Peas in the Universe, Goodwill and a History of Zooite Collaboration on the Peas Project

Warning: This “History of the Peas” is rather long. At Carie’s request, Rick wrote a shorter version here.

The SDSS telescope has five colour filters, one of which is green. Like a rainbow played backwards as it splits in a prism, the colours from all filters are shown to us all at once, so we see them mixed and averaged out – usually twinkling blue star formation, golden ellipticals, and red faraway objects or nearby stars. When an object moves relative to Earth while the SDSS telescope images it, sometimes only gets through the green filter at one given time and thus leaves a pure green image in our pictures – which is usually the case with a camera glitch, one of the three images of asteroids, or satellite trails.

Some objects, though, seem to be green in their own right. We were all so busy in the first month of Galaxy Zoo trying to work out what pretty much anything was, and getting used to a hundred and one things new and strange, that not all of us (certainly not me) paid much attention to the random greenness. Those who did found a great variety of forms:

This picture includes two satellite trails, a camera slip during the image of a star, two nebulae, a possible meteor trail, a colour abberation around an ordinary yellow galaxy, a camera colour error when photographing a star, and spillage of light onto another photographic plate when imaging the red star at the bottom . . . and one genuinely green object, second lowest on the left and far from the most obviously spectacular. To make life even more complicated, if we could see these with our own unenhanced vision, we’d see red. SDSS converts and exaggerates some colours to give us extra detail.

The earliest topic to contain a green object we now know to be special is Topic No. 158, “The Green Galaxy”, on Day 2 of the forum’s existence. Nightblizzard first posted two galaxies with apparent greenness around them, which we now know is caused by the disturbance of light in our own atmosphere. “Blue and red stars exist… sometimes I’ve even seen green stars,” he wrote. “Not sure whether it’s the camera or whether they really exist.” Crispin posted a green-edged star, and Kevin explained about the colour-banding (which Budgieye has since written up in more detail). But then Nightblizzard posted a galaxy of pure, smooth yellowish-green, probably distant, and definitely not a star. Nobody thought much more of it until a long time later.

On August 11th, 2007, Pat found a bright green galaxy, and spotted something else about it. “Is this a quasar?” she asked . Nobody quite knew. Megalodon99, mitch.wheat and jgzoo all thought it might be, though the spectrum didn’t. Whatever it was, it seemed very energetic, judging by its extremely strong emission lines which indicate very hot, often ionised gas.

Next day, Hanny found a similar object, and dubbed it, “Give peas a chance!” I barely knew what a quasar was at the time, or how to spot one using the SDSS pages (it’s amazing how much you learn here, but it takes time!). But they did look like peas, these little round green objects, and several of us roared with laughter.

“Are you collecting them for dinner?”
“That’s TERRIBLE, Hanny. Are you going to put it in the Astronomically Awful Jokes thread as well? Or would that make it appear to be vegetating? If it gets away, it’s an escapee.”
“Peas stop Alice!”
More peas were flung into the pan . . .
“It’s at this point, during this fruitful enterprise, that we discover that unfortunately we have all gone bananas. But never mind, I think everyone is appeased.”
“Judging by the colour, this should be of the mushy type.”
“Ieuw.”
“I was wondering if maybe you’d found a pulse-ah! Or maybe it was a lentle-icular?”
“I think there is definitly some alien kid up there not liking his dinner.”
More peas appeared on various topics, and now they had somewhere to be merged into. We started to think it was time to make pea soup, which would certainly be “souper”! “Oh, pea-have!” said one zooite, and another, finding a long green streak across a whole galaxy, dubbed it the Pea Soup Way.

It’s not the first time that a joke ends up being taken seriously. The Monster Raving Loony Party’s policies ended up finding their way into the law. Inventions initially proposed as jokes include tinned custard, British Summer Time and, one could argue, the idea of black holes. In any case, round green things began to flood in. “I found this, and it was suggested I put it on this thread . . .” “Here’s my contribution to the meal . . .” “The largest pizza in the Universe . . .”

“Are these ‘peas’ just any random green pea like object or a special type of object?” asked marke on 21st August.
“Random green pea like objects I think,” replied the knowledgeable Fluffyporcupine. “Some are stars some are possible quasars… some maybe alien bogies.”

“I never realized just how many peas there are in the universe until now,” wrote EricFDiaz. “They talk about stars, galaxies, nebulae, planets, etc. in astronomy courses, but they never mention the peas. It must be a big secret among professional astronomers. They probably want all the peas for themselves to eat. This has been a most enlightening experience. ROFL!!!”

We had great but uneventful fun until mid-December 2007. The breakthrough took place on a different thread from the peas – which Nightwatch called “What is this green coloured thingy?”. The thingy in question was 587728917372993708. “Apart from a pea for Hanny?” replied Fluffy. “The spectrum suggests something to do with star formation to me. Sorry not much help.” In fact, it was a crucial insight. The spectrum contained similar emission lines to Pat’s possible quasar. These very high peaks indicate great heat, which in turn indicates star formation.

Arralen joined the discussion. “It’s not a quasar . . . spectrum lines are not broad enough: http://www.sdss.org/gallery/gal_zqso.html
Looks rather odd to me – its a bit away from us (z=0.2817) and rather small, yet it must be quite bright, as the signal/noise ration of the spectrum is exceptional. And yet no QSO, AGN or somesuch ?!”

Rick Nowell found and posted an article about novae, the tiny equivalent of supernovae, and their green appearance. Fluffy pointed out something even more important than her earlier finding: that the spectrum suggested a lot of doubly ionised oxygen. (This shows up as OIII. OI means not ionised, and each I after that indicates one electron knocked off the atom – usually, again, by intense heat causing fierce collisions between atoms.) Rick was intrigued. “A galaxy surrounded by ionising oygen is not something I’ve heard of – is there a name for such an object? It would be have to be very energetic to do the ionising bit.” He went and did some more research and found that quasars, too, are surrounded by very energetic, ionised gas. There was also something to do with Dark Matter, which we haven’t looked at yet to the best of my knowledge.

The peas’ fate was sealed when Kevin arrived on the scene. “The fact that the image is green comes from the fact that the powerful [OIII] line sits in the r-band (which is the green in RGB). Whatever it is it has extremely powerful emission lines, but the ratios look very weird to me. It’s definitely a strange object. I will have a more detailed look at it in the office tomorrow.” His conclusion next day: “Peas are galaxies with enormously powerful emission lines.”

It still sounded pretty funny at the time. Now it sounds normal, but normality has a beginning . . .

At that point, Rick began to make lists of all the “peas” in the thread. “I was bored,” he claimed later. But it was a very impressive list, 39 peas with those particular emission lines. He later deleted the “original original” and made a detailed, professional list in its own topic. “Great job Rick!” said Fluffy. Rick then went to look up various papers to see if this was something known yet. He felt it was established that peas were emission line galaxies, and a little later announced that 53 of our objects in the thread (far fewer than the total of postings) were such a thing.

Learning to read spectra wasn’t part of everyday zoo life at that point, and it took some of us, definitely including me, a while to get the idea. On January 3rd 2008, Starry Nite asked: “Could someone “in the know” post the criteria we should be looking for when searching for peas? Obviously, green color on SDSS is a start. And ObjId should say galaxy, not star. But what if there’s no spectra listed? And what does it mean that some peas are blue-green, while others are yellowish or brown-green? Please help the ignorant (me)!” Hanny hadn’t meant this topic as a collection point for anything specific – but that was what it was becoming. Rick replied:

“Starry Nite: the Pea galaxies that have caused interest are deep green Emission Line Galaxies (ELG’s). They are rare, but GZ seems to have found a few (54). They have a lot of ‘doubly ionised oxygen’ in them, which causes them to appear green.” A few days earlier, Kevin had made a pea Object of the Day.

Then things began to get a bit more complicated.

Arralen posted a galaxy with just the right spectrum – but just the wrong appearance! It was a beautiful starforming oval – with a high emission line of doubly ionised oxygen. Clearly such a spectrum did not necessarily produce a pea-like appearance.

“I was interested in the green peas, but what really got me hooked was finding a red object which Arralen said could also be a pea,” Starry Nite told me when I let the pea experts know I was polishing this post. “That and the pretty, candy-like colors of course. I think I was the first person to start collecting peas over their whole redshift range, but it wouldn’t have started without his observation.”

On January 15th, Starry asked if we could become more systematic. “I’m wondering if anyone knows if there’s a way to search the SDSS by spectra. Is the data from the spectral chart of each object that has one organized in a way that someone could search for all objects with an O III level of a certain value? Since all those charts are just organized numbers, seems like a simple program could flag everything that had a high number for O III related to low values for the other parts of the spectrum. Anyone know if such a thing is possible with the way SDSS is currently organized? Oh, and if someone does create such a program, would you consider calling it ‘Pea-Picker’?” There was no immediate answer to that. A few posts later he added, “By the way, Hanny, you found rare green peas, then a blue Voorwerp, so I believe the next in the rainbow you have to find is something yellow that sets the world of astronomy a-buzz!” Well, I think purple – if you’re going from green to blue. Speaking of which, the next not-green pea to appear was purple. Arralen had found the right spectrum again. By this time, nearly everyone was including a spectrum with their object. “More Peas not War,” said Galaxy Hunters Inc.

Starry Nite asked if blue “peas” were peas that were simply closer than green ones – and if red ones were further away. He wanted, as he later put it, to “figure out what kinds of objects have OIII-dominant charts and how they relate to each other. What I think I’ve pieced together could all be wrong but I tried to correlate the little blobs we see at great distance with the same type of object when they’re much closer and look different.” At this point it was asked if this wasn’t messing up Hanny’s thread altogether, the original point of which was to find green things. “I don’t care about what color the peas/beans have,” laughed Hanny, and it was clear that the effort being made to develop our own system was roundly admired.

Starry Nite was taking this very seriously. “Now, to prove that I’m not obsessed with them, a list of all the Peas with spectral charts arranged by z-shift: http://www.galaxyzooforum.org/index.php?topic=3638.msg108116#msg108116]
Characteristics of a “Pea” galaxy A.K.A. OIII galaxy:
1. Mostly-flat spectral chart except for:
2. An extreme peak at OIII (double-ionized oxygen emission line),
3. If there is a peak at OII, it must be shorter than the OIII peak,
4. Other peaks must all be smaller than the OIII.
5. Any H-peaks should be narrow, not wide-based which might indicate a quasar (with a redshift z<0.3).
6. Redshift range (z) of approximately 0.14 – 0.35 for a green color on SDSS.”

“UMmmm. Starry ………..Just so I understand……………….What is your definition of obsession?” queried Galaxy Hunters Inc.

For some time after this, the thread became dedicated to OIII-peaks and debates about double-posting (as Starry Nite was creating a new thread). Rick and Starry continued to write great posts. “I wish I could write a paper about it!” said Rick on February 9th.

We had our first Galaxy Zoo Get-Together just around then and Half65 reported a mass peas breakout. Fortunately, puns, poetry, peals of laughter and pea soup got them all back!
Starry wrote:
“And though the peas are rather small
They had to count them all
Now they know how many peas
It takes to fill the Albert Hall”.
After that they behaved.

On February 19th, our third major player, Laihro, joined the pea hunt. By now a pea spectrum could mean all sorts of things, and yet they just knew it all must mean . . . something. They worked harder and harder to find out what this might be. Peas were often compared to the QSO thread’s collection, and suddenly Wolf-Rayet galaxies were trying to sneak in as well!

Discoveries in science are seldom made with a leap and a “Eureka!” but by relentless arguing, refining, testing, and making tighter and tighter definitions. Starry gave us some more pointers at the end of February, and wrote a terrifyingly long catalogue! FermatsBrother saw a pattern, and put the peas into a beautiful table. He put the redshift (from which we can calculate a galaxy’s distance from us) into the first column, followed successively by the wavelenth of OIII (which would increase with increasing redshift), the actual colour the light of this wavelength would be, and the colour of the object at these redshifts and wavelenths that appear on our screen. This clearly indicated that peas must change colour according to how far they are from us, as Rick reflected.

On March 13th, FermatsBrother then went to CasJobs in SDSS to run a program to find all the peas from redshift 0.27 to 0.30 of the right colour. He called it his “pea harvester” and posted several results, some of which had been found through classification, others not. “What seems to be astonishing, is that most of them appear to be the same angular size!” he noted.

Starry Nite challenged him to describe his method:

“If you copy/paste the following program [into CasJobs] you will get a list of all the “green” Objects for z 0.27 to 0.30.
The selection criteria for “green” is that the [g’ Magnitude value > r’ Magnitude value] and that the [i’ Magnitude value > r’ Magnitude value] as occurs in the green Objects that have already been found.
I’ve also inserted a criterion for the zconf as 0.9 quite arbitrarily.
I’ve put a sort on the z value, but this could equally well be the ObjID.
The program as presented picks up many of the peas you already have.
Have a go with it, and try changing the parameter values to see what you get !
Cheers – Fermats Brother”

His program was:

---------------------------------------------------------------------------

select

PhotoObj.modelMag_g, PhotoObj.modelMag_r, PhotoObj.modelMag_i, PhotoObj.ObjID

from

PhotoObj, specObj

where

specObj.bestObjid = PhotoObj.ObjID AND

specObj.zconf > 0.9 AND

specobj.z between 0.27 and 0.30 AND

PhotoObj.modelMag_g - PhotoObj.modelMag_r > 0.4 AND

PhotoObj.modelMag_i - PhotoObj.modelMag_r > 0.4

Order by specObj.z

---------------------------------------------------------------------------

Starry Nite immediately tried it out and found several dark red peas. Then on March 21st he noticed that the spectral chart of Hanny’s Voorwerp might as well be a pea spectrum! Doubly ionised oxygen was turning up all over the Universe!

Next day Rick offered us our “Celebratory List and Counting!” He couldn’t resist quoting Kevin’s old words: “I wonder if the “pea selection” yields a sample of interesting objects….” It certainly did.

Laihro, who is a genius with computers, was getting quietly but increasingly involved. Following the thread at this point becomes rather like a detective story because he seems to have made a huge search of 7,400 objects – but later deleted his post. Sherlock Holmes has to get out his microscope and rely on posts by FermatsBrother asking more and preserving a few remains, and a reply by Laihro denying all knowledge of astrophysics! Nevertheless, bowing to pressure, on April 9th he published his SQL query. He set up a thread specifically to find and study OIII objects – which is pretty technical but you feel you learn a lot while reading it, and it’s an excellent store of links! Edd came in to talk about the data too.

Twelve days later Laihro extended an open invitation to discuss the definition of OIII objects. By this time, the emphasis of the thread was very strongly on OIII objects, rather than little round green things. Laihro wrote another list here.

And so the hunt continued! ElisabethB and Galaxy Hunters Inc were particularly active, besides our three well-known geniuses. More and more peas appeared – until the middle of May, when the great fire of servers put out our beloved forum for many weeks. The posts stop abruptly. Then, six weeks later, on July 8th, we see a post that makes you sit up slightly because the poster has one blue star rather than the usual one to five yellow ones.

“Hi, I’ve started a new thread called ‘Peas Project’ to collect a sample of Peas for further investigation. Hopefully we can gather enough of these interesting new objects to find out what they are and why they’re so special. Carie.”

“Great news!” grinned the pea maniacs, who had been continuing their work on the temporary forum anyway.

So we headed along to Carie’s thread and it was a request for a sample of a particular class of peas. The ones she and Kevin were after were “green, compact . . . have a big bright OIII line . . . [and] should be at a redshift z > 0.15 and z < 0.45.” People welcomed her to the zoo and asked if that meant we should go through the whole peas thread to dig out the relevant candidates. Elisabeth recommend Laihro and Rick as good points of contact. (Sadly, Starry Nite had been away for a while by then. I’m so glad he’s been back more often recently!)

We discussed the method a little more. “As I said before: Rick Nowell!!!” said Els.

“Pea-hunters of the world unite!” replied Rick.

Rick and Laihro began to pare down their by this point extremely extensive lists and debate the best methods of doing this. But of course everybody wanted to continue collecting, and newcomers wanted to know what was going on. Carie wrote, “I think part of the reason we’re so interested in the Peas is that they don’t fall so neatly into any of the typically studied categories. They appear to be quite small and compact, but certainly aren’t stars or distant quasars. At first glance, they have emission lines like star forming galaxies and buried AGN, but they are much smaller then what we’d expect to see at these nearby redshifts.”

Frustratingly, getting an ideal pea sample off the forum was proving less and less possible. Newcomers weren’t sure whether they should post their peas in “Give peas a chance!” or in “Peas Project”; it still wasn’t easy for everybody to decipher spectra. However detailed the first post, no set of long instructions can be instantly memorised, and forum users are often more inclined to read the most recent post and make assumptions about what’s being collected from there. That’s not to criticise. It’s human nature. Stopping and studying isn’t something you really feel like doing until you’ve already plunged in and got really excited.

Waveney neatly solved the problem in exactly the same fashion as he tackled the mergers. He entered every collected pea into a sorting program and gave us the option of “green”, “not green” or “greenish”. It was a sample of about 2000, and many of us challenged ourselves to do the whole lot in one evening. It only took a couple of days to have a definite result. There was little disagreement about what was what colour. (It was a purely visual check; redshift and spectrum was checked automatically.)

But we still didn’t know what peas are! Kevin announced that many were AGNs, but their galaxy type remained infuriatingly uncertain. (He even obtained a Hubble picture of one which revealed no clear structure!)

On July 22nd, Kevin wrote, “Carie and me have defined a sample of peas and downloaded their spectra. We’re currently processing their spectra with a code called GANDALF to measure their emission line strenghts. If nothing goes wrong, this will take about 1.5 days, after which we can start doing emission line diagnostic diagrams and learn more about what powers peas.”

We all got silly with excitement at that. Half65 suggested that GANDALF might be a tool better made for studying ring galaxies, to a chorus of groans; next, a string of suggestions beginning with “P” was put forward as to what powers the elusive things. (GANDALF measures and filters out unwanted emission lines from stars and gas.)

On July 25th, Kevin announced that the plots Carie had created shed no light yet, green or otherwise!

On August 19th, Carie gave us an update on the plots, their colour-magnitude, and their power source, which was a huge rate of star formation. The Milky Way’s rate is one or two stars an Earth year; peas appear to produce around 40! (At which point Kevin turned up and remarked that that’s not really a very fast rate; early galaxies would have been up in the thousands. Blue compact galaxies also produce many more. But the small size of the peas still made it surprising.)

Over August and September, Kevin and Carie “shelled the peas” while zooites continued to hunt (although the official hunt was over, volunteer curiosity was not). On October 2nd, Carie showed us another graph which demonstrated that very few were AGN, and still refused to put them into a specific class of galaxy. She then added another graph with corrections six days later. The same day, Carie, Kevin and Waveney organised another brief project to sort 439 peas a little further.

Kevin and Carie applied for Hubble time on the peas project in December, and in January were debating names for the peas paper – which they submitted on April 14th. The volunteer Citizen Scientists who planted the peas are now enjoying watching them grow. Rick, Laihro, Waveney, and many others who’d contributed were each given copies of the journal at Oxford on Sunday 21st June, and the peas will be a main focus of Pulse-Project’s upcoming documentary about Galaxy Zoo. Give peas a chance? Yep – we all did; and, thanks to so many people’s hard work, so will science.

— Alice