I had used this post previously to help me write a paper a while back. The paper is out:
http://arxiv.org/abs/1106.1765
I am now struggling to finish my thesis for Monday and using the blog again to write up this particularly mentally blocked section :) will be back in writing action after it’s all done.
———below is the ooold part of this post——
ok so here’s the thing. i actually hate writing. or working maybe. and im having a terribly hard time trying to write a paper on what i’ve been working on these last few months. so im going to use you all to help me write it. hehe =) this way it doesnt feel like work. it feels like procrastination. and my thoughts would be more fluent, and i’d strive to be clearer. so i’ll explain it all. you dont have to read it or anything though hehe, but its nice if you want to. atleast the overall idea. in some parts im also going to pretend that you are interested in things like errors. i have made the post private, as i have heard horror stories of research being stolen, and im not sure if this is worthy of theft, but i dont want to risk it as i have spent a lot of time on it and it is very crucial to my career. i should also mention that some explanations are beyond the scope of this article, and you will have to look them up, for example the standard cosmological model.
UPDATE: I HAVE WRITTEN AN OK DRAFT NOW, THANK YOU FOR YOUR HELP. I WILL NOT TORTURE YOU ANY MORE WITH ALL THE CRAZY DETAILS, BUT I’LL LEAVE THE GENERAL BIT ON. FOR THOSE INTERESTED IN DETAILS, I’LL LINK YOU TO THE FINAL PAPER WHEN ITS DONE =)
general background (for non-physicists and people not interested in the subject) :
the \lambda CDM model is the most widely accepted model of cosmology, the one that is used to understand how the universe as we know it today came to be. how the various structures in the universe like galaxies and clusters of galaxies formed. according to this theory, ordinary matter, known as baryonic matter contributes to only about 4% of the energy of the universe. the rest of the matter in the universe – we dont know much about it – is given the name dark matter, as they do not emit any electromagnetic signals (x-rays, gamma rays, visible light etc). this dark matter contributes to 22% of the energy content of the universe, and is assumed to be cold, that is, these particles are not going too fast. so we have CDM – Cold Dark Matter. the thing about dark matter is that these particles only interact with other matter and with each other through gravity, as far as we know. they were initially proposed to account for the extra gravity that is inferred from the motion of stars. \lambda is the component which makes up the rest of energy density (~74%), and it is called dark energy. and we don’t need to go into that here. these days, there has been mounting evidence that cosmological structures form hierarchially - small galaxies merge to form bigger galaxies. small structures merge to form bigger structure. these structures are embedded in dark matter. so you can picture the universe as a neural web of dark matter – there are stars and galaxies embedded in dark matter and these are connected to each other by strands and filaments of dark matter. the large masses of dark matter are attracted to each other due to gravitation and so merge, thus causing the galaxies to collide and merge. here you can visualise all of this :
http://www.youtube.com/watch?v=X7APek5SQ2s&feature=related
http://www.cita.utoronto.ca/~dubinski/nbody/Mice_long.mpg – colliding galaxies
so you see, galaxies are immersed in dark matter, which is much more extended than the baryonic (ordinary) part of the galaxy. in the galaxy disk, where all the stars are, and lots of things are going on like supernovae, it is difficult to purely study the effect of dark matter. moreover, there’s more visible matter here than dark matter. so we have to look beyond the outer edges of the main stellar part of the galaxy and focus on the dark matter around it – the dark matter halo.
A study of the global features of the halo of a galaxy like its shape, extent, rotation etc are important in understanding the nature of dark matter particles themselves. The DM halo is in general ellipsoidal, but we will assume its spheroidal (valid assumption for spiral galaxies). We want to find its flatness q, that is the ratio between the vertical axis and equatorial axis. The halo can be either spherical ( if c = a), oblate ( c < a), or prolate (c > a).
N-body simulations of gravitational collapse of density peaks show that a flattened halo implies that the DM particles are cold, whereas a spherical one implies that they are hot. A massless spherical halo component would mean that the ‘missing mass’ is due to a modification of gravity rather than due to undetectable matter. The study of the distribution of the halo could also tell us about the local orientation of the cosmic web, since the smallest axis tends to be normal to the sheet in which it lies. A survey of galaxies and their halos could tell us how the shape of the halo relates to the main stellar component, as well as recent mergers.
Various methods have been developed to study the properties of dark halos, but some of the results obtained are inconclusive and some of them seem to contradict each other, and there is no real consensus on several of these properties. For example, efforts to constrain the flattening of the NGC 5907 halo by N-body simulations of its tidal stream (image) show that similar streams are produced for different values of q. Nevertheless, these tidal streams are important probes of dark halos and in this paper, we improve on the existing methods based on them. These stellar streams are left behind by a satellite galaxy as it falls into the host galaxy and trace out the orbit of its progenitor. We use maximum likelihood estimation (MLE) to constrain the parameters of the halo potential using the guiding center orbit of the stream. 
Besides tidal streams, gravitational lensing has proved to be a powerful tool in constraining the flattening parameter as it is sensitive to the ellipticity of the lens, but the values obtained by strong lensing are also sensitive on how well the substructures in the halo have been accounted for. Weak lensing, being much less sensitive to the substructure, provides a better way to measure the halo profile. Other methods that have been used involve HI, X-ray gas, polar rings and velocity distributions of halo stars. Fitting the warps in the gas in spiral galaxies have also been used to study the DM halo profiles.
TO CUT A LONG STORY SHORT :
We find the combination of parameters that generate the best fitting orbit to the stream using Monte Carlo Markov Chains. And it works! We can get back the flatness parameter q of the potential that we make the fake data with.
