Modeling Energetic Light From Our Neighboring Galaxy
Sep 5, 2023
Gary Forman is a graduate student in Dr.Paul Ricker's research group. His current interests lie in modeling light from extraterrestrial bodies, like gamma rays from star development and star's undergoing supernovae, to develop simulations and models of astronomical phenomena. This presentation also concerns finding methods to describe the motion of cosmic rays.
OK.
Hello, everyone.
Again, my name is Gary Forman.
I'm from the Department of Astronomy.
And today, I'll tell you a little bit
about my work in modeling the most energetic light
from our neighboring galaxy.
So first, let me get you oriented with the picture
I have plotted here.
This is real data from the Fermi Gamma-ray Space Telescope,
which has been performing an all-sky survey for the last
about 4 and 1⁄2 years.
And to give you a sense of how large this image on the sky
is, this white streak here is 1 degree.
That's about twice the size of the full moon.
So we have a pretty big coverage in the night sky
with this image.
And so the first part of my title
is the most energetic light.
And by that, I mean gamma rays.
So gamma rays are more energetic than the ultraviolet light that
is at the high end of the spectrum of the sun,
more energetic than the X-rays that we use to image
our bones at the doctor.
They're so energetic that they're not
being generated by thermal processes in the universe.
These photons are strictly non-thermal,
just because nothing can achieve the temperatures that
would be required to create them.
The second part of my title was our neighboring galaxy.
And by that, I mean the Large Magellanic Cloud.
The nice thing about the Large Magellanic Cloud
is it's close.
It has a well-known distance of about 160,000 light years.
By Earth scales, this sounds far away.
But given the size of the observable universe
at about 14 billion light years, this
is in our galactic neighborhood.
It's also nice that it's face on.
So we can imagine maybe some of our studies
would be easier done in our own galaxy.
The Milky Way, it's closer.
But the problem with the Milky Way is it's viewed edge on.
So you can imagine kind of like trying
to pick out the best slice of pizza
by looking at the pizza edge on.
It's hard to do.
You can't see which slice has the most toppings.
You can't see which is the biggest slice.
Well, with the Large Magellanic Cloud,
we have our pizza face on.
We can see all the pepperoni and cheese quite easily.
And important for my study is that this galaxy
has been very significantly detected by the Fermi gamma ray
space telescope.
OK, so what's our picture?
We said that gamma rays are not being
generated by thermal processes.
So how do we think they get there?
Well, the story begins with star formation.
Gas that's available in the Large Magellanic Cloud
is, through some processes, mainly gravitation,
able to collapse to form stars.
The most massive of these stars that we create
are going to burn hottest, live fastest, and die youngest.
And when they die, they're going to cause
explosions known as supernovae.
These supernovae are going to shock the surrounding
interstellar material.
And across these shock fronts, we're
going to accelerate charged particles known
as cosmic rays.
These cosmic rays consist mostly of electrons and protons,
but also include heavier atomic nuclei as well.
And there's a positron component.
We have our cosmic rays.
If we consider the protons, those
can now collide with thermal ambient protons
in the interstellar medium.
And when these collisions happen,
one product is a neutral pion.
These neutral pions have a very short lifetime,
and they'll decay to the gamma rays
that we've seen in the picture shown on the first slide.
So this model has been around for something like 70 years.
But it's only now that, with the great data from the Fermi
Telescope, that we're able to start testing it.
And one way to test it, is what I'm doing,
is by comparing what we think are the relevant galactic
components of the Large Magellanic Cloud
to the gamma ray mission.
So on the far left here, we have a map
of the thermal ambient protons, ionized hydrogen.
These are the colliders for our cosmic ray protons.
And on the right, we have a couple
of maps of star formation rate.
We said star formation was the starting place
for how to create cosmic rays.
Now, using least squares likelihood fitting,
we fit a combination of these maps to the gamma ray
data below on the left.
And our best fit results in a residual
that doesn't really seem to indicate that there
were any high counts there at all.
So we think we're doing well with this fit,
and we're supporting the picture
that I painted in the previous slide.
OK, so now we have an idea based on observations
of how we think these cosmic rays are being
accelerated through supernovae.
So what simulations can do is help
us understand how these charged particles move
around the galaxy once they have been accelerated.
So to do our work, we're going to use the adaptive mesh
magnetohydrodynamics code FLASH.
It's an open source code widely used in the astronomy field.
FLASH also has a particle mesh gravity solver,
which we heard about a little bit in an earlier talk.
And recently, a cosmic ray propagation module
has been added to the code.
Cosmic rays are treated as a fluid on the numerical grid.
And this fluid undergoes, or the equation of motion
depends on advection, diffusion, sourcing, and energy losses.
OK, so as I said, we want to better describe
the motions of cosmic rays after they've been accelerated.
To do this, we really want to make a realistic simulation
of the Large Magellanic Cloud.
And our simulations are going to depend
on quite a bit of physics, not least of which
includes magnetic field topology.
We're talking about charged particles,
so their motions will care about the presence
of magnetic fields.
We need to know about the density,
the presence of the thermal interstellar medium,
because those are the particles that
when we collide our protons with are
going to generate the gamma rays.
Star formation is important for sourcing of our cosmic rays.
We'll wish to model energy losses.
And also important is properly modeling
the gravitational potential.
So on the right is a simulation that we've
done of the LMC's potential.
The way it's initialized is that this disk shown
will maintain its profile in isolation,
but we've introduced tidal forces
from the Large Magellanic Cloud's neighbors.
And we'll see this has a profound effect
on the morphology.
And again, we're looking at the disk of the Large Magellanic
Cloud face on.
So we see some tidal stretching and development
of spiral arm structure.
So I'll end with some future work that we wish to perform.
We'd like to continue adding physics
to our Large Magellanic Cloud simulation.
And we'll constrain the numerical model
for cosmic rays by comparing our simulated data
with the gamma ray observations.
Additionally, we'd like to investigate
how we might enhance our calculations by accelerating
the flash code using the GPU architecture.
Currently, flash is CPU only.
And additionally, we'd like to perform
a study of an additional nearby galaxy called
the Small Magellanic Cloud to see
whether our gamma ray model holds
in a pretty different environment.
So I'll end with that.
And thank you very much.
Well, thank you.
We have time for a couple of questions.
I'll start.
How large are those data sets, the experimental one?
Because your image looks like 200 by 200 pixels,
but it's probably more than that.
So the image I showed is about 200 by 200 pixels.
It's an all-sky survey, though, so we
haven't been trained on just the LMC for the full time.
And the resolution of gamma ray astronomy
is also pretty low as well.
So actually, our pixel sampling is
beyond the true resolution of the telescope.
Other questions?
Well, I just wanted to thank all the speakers today.
Thank you.

