First and foremost, it is a matter of avoiding light pollution, having clear skies, and having as little atmosphere above the telescope as possible. Thus, you want to build it in a remote, dry, high site. There are of course higher sites than this (1800 metres). I guess that they wanted it to be...
My data hasn't got that low SNR (then it would have been totally worthless), I've just zoomed in on the baseline, since the baseline was what I thought was relevant to the problem that I tried to solve by this post (the spectral lines are cutoff, as you can see in my first posts). If I put a...
Here are some spectra found by Google image search on "millimeter astronomy spectrum" or similar:
I could also give you links to a gazillion articles in Astronomy&Astrophysics and Astrophysics Journal using the same scale...
Perhaps parts of this text may be a good reference, even though I...
Low-pass filtering the data and then subtracting that result from the data produces the same result as high-pass filtering the data (not so surprising).
Sounds fairly reasonable, I believe. This is what they call a superheterodyne receiver, if that helps. And remember, I'm an astronomer, not a...
I'll have to think about this and get back to you tomorrow... but I'm at least quite sure that the temperature on the y-axis is proportional to a system voltage.
The noise in my graphs is more or less purely receiver noise.
Short version: I think that I've found an acceptable solution to the problem, thanks to the suggestions from sophiecentaur!
I made a code (now in Python, since it's more comfortable to me) that creates a baseline from the mean of a group of channels for each channel. The mean is weighted by a...
I also thought this was a bit confusing when I was new to radio astronomy (and then I got used to it...). If I would look at a completely blank spot on the sky, I would get noise around zero (positive and negative). If I take a very short integration, the noise would have large amplitude around...
You definitely can have a measured flux density that is negative, since your baseline is usually a black-body continuum (which can be assumed to be flat at these relatively narrow bands). And the noise is fluctuating around zero, equally many wiggles below as above. These are simply fluctuations...
Haha, I agree axes should have labels, point taken ;)
The y-axis represents flux density, which is usually measured in Kelvin(!), but this is now an arbitrary unit. The x-axis represents frequency, but is now just channel number. It is all completely linear. So the detector reads off how much...
First, thanks for all help so far.
I'm sorry, but I don't really understand what you mean. How is this similar to an HPF, and what should I do with the baselines?
Ok, I gave it a go. I know very little about filters (I've never really had a use for them before), but found a page claiming...
The sinusoidal baseline is an artifact. This is radio astronomy data, and a non-flat baseline does not make sense. The receiver people at the telescope are aware of the problem, and the telescope is nowadays producing nice flat baselines - but my data are from before that was fixed, and they...
Hi,
I have a bunch of spectra which happen to show some sinusoidal baselines, like this:
I would like to subtract these baselines somehow. Of course I can try to fit sinusoidals using for instance Matlab's lsqcurvefit, but it seems like I have to be very careful with my input guesses not to...