### Debunking The CMB Curve

Here are a couple of blackbody curves to

give some idea of their shape.

When NASA's FIRAS instrument on the COBE satellite measured the CMB spectrum and gave this:

(click on the picture to give the link)

Everyone was happy because the curves look identical!

BUT, wait a minute, that wavelength scale looks a bit funny. Lets look at the divisions, 1, 0.2, 0.1, 0.07 don't they seem a little strange? A typical blackbody curve has intensity plotted against wavelength - with a linear scale. I mean, if it is an intensity/frequency diagram then why not put the frequency scale at the bottom? Why put wavelength at the bottom of your graph, with a strange non linear scale to boot, if you are plotting against frequency?

When it comes to the CMB, most curves are plotted against frequency. For instance, Ned Wrights pages plots the intensity against 'wavenumber [/cm]'. Why not use the traditional 'wavelength/cm'? They look similar but 'wavenumber[/cm]' is nothing like the 'wavelength/cm'. What Ned plots is the intensity against the 'number of waves in one centimetre'. Why 'one centimetre'? why not one metre or even number of waves in one second or even one week!

Is it because 'wavenumber[/cm]' looks like 'wavelength/cm' to a casual observer? No that can't be it - can it?

Lets look elsewhere, and go back to nasa.gov this curve plots intensity against 'waves/centimeter' Now, this is not 'wavelength in centimetre', but how many waves there are in one centimetre. You see when we draw graphs we commonly state the quantity divided by the unit. 'Wavelength measured in centimetre' would be written as 'wavelength/cm'. What they are all plotting is intensity against frequency curves - nothing wrong in that, but why don't they state that it is frequency in hertz, Hz, that they are plotting. Why all these referrals to things sounding like 'wavelength' when it is frequency they are plotting?

Now, I am not in the least bit cynical (!), but someone who was might think that it was all a con. That is, trying to make us think that these 'intensity frequency' graphs are 'intensity wavelength' graphs. The graphs 'look' like the traditional blackbody curves only when they are drawn against frequency. What does the Firas data 'look' like when it is drawn in the 'blackbody' way of intensity against wavelength? Lets have a look:

CMB curve                                                 Typical Blackbody curves

Not as convincing is it?

The point I am making is that there is no problem in drawing intensity frequency graphs as they do, but if you must do this, then call the graphs just that. Label the axis 'frequency/Hz'. Don't use 'waves/cm', or wavenumber[/cm] and certainly don't label the bottom axis in wavelength and then use a non linear wavelength scale.

If you are drawing an intensity frequency graph label it that way. The data is blackbody as stated. OK, the curve doesn't 'look' like a typical blackbody curve when you do that, but so be it. If the Big Bang Theory stands up to testing, then there should be no problem in drawing and labelling the graphs in the standard way.

HOME