was chosen to determine the luminosities of the GPs using Equivalent width (Eq.Wth.). The histogram on the right shows on the horizontal scale the Eq.Wth. of a comparison of 10,000 normal galaxies (marked red), UV-luminous Galaxies (marked blue) and GPs (marked green).[1] As can be seen from the histogram, the Eq.Wth. of the GPs is much larger than normal for even prolific starburst galaxies such as UV-luminous Galaxies.[34]

Within the Cardamone et al. paper, comparisons are made with other compact galaxies, namely Blue Compact Dwarfs Galaxies and UV-luminous Galaxies, at local and much higher distances.[35] The findings show that GPs form a different class of galaxies than Ultra Blue Compact Dwarfs, but may be similar to the most luminous members of the Blue Compact Dwarf Galaxy category.[36] The GPs are also similar to UV-luminous high redshift galaxies such as Lyman-break Galaxies and Lyman-alpha emitters.[37][38][39] It is concluded that if the underlying processes occurring in the GPs are similar to that found in the UV-luminous high redshift galaxies, the GPs may be the last remnants of a mode of star formation common in the early Universe.[1][40][41]

Histogram showing reddening values for GPs

GPs have low interstellar reddening values, as shown in the histogram on the right, with nearly all GPs having E(B-V) ≤ 0.25. The distribution shown indicates that the line-emitting regions of star-forming GPs are not highly reddened, particularly when compared to more typical star-forming or starburst galaxies.[1] This low reddening combined with very high UV luminosity is rare in galaxies in the local Universe and is more typically found in galaxies at higher redshifts.[42]

Cardamone et al. describe GPs as having a low metallicity, but that the oxygen present is highly ionized. The average GP has a metallicity of log[O/H]+12~8.69, which is solar or sub-solar, depending on which set of standard values is used.[1][43][44][45][46] Although the GPs are in general consistent with the mass-metallicity relation, they depart from it at the highest mass end and thus do not follow the trend. GPs have a range of masses, but a more uniform metallicity than the sample compared against.[47] These metallicities are common in low mass galaxies such as Peas.[1]

An example of a GP spectrum made using GANDALF.

As well as the optical images from the SDSS, measurements from the GALEX survey were used to determine the ultraviolet values.[48] This survey is well matched in depth and area, and 139 of the sampled 251 GPs are found in GALEX Release 4 (G.R.4).[49] For the 56 of the 80 star-forming GPs with GALEX detections, the median luminosity is ~30,000 million $L_{\odot}$ (~30,000 million solar luminosities).

When compiling the Cardamone paper, spectral classification was made using Gas And Absorption Line Fitting (GANDALF).[1] This sophisticated computer software was programmed by Marc Sarzi, who helped analyze the SDSS spectra.[50]

## Analysis of the Cardamone 2009 paper

These values are from Table 4, pages 16–17 of Cardamone 2009 et al., which shows the 80 GPs that have been analyzed here.[1] The long 18-digit numbers are the SDSS reference numbers, which link to the appropriate entry at the SDSS Skyserver website.

r-i vs g-r color-color diagram for 251 GPs (green crosses), a sample of normal galaxies (red points) and all quasars (purple points)
Greatest Least Average Nearest to Average
Distance z=0.348
(587732134315425958)
z=0.141
(587738947196944678)
z=0.2583 z=0.261
(587724240158589061)
Mass 1010.48 M
(588023240745943289)
108.55 M
(587741392649781464)
109.48 M 109.48 M
(587724241767825591)
Rate of star-forming 59 M/yr
(587728906099687546)
2 M/yr
(588018090541842668)
13.02 M/yr 13 M/yr
(588011122502336742)
Luminosity ([OIII] Eq.Wth.) 238.83 nm
(587738410863493299)
1.2 nm
(587741391573287017)
69.4 nm 67.4 nm
(588018090541842668)
Luminosity (UV) 36.1×1036 W
(587733080270569500)
1.9×1036 W
(588848899919446344)
12.36×1036 W 12.3×1036 W
(588018055652769997)

Color selection was by using the difference in the levels of three filters, in order to capture these color limits: u-r ≤ 2.5 (1), r-i ≤ -0.2 (2), r-z ≤ 0.5 (3), g-r ≥ r-i + 0.5 (4), u-r ≥ 2.5 (r-z) (5).[1] If the diagram on the right (one of two in the paper) is looked at, the effectiveness of this color selection can be seen. The color-color diagram shows ~100 GPs (green crosses), 10,000 comparison galaxies (red points) and 9,500 comparison quasars (purple stars) at similar redshifts to the GPs. The black lines show how these figures are on the diagram.

One of the original ways of recognizing GPs, before SQL programming was involved, was because of a discrepancy about how the SDSS labels them within Skyserver.[51] Out of the 251 of the original GP sample that were identified by the SDSS spectroscopic pipeline as having galaxy spectra, only 7 were targeted by the SDSS spectral fibre allocation as galaxies i.e. 244 were not.[1][52]

## Two Papers by R.Amorin, J.M.Vilchez and E.Perez-Montero

In June 2010, authors R. Amorín, E. Pérez-Montero and J.M. Vílchez published a paper in The Astrophysical Journal letters titled "On the Oxygen and Nitrogen Chemical Abundances and the Evolution of the "Green Pea" Galaxies", which disputes the metallicities calculated in the original Cardamone et al. GPs paper [1][3] Amorin et al. use a different methodology from Cardamone et al. to produce metallicity values more than one fifth (20%) of the previous values (about 20% solar or one fifth solar) for the 80 'starburst' GPs. These mean values are log[O/H]+12~8.05, which shows a clear offset of 0.65dex between the two papers' values. For these 80 GPs, Amorin et al., using a direct method, rather than strong-line methods as used in Cardamone et al., calculate physical properties, as well as oxygen and nitrogen ionic abundances.[53] These metals pollute hydrogen and helium, which make up the majority of the substances present in galaxies. As these metals are produced in Supernovae, the older a galaxy is, the more metals it would have. As GPs are in the nearby, or older, Universe, they should have more metals than galaxies at an earlier time.

N/O vs. O/H abundance ratio

Amorin et al. find that the amount of metals, including the abundance of nitrogen, are different from normal values and that GPs are not consistent with the mass-metallicity relation, as concluded by Cardamone et al.[1][54] This analysis indicates that GPs can be considered as genuine metal-poor galaxies. They then argue that this oxygen under-abundance is due to a recent interaction-induced inflow of gas, possibly coupled with a selective metal-rich gas loss driven by Supernovae winds and that this can explain their findings.[47][55] This further suggests that GPs are likely very short-lived as the intense star formation in them would quickly enrich the gas.[3]

O/H vs. stellar mass

In May 2011, R.Amorin, J.M.Vilchez and E.Perez-Montero published a conference proceeding paper titled "Unveiling the Nature of the "Green Pea" galaxies".[15] In it they review recent scientific results and announcing a forthcoming paper on their recent observations at the Gran Telescopio Canarias.[15] This paper is also a modified report of a presentation at the Joint European and National Astronomy Meeting (JENAM) 2010.[56] They conclude that GPs are a genuine population of metal-poor, luminous and very compact starburst galaxies. Amongst the data, five graphs illustrate the findings they have made. Amorin et al. use masses calculated by Izotov, rather than by Cardamone.[3][14] The metallicities that Amorin et al. use agree with Izotov's findings, or vice-versa, rather than Cardamone's.[3][14]

The first graph (on the left; fig.1 in paper) plots the nitrogen/oxygen vs. oxygen/hydrogen abundance ratio. The 2D histogram of SDSS star forming galaxies is shown in logarithmic scale while the GPs are indicated by circles. This shows that GPs are metal-poor.

N/O vs. stellar mass

The second graph (on the right; fig.2 in paper) plots O/H vs. stellar mass. The 2D histogram of SDSS SFGs is shown in logarithmic scale and their best likelihood fit is shown by a black solid line. The subset of 62 GPs are indicated by circles and their best linear fit is shown by a dashed line. For comparison we also show the quadratic fit presented in Amorin et al. 2010 for the full sample of 80 GPs. SFGs at z ≥ 2 by Erb et al. are also shown by asterisks for comparison.[3][57]

O/H vs. B-band (rest-frame) absolute magnitude

The third graph (on the left; fig.3 in paper) plots N/O vs stellar mass. Symbols as in fig.1.

Gas Mass Fraction v. Metallicity

The fourth graph (on the right; fig.4 in paper) plots O/H vs. B-band (rest-frame) absolute magnitude. The meaning of symbols is indicated. Distances used in computing (extinction corrected) absolute magnitudes were, in all cases, calculated using spectroscopic redshifts and the same cosmological parameters. The dashed line indicates the fit to the HII galaxies in the Luminosity-Metallicity Relationship (MZR) given by Lee et al. 2004.[58]

The fifth graph (on the left; fig.5 in paper) plots gas mass fraction vs. metallicity. Different lines correspond to closed-box models at different yields, as indicated in the legend. Open and filled circles are GPs which are above and below the fit to their MZR. Diamonds are values for the same Wolf-Rayet galaxies as in Fig. 4.[3]

## GTC-OSIRIS Spectrophotometry

In February 2012, authors R. Amorin, E. Perez-Montero, J. Vilchez and P. Papaderos published a paper titled "The star formation history and metal content of the "Green Peas". New detailed GTC-OSIRIS spectrophotometry of three galaxies" in which they presented the findings of observations carried out using the Gran Telescopio Canarias at the Roque de los Muchachos Observatory. They gather deep broad-band imaging and long-slit spectroscopy of three GPs using high precision equipment.[8]

Their findings show that the three GPs display relatively low Extinction (astronomy), low oxygen abundances and high nitrogen-to-oxygen ratios.[8] Also reported are the clear signatures of Wolf–Rayet stars, of which a population are found (between ~800 and ~1200).[8] A combination of population and evolutionary synthesis models strongly suggest a formation history dominated by starbursts.[8] These models show that these three GPs currently undergo a major starburst producing between ~4% and ~20% of their stellar mass. However as these models imply, they are old galaxies having formed most of their stellar mass several billion (gigayear) ago.[8] The presence of old stars has been spectroscopically verified in one of the three galaxies by the detection of Magnesium.[8] Surface photometry, using data from the Hubble Space Telescope archive, indicates that the three GPs possess an exponential low surface brightness envelope (see Low surface brightness galaxy).[8] This suggests that GPs are identifiable with major episodes in the assembly history of local Blue Compact Dwarf galaxies.[8]

The three galaxies are (using SDSS references):[8]

## Comparison of Green Peas to Luminous Compact Galaxies

In February 2011, Yuri Izotov, Natalia Guseva and Trinh Thuan published a paper titled "Green Pea Galaxies and Cohorts: Luminous Compact Emission-line Galaxies in the Sloan Digital Sky Survey", examining the GPs and comparing these to a larger set of 803 Luminous Compact Galaxies (LCGs).[14] They use a different set of selection criteria from Cardamone et al. These are: a) a high extinction-corrected luminosity > 3x10^40 Ergs s^-1 of the hydrogen beta emission line; (see Hydrogen spectral series) b) a high equivalent width greater than 5 nm; c) a strong [OIII] wavelength at the 436.3 nm emission line allowing accurate abundance determination; d) a compact structure on SDSS images; and e) an absence of obvious active galactic nuclei spectroscopic features.[14]

Its conclusions (shortened) are:

1. The selected galaxies have redshifts between 0.02 and 0.63, a range equal or greater than a factor of 2 when compared to the GPs. They find the properties of LCGs and GPs are similar to Blue Compact Dwarf galaxies. Explaining how the colours of emission-line galaxies change with distance using SDSS, they conclude that GPs are just subsamples within a narrow redshift range of their larger LCG sample.[14]
2. Although there were no upper limits on the Hydrogen beta luminosities, it was found that there was a 'self-regulating' mechanism which bound the LCGs to a limit of ~3x10^42 Ergs s^-1.[14]
3. In the [OIII] wavelength 500.7 nm ratio to hydrogen beta vs [NII] wavelength 658.3 nm ratio to hydrogen alpha, LCGs occupy the region, in the diagnostic diagram, of star-forming galaxies with the highest excitation. However, some active galactic nuclei also lie in this region in the diagnostic diagram.[14]
4. The oxygen abundances 12 + log O/H in LCGs are in the range 7.6-8.4 with a median value of ~8.11, confirming Amorin et al.'s analysis of a subset of GPs.[3][14] This range of oxygen abundances is typical of nearby lower-luminosity Blue Compact Dwarfs. These results show that the original Cardamone et al. median oxygen abundance of 12 + log O/H = ~8.7 is overestimated, as a different, Empirical evidence method was originally used, rather than the direct method by Amorin et al. and Izotov et al.[1] There is no dependence of oxygen abundance on redshift.
5. In the luminosity-metallicity diagram (fig. 8 in paper), LCGs are shifted by ~2 magnitudes brighter when compared to nearby emission-line galaxies. LCGs form a common luminosity-metallicity relation, as for the most actively star-forming galaxies. Some LCGs have oxygen abundances and luminosities similar to Lyman-break galaxies (LBGs), despite much lower redshifts, thus enabling the study of LBGs through LCGs.[14]

## Radio detection of Green Peas

In February 2012, authors Sayan Chakraborti, Naveen Yadav, Alak Ray and Carolin Cardamone published a paper titled "Radio Detection of Green Peas: Implications for Magnetic Fields in Young Galaxies" which deals with the magnetic properties of the GPs.[19] In it, they describe observations which have produced some unexpected results raising puzzling questions about the origin and evolution of magnetic fields in young galaxies.[19] The ages are estimated from looking at the star formation that the GPs currently have ongoing and then estimating the age of the most recent starburst. GPs are very young galaxies, with models of the observed stellar populations indicating that they are around 10^8 (one hundred million) years old (1/100th the age of the Milky Way).[19] There is some question as to whether the GPs all started from the same starburst or if multiple starbursts went on (much older stellar populations are hidden as we can't see the light from these).

Using data from the Giant Metrewave Radio Telescope (GMRT) and archive observations from the Very Large Array (VLA), Chakraborti et al. produced a set of results which are based around the VLA FIRST detection of stacked flux from 32 GPs and three 3-hour low frequency observations from the GMRT which targeted the three most promising candidates which had expected fluxes at the milli-Jansky (mJy) level.

Chakraborti et al. find that the three GPs observed by the GMRT have a magnetic field of B~39 μG, and more generally a figure of greater than B~30μG for all the GPs. This is compared to a figure of B~5μG for the Milky Way.[19] The present understanding is of magnetic field growth based on the amplification of seed fields by dynamo theory and its action over a galaxy's lifetime.[19] The observations of GPs challenge that thinking.

Given the high star-forming rates of GPs generally, they are expected to host a large number of Supernovae. Supernovae accelerate electrons to high energies, near to the speed of light, which may then emit synchrotron radiation in radio band frequencies.

## Two papers by A. Jaskot and M.S. Oey

In April 2013, authors A. Jaskot and M. Oey published a paper in the Astrophysical Journal titled "The Origin and Optical Depth of Ionizing Radiation in the "Green Pea" Galaxies".[22] Six "extreme" GPs are studied.

Six Extreme GPs

Using these, they endeavour to narrow down the list of possibilities about what is producing the UV-radiation and the substantial amounts of high-energy photons that might be escaping from the GPs.[22] Through trying to observe these photons in nearby galaxies such as the GPs, our understanding of how galaxies behaved in the early Universe might well be revolutionised. It is reported that the GPs are exciting candidates to help astronomers understand a major milestone in the development of the cosmos 13 billion years ago, during the epoch of reionization.[59]

Following on from the above Jaskot and Oey paper, observations on the Hubble Space Telescope, totalling 24 orbits, were taken in December 2013.[23] The Cosmic Origins Spectrograph and the Advanced Camera for Surveys were used on four of the "extreme" GPs. The abstract for HST Observing Program 13293 states that "These observations will either reveal the Green Peas as a class of galaxies having substantial LyC escape fractions or demonstrate that even some of the most extreme galaxies in the nearby Universe are optically thick."[23]

In February 2014, authors A. Jaskot and M. Oey published a conference report titled "The Origin and Optical Depth of Ionizing Photons in the Green Pea Galaxies".[25] This will appear in "Massive Young Star Clusters Near and Far: From the Milky Way to Reionization", based on the 2013 Guillermo Haro Conference. In the publication, Jaskot and Oey write: "We are currently analyzing observations from IMACS and MagE on the Magellan telescopes and COS and ACS on Hubble Space Telescope (HST) to distinguish between WR (Wolf-Rayet star) and the shock ionization scenarios and confirm the GPs’ optical depths.[25] The absence of WR features in the deeper IMACS spectra tentatively supports the shock scenario, although the detection limits do not yet definitively rule out the WR photoionization hypothesis.[25] The MagE spectra will settle the question of whether WR stars are present and reveal weak shock lines, if they exist.[25] Our latest HST observations demonstrate that the extreme GPs are Lyman-alpha emitters (LAEs), with the strongest Lya emission present in objects that lack absorption lines from the neutral interstellar medium (ISM), such as Carbon II (wavelength133.5nm).[23][25] The absence of strong C II absorption implies that these GPs may be optically thin along our line of sight. The likelihood of shock ionization in the GPs combined with the Lya emission suggests a possible explanation for low optical depths in the GPs. Supernova feedback 3-5 Myr after a burst may create holes in the ISM, allowing LyC photons from the remaining massive stars to escape. Supernova-driven outflows should likewise enhance Lya escape, consistent with our HST observations."[25]

## References

1. C. Cardamone et al. (2009). "Galaxy Zoo Green Peas: Discovery of A Class of Compact Extremely Star-Forming Galaxies". Monthly Notices of the Royal Astronomical Society 399 (3): 1191. arXiv:0907.4155. Bibcode:2009MNRAS.399.1191C. doi:10.1111/j.1365-2966.2009.15383.x.
2. ^ M. Jordan Raddick et al. (2010). "Galaxy Zoo:Exploring the motivations of citizen science volunteers". Astronomy Education Review 9 (1): 010103. arXiv:0909.2925. Bibcode:2010AEdRv...9a0103R. doi:10.3847/AER2009036.
3. R. O. Amorín, E. Pérez-Montero, J.M. Vílchez. "On the oxygen and nitrogen chemical abundances and the evolution of the "green pea" galaxies". arXiv:1004.4910 [astro-ph.CO]. Bibcode arXiv:1004.4910v2.
4. ^ a b "Galaxy Zoo Hunters Help Astronomers Discover Rare 'Green Pea' Galaxies". Yale Bulletin. July 27, 2009. Retrieved 2009-12-29.
5. ^ Paul J.McMillan. "Mass models of the Milky Way". arXiv:arXiv:1102.4340v1.
6. ^ Laura Chomiuk, Matthew S. Povich (October 18, 2011). "Toward a Unification of Star Formation Rate Determinations in the Milky Way and Other Galaxies". arXiv:arXiv:1110.4105v1.
7. ^ "New Image of Infant Universe reveals era of first stars, age of cosmos, and more". NASA. 2003. Retrieved 2010-01-16.
8. R. Amorin, E. Perez-Montero, J.M. Vilchez, P.Papaderos. "The star formation history and metal content of the 'Green Peas'.New detailed GTC_OSIRIS spectrophotometry of three galaxies". arXiv:1202.3419 [astro-ph.CO].
9. ^ C. Lintott, K. Schawinski, S. Bamford, A. Slosar,. "Galaxy Zoo 1 : Data Release of Morphological Classifications for nearly 900,000 galaxies". arXiv:1007.3265 [astro-ph.CO]. Bibcode arXiv:1007.3265v4.
10. ^ "SDSS Color". Sloan Digital Sky Survey. Retrieved 2010-01-23.
11. ^ "SAO/NASA Astrophysics Data System (ADS)". Harvard-Smithsonian Center for Astrophysics. Retrieved 2014-03-13.
12. ^ a b John J. Salzer, Anna L.Williams, & Caryl Gronwall. "A Population of Metal-Poor Galaxies with ~L* Luminosities at Intermediate Redshifts". arXiv:0903.3948 [astr-ph.CO]. Bibcode arXiv:0903.3948v1.
13. ^ J. J. Salzer, C. Gronwall, V. A. Lipovetsky, A. Kniazev, J. W. Moody, T. A. Boroson, T. X. Thuan, Y. I. Izotov, J. L. Herrero, L. M. Frattare. "The KPNO International Spectroscopic Survey. I. Description of the Survey". arXiv:arXiv:astro-ph/0004074 [astr-ph.CO]. Bibcode arXiv:astro-ph/0004074v2.
14. Y. Izotov, N. Guseva, T. Thuan (2010). "Green Pea Galaxies and cohorts: Luminous Compact Emission-Line Galaxies in the Sloan Digital Sky Survey". arXiv:1012.5639 [astro-ph.CO]. Bibcode arXiv:1012.5639v1.
15. ^ a b c d R. Amorin, R. Perez-Montero and J.Vilchez (2011). "Unveiling the Nature of the "Green Pea" galaxies". arXiv:1105.1477 [astro-ph.CO].
16. ^ a b Y.I. Izotov, N.G. Guseva, K.J. Fricke and C. Henkel. "Star-forming galaxies with hot dust emission in the Sloan Digital Sky Survey discovered by the Wide-field Infrared Survey Explorer (WISE)". arXiv:1111.5450 [astro-ph.CO].
17. ^ a b c d L.S. Pilyugin, J.M. Vilchez, L. Mattsson, T.X. Thuan. "Abundance determination from global emission-line SDSS spectra: exploring objects with high N/O ratios". arXiv:1201.1554 [astro-ph.CO].
18. ^ a b S.A. Hawley (2012). "Abundances in Green Pea Star-forming Galaxies". Publications of the Astronomical Society of the Pacific. Bibcode:2012PASP..124...21H. doi:10.1086/663866.
19. S. Chakraborti, N. Yadav, C. Cardamone and A. Ray (2011). "Radio Detection of Green Peas: Implications for Magnetic Fields in Young Galaxies". arXiv:1110.3312 [astro-ph.CO]. Bibcode arXiv:1110.3312v1.
20. ^ a b Ricardo Amorín, José M. Vílchez, Guillermo Hägele, Verónica Firpo, Enrique Pérez-Montero and Polychronis Papaderos. "Complex gas kinematics in compact, rapidly assembling star-forming galaxies". arXiv:1207.0509 [astro-ph.CO]. Bibcode 2012ApJ...754L..22A.
21. ^ a b S. Parnovsky, I. Izotova and Y. Izotov. "H alpha and UV luminosities and star formation rates in a large sample of luminous compact galaxies". arXiv:1209.3498 [astro-ph.CO]. Bibcode 2012arXiv1209.3498P.
22. ^ a b c d A. Jaskot and M. Oey. "The Origin and Optical Depth of Ionizing Radiation in the "Green Pea" Galaxies". arXiv:1301.0530 [astro-ph.CO]. Bibcode arXiv:1301.0530v1.
23. ^ a b c d "HST OBSERVING PROGRAM 13293". Space Telescope Science Institute. Retrieved 2014-03-13.
24. ^ a b c d Y. I. Izotov, N. G. Guseva, K. J. Fricke and C. Henkel. "Multi-wavelength study of 14000 star-forming galaxies from the Sloan Digital Sky Survey". arXiv:1310.1559 [astro-ph.CO]. Bibcode arXiv:1310.1559v1.
25. A. Jaskot, M. Oey. "The Origin and Optical Depth of Ionizing Photons in the Green Pea Galaxies". arXiv:1402.4429 [astro-ph.CO]. Bibcode arXiv:1402.4429v1.
26. ^ T. Heckman et al. (2005). "The Properties of Ultraviolet-Luminous Galaxies at the Current Epoch". Astrophysical Journal 619 (1): L35. arXiv:astro-ph/0412577. Bibcode:2005ApJ...619L..35H. doi:10.1086/425979.
27. ^ A. Bauer, N. Drory, G. Hill, G. Feulner (2005). "Specific Star Formation Rates to Redshift 1.5". Astrophysical Journal 621 (2): L89. arXiv:astro-ph/0412358. Bibcode:2005ApJ...621L..89B. doi:10.1086/429289.
28. ^ D. Darg et al. (2010). "Galaxy Zoo: The fraction of merging galaxies in the SDSS and their morphologies". Monthly Notices of the Royal Astronomical Society 401 (2): 1043. arXiv:0903.4937. Bibcode:2010MNRAS.401.1043D. doi:10.1111/j.1365-2966.2009.15686.x.
29. ^ J. Baldwin, M. Phillips, R. Terlevich (1981). "Classification parameters for the emission-line spectra of extragalactic objects". Publications of the Astronomical Society of the Pacific 93 (551): 5. Bibcode:1981PASP...93....5B. doi:10.1086/130766.
30. ^ L. Kewley et al. (2001). "Theoretical Modeling of Starburst Galaxies". Astrophysical Journal 556 (1): 121. arXiv:astro-ph/0106324. Bibcode:2001ApJ...556..121K. doi:10.1086/321545.
31. ^ B. Groves, L. Kewley (2008). "Distinguishing Active Galactic Nuclei and Star Formation". In J.H. Knapen, T.J. Mahoney, A. Vazdekis. Pathways through an Eclectic Universe. ASP Conference Series 390. p. 283. arXiv:0707.0158. Bibcode:2008ASPC..390..283G.
32. ^ G. Kauffmann et al. (2003). "Stellar Masses and Star Formation Histories". Monthly Notices of the Royal Astronomical Society 341 (1): 33. arXiv:astro-ph/0204055. Bibcode:2003MNRAS.341...33K. doi:10.1046/j.1365-8711.2003.06291.x.
33. ^ "SDSS Spectra". Sloan Digital Sky Survey. Retrieved 2010-01-17.
34. ^ C. Hoopes et al. (2007). "The Diverse Properties of the Most Ultraviolet-Luminous Galaxies Discovered by GALEX". Astrophysical Journal 173 (2): 441. arXiv:astro-ph/0609415. Bibcode:2007ApJS..173..441H. doi:10.1086/516644.
35. ^ O. Vaduvescu, M. McCall, M. Richer (2007). "Chemical Properties of Star-Forming Dwarf Galaxies". Astronomical Journal 134 (2): 604. arXiv:0704.2705. Bibcode:2007AJ....134..604V. doi:10.1086/518865.
36. ^ M. Corbin et al. (2006). "Ultracompact Blue Dwarf Galaxies: HST Imaging and Stellar Population Analysis". Astrophysical Journal 651 (2): 861. arXiv:astro-ph/0607280. Bibcode:2006ApJ...651..861C. doi:10.1086/507575.
37. ^ M. Bremer et al. (2004). "The Properties of Galaxies at z~5". Monthly Notices of the Royal Astronomical Society 347 (1): L7. arXiv:astro-ph/0306587. Bibcode:2004MNRAS.347L...7B. doi:10.1111/j.1365-2966.2004.07352.x.
38. ^ C. Gronwall et al. (2007). "Lyα Emission-Line Galaxies at z = 3.1 in the Extended Chandra Deep Field-South". Astrophysical Journal 667 (1): 79. arXiv:0705.3917. Bibcode:2007ApJ...667...79G. doi:10.1086/520324.
39. ^ L. Pentericii et al. (2009). "The physical properties of Lyα emitting galaxies: not just primeval galaxies?". Astronomy & Astrophysics 494 (2): 553. arXiv:0811.1861. Bibcode:2009A&A...494..553P. doi:10.1051/0004-6361:200810722.
40. ^ E. Gawiser et al. (2007). "Lyα-Emitting Galaxies at z = 3.1: Progenitors Experiencing Rapid Star Formation". Astrophysical Journal 671 (1): 278. arXiv:0710.2697. Bibcode:2007ApJ...671..278G. doi:10.1086/522955.
41. ^ M. Giavalisco et al. (2004). "The Rest-Frame Ultraviolet Luminosity Density of Star-forming Galaxies at Redshifts z > 3.51". Astrophysical Journal 600 (2): L103. arXiv:astro-ph/0309065. Bibcode:2004ApJ...600L.103G. doi:10.1086/381244.
42. ^ K. Masters et al. (2010). "Galaxy Zoo: Dust in Spirals". Monthly Notices of the Royal Astronomical Society 404 (2): 792. arXiv:1001.1744. Bibcode:2010MNRAS.404..792M. doi:10.1111/j.1365-2966.2010.16335.x.
43. ^ N. Grevesse, A. Sauval (1998). "Standard Solar Composition". Space Science Reviews 85 (1/2): 161. Bibcode:1998SSRv...85..161G. doi:10.1023/A:1005161325181.
44. ^ C. Allende Prieto, D.L. Lambert, M. Asplund (2001). "The Forbidden Abundance of Oxygen in the Sun". Astrophysical Journal 556 (1): L63. arXiv:astro-ph/0106360. Bibcode:2001ApJ...556L..63A. doi:10.1086/322874.
45. ^ M. Asplund, N. Grevesse, A.J. Sauval (2005). F.N. Bash & T.G Barnes, ed. "Cosmic abundances as records of stellar evolution and nucleosynthesis". ASP Conference Series 336. p. 25. arXiv:astro-ph/0410214. Bibcode:2005ASPC..336...25A.
46. ^ S. Basu, H.M. Antia (2007). "Helioseismology and Solar Abundances". Physics Reports 457 (5-6): 217. arXiv:0711.4590. Bibcode:2008PhR...457..217B. doi:10.1016/j.physrep.2007.12.002.
47. ^ a b C. Tremonti et al. (2004). "The Origin of the Mass-Metallicity Relation: Insights from 53,000 Star-forming Galaxies in the Sloan Digital Sky Survey". Astrophysical Journal 613 (2): 898. arXiv:astro-ph/0405537. Bibcode:2004ApJ...613..898T. doi:10.1086/423264.
48. ^ "GALEX Observes the Universe". NASA. 2003. Retrieved 2010-01-16.
49. ^ P. Morrissey et al. (2007). "The Calibration and Data Products of GALEX". Astrophysical Journal Supplement 173 (2): 682. Bibcode:2007ApJS..173..682M. doi:10.1086/520512.
50. ^
51. ^ "SDSS Skyserver". Sloan Digital Sky Survey. Retrieved 2010-01-17.
52. ^ C. Stoughton et al. (2002). "Sloan Digital Sky Survey: Early Data Release". Astronomical Journal 123 (1): 485. Bibcode:2002AJ....123..485S. doi:10.1086/324741.
53. ^ E. Pérez-Montero, T. Contini (2009). "The impact of the nitrogen-to-oxygen ratio on ionized nebulae diagnostics based on [NII] emissionlines". Monthly Notices of the Royal Astronomical Society 398 (2): 949. arXiv:0905.4621. Bibcode:2009MNRAS.398..949P. doi:10.1111/j.1365-2966.2009.15145.x.
54. ^ J. Lequeux et al. (1979). "Chemical composition and evolution of irregular and blue compact galaxies". Astronomy and Astrophysics 80 (2): 155–166. Bibcode:1979A&A....80..155L.
55. ^ F. Kristian, R. Davé (2008). "The Origin of the Galaxy Mass-Metallicity Relation and Implications for Galactic Outflows". Monthly Notices of the Royal Astronomical Society 385 (4): 2181. arXiv:0704.3100. Bibcode:2008MNRAS.385.2181F. doi:10.1111/j.1365-2966.2008.12991.x.
56. ^ "JENAM_2010". Retrieved 2011-06-21.
57. ^ D.K. Erb, A.E. Shapley, M. Pettini, C.C. Steidel, N.A. Reddy, K.L. Adelberger. (2006). "The Mass/Metallicity Relation at z=2". Astrophysical Journal. 644, 813, 2006. arXiv:astro-ph/0602473. Bibcode:2006ApJ...644..813E. doi:10.1086/503623.
58. ^ J.C. Lee, J.J. Salzer, J. Melbourne (2004). "Metal Abundances of KISS Galaxies. III. Nebular Abundances for Fourteen Galaxies and the Luminosity-Metallicity Relationship for H II Galaxies". Astrophysical Journal 616: 752–767. arXiv:astro-ph/0408342. Bibcode:2004ApJ...616..752L. doi:10.1086/425156.
59. ^ "Michigan_Uni_PR". University of Michigan. Retrieved 2014-03-27.