Besides all the good ideas that had been written here. I think it's posible to add an equivalent target plane to your system and get a diagnose of the both the beam waist and the plane surface. For this, if it's possible, add a beam splitter before the lens you are using to focus the beam and this will redirect a portion of ligth that comes from the reflection in the surface, so you can acquire an image of the beam waist.
We tried this while doing some laser material machining.
Take in mind the changes in power/ intensity/ fluence in the interaction due the addition of optics.
You can readilly observe CO2 laser spots up to power densities of 16 W/cm2 using thermal imaging plates found at http://www.cascadelaser.com/thimpl1.html . At higher denities you can use thermal imaging paper as used in thermal printers to take a burn pattern.
Besides all the good ideas that had been written here. I think it's posible to add an equivalent target plane to your system and get a diagnose of the both the beam waist and the plane surface. For this, if it's possible, add a beam splitter before the lens you are using to focus the beam and this will redirect a portion of ligth that comes from the reflection in the surface, so you can acquire an image of the beam waist.
We tried this while doing some laser material machining.
Take in mind the changes in power/ intensity/ fluence in the interaction due the addition of optics.
there are a lot good ideas here. It just depends on if you can spent money for it or you want to try it on your own. On your own I think that the idea with the beam splitter is good. You could try to lower the intensity far enough so you could use a CCD camera. Since I luckily have access to Focus-Monitor build by Primes, I can measure spot size and a caustic directly. Here's a link to that:
First of all, it depends of the working regime of your laser: if it is pulsed or cw. Let's suppose it is cw. 100W beam power is not so high.
You can measure the spot size and the mode structure of the beam by several methods. One of the most recommended is the rotating needle method which consists in introducing a reflecting thin piece in the emitted laser beam and observe/measure the dimensions of the reflected beam that will have only some percents power of the emitted laser beam. By adjusting the dimenisons of the cross section of the needle, the rotating speed, etc you may find the optimal conditions to measure the spot size, the mode structure, etc
I think the most straightforward method is to use a piece of polished plexiglas (polymethylmetacrilate, PMMA) in the focal plane. The beam readily leaves its printout and power density is conveniently higher than for thermal printer paper. I suggest to tap 4% of the beam first, using NaCl or KCl wedge or whatever is handy. Smells bad, but quick, cheap and reliable.
I would propose the same method that Paul Mikheyev indicated, particularly if your CO2 laser is a pulsed laser: at each pulse, the laser will ablate the plexiglass plate. After several pulses, the depth of ablation will be proportional to the fluence of your laser beam. We experiented this method on large discharge CO2 lasers in the 80th.
The knife edge method works well as suggested by Leopoldo Martin. I'll just add that Spiricon makes an automated scanning slit profilometer which we use for this exact application of measuring a ~100 um spot from a 100 W CO2 laser. Make sure to get the one with the pyroelectric detector. http://www.ophiropt.com/laser-measurement-instruments/beam-profilers/products/slit-based-profilers/nanoscan