Can we trust our measurements in cosmology?

[sithe]

Morning everyone,

We know of some illusions in our everyday experience, such as an echo making us believe someone is talking back at us, or mirages in the desert, or mirrors making us think there are two symmetric objects when one is only the reflection of another.

I was wondering about such illusions in cosmology. I know of two, when we look at a star whose light has been distorted by gravity, we might think the star is in one place when it is actually in another. Also we sometimes see two galaxies when there is actually one galaxy because of the distorting effect of dark matter.

How do we account for such possibilities / anomalies when doing measurements in cosmology, e.g. when looking at radiation from distant galaxies / stars / objects. Or even when looking at complex things like cosmic background radiation etc.

Thank you for any insight you can give.

Sithe

 

[mathman]

I think you need to be more specific.

http://en.wikipedia.org/wiki/Gravitational_lens

The above may be an answer to some of your questions.

 

[Nugatory]

How do we account for such possibilities / anomalies when doing measurements in cosmology, e.g. when looking at radiation from distant galaxies / stars / objects. Or even when looking at complex things like cosmic background radiation etc.

If we know about the existence of the illusion, as in your example of the star whose position is distorted by gravity "bending" the light from it, we also know to do the necessary calculations to compensate for it. Likewise, we understand mirrors and echoes, draw correct conclusions about what's really going on even when our senses suggest an illusion instead,

Another example: Light bends as it passes through the surface of water, so if you see a fish under the water, it's not where you think it is - but that doesn't stop people from spearing and harpooning fish with quite remarkable success.

And another more mathematical example: You have no problem understanding that what you're seeing when you look at Mars in a telescope isn't what's happening on Mars right now, it's what was happening on Mars a few minutes ago when the light you're seeing left Mars.

Of course there's always the possibility that we're being fooled by an illusion we don't know about... But it is surprisingly difficult to find plausible (as opposed to, for example, malevolent fairies perturbing our measuring instruments) illusions that work everywhere, without eventually slipping somewhere. The flat-earth illusion has taken many people in, but even 2000 years ago people were looking at the Earth's shadow on the moon, the behavior of ships sailing below the horizon, the behavior of shadows at high noon, and seeing through the illusion.

 

[Chalnoth]

Morning everyone,

We know of some illusions in our everyday experience, such as an echo making us believe someone is talking back at us, or mirages in the desert, or mirrors making us think there are two symmetric objects when one is only the reflection of another.

I was wondering about such illusions in cosmology. I know of two, when we look at a star whose light has been distorted by gravity, we might think the star is in one place when it is actually in another. Also we sometimes see two galaxies when there is actually one galaxy because of the distorting effect of dark matter.

How do we account for such possibilities / anomalies when doing measurements in cosmology, e.g. when looking at radiation from distant galaxies / stars / objects. Or even when looking at complex things like cosmic background radiation etc.

Thank you for any insight you can give.

Sithe

The basic solution is, well, to use science. This breaks down roughly as:

1. Use an explanatory model that makes definitive predictions about how various observations should be related to one another.
2. Perform many independent observations to support or falsify the model. The more independent observations we have which test different predictions of the model, the less likely it is that our model accidentally agrees with the observations.

For example, take Big Bang Nucleosynthesis and the Cosmic Microwave Background. BBN is our name for the physics that formed the light elements in the early universe, and our understanding of the physics in operation there makes some very specific predictions as to what the abundances of light elements should be. This observation is performed by observing stars in places where very little stellar nucleosynthesis has taken place, such as dwarf galaxies or very far-away objects.

The exact same theoretical model which describes BBN also makes corresponding predictions about the Cosmic Microwave Background. Specifically, BBN depends critically on the ratio of photons to protons/neutrons. And the CMB itself gives an accurate measurement of this ratio.

The two observations agree.

All told, there are a little over a dozen different types of major cosmological observations, and a number of repetitions of each type of observation. So far the simplest model that fits all of these observations is the [itex]\Lambda[/itex]CDM model with cosmic inflation, which includes a cosmological constant and dark matter, as well as a really simple model of inflation.

 

[Low-Q]

...

And another more mathematical example: You have no problem understanding that what you're seeing when you look at Mars in a telescope isn't what's happening on Mars right now, it's what was happening on Mars a few minutes ago when the light you're seeing left Mars.

I have a question about this "effect". What we observe on Mars from a telescope, isn't that happening right now? I assume that if an observer travels to Mars (hypotetically) in no time and discover a specific hurricane, the observer on Earth would see the same hurricane at the same time through his (Very expensive) telescope, and not 20 minutes later. However, the observer on Earth will see that the traveller lands on Mars 20 minutes after the hurricane occoured...arrgh.. this is mind twisting stuff

Anyways, when we on Earth observe the universe, I find it a little hard to understand how we can calculate far far distances, the age of the universe etc. if the universe has expanded in the mean time.

Vidar

 

[sithe]

Thank you for that input.

Let's take the CMB as a specific example. We observe it today but it was emitted 13 billion years ago. What distortions could have happened since, space is expanding, there is dark matter, dark energy, speed of light is not constant under different conditions, etc etc ... there are all these effects and how many other unknown ones.

I guess my question is how confident are we that there isn't some effect that is not accounted for. I hear you about the consistencies that we get from using different observation approaches. But all our observations are done through a very tiny window looking out at a possibly infinity universe.

Would it be correct to say we can at best have working models, but we should always be open to possible fundamental illusory effects. Or are we very confident that no such fundamental errors exist.

 

[cristo]

I have a question about this "effect". What we observe on Mars from a telescope, isn't that happening right now? I assume that if an observer travels to Mars (hypotetically) in no time and discover a specific hurricane, the observer on Earth would see the same hurricane at the same time through his (Very expensive) telescope, and not 20 minutes later. However, the observer on Earth will see that the traveller lands on Mars 20 minutes after the hurricane occoured...arrgh.. this is mind twisting stuff

No, what we mean by "observe" something is that we see the light coming from that event. The speed of light is finite, and so this puts a restriction on how we observe things. When we look at mars, we see light that was emitted a few minutes ago, so if a hurricane started now, then it would be a few minutes before we could see it. Likewise, if an observer instantaneously appeared on Mars at the same time as the hurricane started, we would not see that observer on Earth until the light arrived, and would see the hurricane start and the observer arrive at the same time.

 

[cristo]

Would it be correct to say we can at best have working models, but we should always be open to possible fundamental illusory effects. Or are we very confident that no such fundamental errors exist.

Is this not the definition of science? We can, at all times, at best have a theory that fits all known data at that point. But tomorrow there could come along another observation that cannot be explained by the current best fit model.

Now, in the case of the cosmological model, there are some assumptions that come in. The main assumption is known as the cosmological principle, and it (loosely) states that we are not in a special place in the universe. Most people would agree that this is not a controversial assumption to make. Of course, we can only say anything about our observable universe, but with the cosmological principle we can extrapolate to the entire universe.

There is an immense wealth of observational evidence supporting the standard cosmological model, however it could be the case that tomorrow there comes some observation that puts the model into doubt. But that is really what makes science exciting!

 

[Low-Q]

There is an immense wealth of observational evidence supporting the standard cosmological model, however it could be the case that tomorrow there comes some observation that puts the model into doubt. But that is really what makes science exciting!

Good point! Questioning science is also science.
For example, why do we look away from the big bang to calculate the age of the universe? How can we look at the farthest object observable and say that this object is as it looked like "right after" the big bang? It took a while before the observable universe expanded that far, didn't it? How much has this object developed right now? Is this a galaxy now with stars and planets just as our own galaxy? How far away is this galaxy right now? I love science - full of surprises every day, and we never really see the whole picture :-)

Vidar

 

[Chalnoth]

I guess my question is how confident are we that there isn't some effect that is not accounted for.

Very confident, at least in terms of the overall cosmology. There are likely to be discrepancies in the details (e.g. the precise nature of inflation, dark matter, and dark energy, where General Relativity gives way to quantum gravity). But not in the overall picture.

I hear you about the consistencies that we get from using different observation approaches. But all our observations are done through a very tiny window looking out at a possibly infinity universe.

Sure, but we're looking at extremely different types of things that, were our models incorrect, shouldn't be at all related. Here's an example with some real data:
http://physics.stackexchange.com/questions/64035/cosmological-triangle-with-planck-results

What you can see here are the experimental constraints on the amount of matter (normal + dark matter) and the amount of dark energy. The different colors here are different experiments, which I'll describe briefly below. But for the moment let me just point out that these three different experiments did not have to agree: they could easily have failed to intersect at the same point.

SNe: These are measurements of supernova brightnesses over distance. These give us a distance/redshift relationship, as the brightness of the supernovae is a reasonably-good proxy for distance. This, in turn, tells us how fast our universe has expanded over time.

BAO: This is a measure of the typical separation between galaxies. Early in the universe, the physics that occurred before the CMB was emitted set up the initial distribution of over-dense and under-dense places in the universe. Over time, on large scales, these under-dense and over-dense regions have expanded away from one another. Thus, the typical separation between galaxies gives us a different measure of how much expansion there has been.

CMB: The CMB is measured in the millimeter wave regime, with the hot and cold spots representing over-dense and under-dense regions in the early universe.

Would it be correct to say we can at best have working models, but we should always be open to possible fundamental illusory effects. Or are we very confident that no such fundamental errors exist.

There's always the possibility. It's just very unlikely. And we expect that if our models are wrong, at some point the observations will demonstrate that.

Heck, if we want to be pedantic, we know that all of our models are wrong on some level. The difficulty lies in finding precisely how they are wrong, and what the right way to correct them is.

 

[sithe]

Thank you Chalnoth ... very useful insights ... it makes very good sense

In reference to the lamda cdm model, do we have any measurement results on dark matter yet, besides the computer modelling which shows there is a lot more mass than is accounted for in visible matter .. so this extra mass must come from dark matter

The confusing thing with dark matter for me is that it has none of the properties of ordinary matter, except it has mass. We think mass comes from the Higgs, does that mean dark matter also contains it, or is there another mechanism by which mass can arise?

 

[Chalnoth]

In reference to the lamda cdm model, do we have any measurement results on dark matter yet, besides the computer modelling which shows there is a lot more mass than is accounted for in visible matter .. so this extra mass must come from dark matter

Personally, I think the best evidence for dark matter is the CMB. But that takes a bit of time to explain, so I'll instead leave with a description of this bit of evidence which is very visceral in its directness:
http://www.preposterousuniverse.com/blog/2006/08/21/dark-matter-exists/

The confusing thing with dark matter for me is that it has none of the properties of ordinary matter, except it has mass. We think mass comes from the Higgs, does that mean dark matter also contains it, or is there another mechanism by which mass can arise?

As far as we can tell, it basically acts like a neutrino but with more mass. So it's not that exotic. This kind of particle is rather common in beyond-standard-model theories. Supersymmetry, for example, requires a number of neutrino-like particles with higher masses to exist.