On UAP Spectroscopy
( namely: the relentless attempt of Wile Coyote to capture Beep-Beep the supersonic roadrunner )
One of my colleagues asked me to post something about the use of spectroscopy with the UAP. Good question. I will try to tell something about my own experience.
In general, taking a spectrum of a UAP (of course at night) is not an easy task. It is often impossible to take a spectrum of such phenomena if they move too fast: in such a case the spectrum will follow the movement of the UAP, and so you will obtain many weak (and/or blurred) spectra at all the UAP positions in the sky. In some cases, if the movement is not erratic it is possible to realign all spectra by software and sum all together in order to increase the S/N ration. If the object is weakly luminous and if it is moving (even slowly and linearly) any spectrum is practically impossible. On the contrary, although being weakly luminous, if the object is standing still for a sufficient long time then it is possible to take a spectrum using relatively long exposures (integration time, technically) such as 30-60 secs or more.
The following is the most favorable situation:
The UAP is very luminous, relatively long-lasting and standing still or extremely luminous and moving with a very little angular velocity (due to its distance) – In such a case it is possible to obtain a good quality spectrum using a relatively short integration time (such as 1-5 secs). In such a situation it is possible to use alternatively a simple diffraction grating of good quality such as Rainbow Optics Spectroscope (ROS), Paton Hawksley, or Star Analyser (typically, all with 100 or 200 lines/mm: see on Google for all three), or (better, if you want a higher spectral resolution) a High Resolution Slit-Free Direct-Viewing Multiplexing Field Spectrograph (https://patents.google.com/patent/US8749781). I have used both solutions in my experience. What I can say is that if – using a simple dispersion grating – you simply take a spectrum of the light in the same field as the light itself, the spectrum will be of too low resolution (below R = 10^2) to be considered scientifically relevant (too many Angstrom/pixel). Of course, putting a dispersion grating in front of the lens and taking a photo at 50-70 mm is very easy and fast, but it is scientifically irrelevant, and line identification can be completely wrong. For instance, if there are spectral (emission) lines and these are very close together they will inevitably blend together to form a “bump” in the spectrum which will be impossible to identify. The only way to solve the problem here is to use a zoom lens of the 35-300 mm type: catch first the light with the spectrum and then (quickly) increase the focal length by moving slightly the lens (left or right in the sky) in order to have only the spectrum in the field, and then start the exposure. This is possible only if the operation is manual and not automatic (unless sophisticated software is used in order to pilot the camera). So, in this specific case, you get relatively few Angstrom/pixel, and the resolution can shift to 5×10^2 or a bit more. This (according to my past tests) is sufficient to resolve close spectral lines. Clearly you have to hope that the phenomenon maintains a high luminosity and a very small or absent movement during the exposure.
I think that the HRSFDVMFS spectrograph (patented by Dr. Ron Masters, with whom I collaborated in the past doing tests and calculations) in its newest version allows a spectral resolution of almost R = 10^4. Here you do not need to put the light source inside a slit for dispersion, and the light is caught for dispersion even if it is not centered. Clearly, using high resolution (the one I used was around R = 10^3, just medium resolution) like this requires not only more work in the post-processing phase as this spectrum is obtained “in pieces” in the “echelle” format (as we use in astronomy), but also a much longer exposure time (such as 60-180 minutes) during the spectrum acquisition phase. In reality, some long lasting and very luminous UAPs do potentially allow the use of this procedure: for instance, when taking spectra of Marfa lights or similar lights, including Hessdalen ones.
Then, you need to take a spectrum (using the same focal length, if possible) of a reference light (namely: a comparison spectrum), such as a streetlight which shows a line spectrum, whose wavelengths are well known previously. In my opinion, Sodium lamps are the best for this task, as you can identify lines quite easily. Mercury lamp (in case added with some other element) is more difficult to use, but not impossible. This operation is absolutely necessary when you have to calibrate your spectrum (which is only in pixel as you get it) in wavelength, using a high-order polynomial function after at least 5 or 6 lines are promptly identified. So, you first calibrate the reference spectrum and then you insert the UAP spectrum in pixel (being extremely careful to fit the extremes correctly), so that also the UAP spectrum is equally calibrated in wavelength. At this point, you can start to attempt line identification, being very careful not to confuse quantum fluctuations of the noise with true spectral lines (whose full width half maximum FWHM are typically a few Angstroms). After all this is done you have to normalize the spectrum to the continuum (rectification procedure) so that you have the zero point to measure the lines: namely to calculate via software the integral of the energy lines take from the continuum.
Regarding software to do all this I have used in the past Visual Spectroscopy (VSPEC: http://astrosurf.com/vdesnoux/ ). More recently I have used RSPEC (https://www.rspec-astro.com/ ), which I prefer as it is much more user friendly.
When all this is done it is good habit to export the table that gives the data points of the spectrum to an Excel file (or equivalent, such as more professional SMONGO or other soft). Clearly it is not possible to calibrate in wavelength a spectrum in pixel which is first plotted on Excel: only the reverse is possible.
Now, I should comment on what came out from all of this in my own (past) experience when I attempted to take spectra of UAPs.
First, on average, I was able to obtain a decent spectrum out of 5 attempts, in general. It is better to take many spectra of the same object using different exposure times (such as: 5, 10, 15, 20, 25, 30 secs and so on), until the spectrum appears to be sufficiently luminous.
Secondly, I could verify that spectra of UAPs can be very different from each other (depending on where the UAP appears in the sky, how much smog in the air, on intrinsic causes, etc). In most of the cases spectra appear to be a continuum without lines, but always a rather multi-peaked or wavy “strange continuum”, which resembles quite much fluorescent lights, in my opinion. In this case any real identification is extremely difficult (I was not able in these specific cases, probably due to too low resolution), and yes, it resembles pretty much human light sources.
As my attention was concentrated on spectral lines, what I didn’t do was another calibration that should be always done anyway: flux calibration (just before doing wavelength calibration) using a known source such as planet Venus and removal of the responsivity curve of the sensor used to detect the spectrum. In this case you can at least see if the spectrum resembles a black body (thermal spectrum), a power-law spectrum or other. Certainly, you can measure the temperature of the light source in such a way, even if no spectral lines are present, but the evidence of a multi-peaked continuum complicates all the thing. This doesn’t mean that it cannot be done, it means that you have to be prepared to face a lot of difficulties. In one case (a spectrum taken by a friend amateur astronomer, which I then carefully analyzed) Oxygen emission lines could be clearly identified, meaning that such a light source was a plasma able to excite/ionize the air. This is clearly the most favorable situation: if lines/blends can be well identified it is possible to determine the temperature of the phenomenon as well, and gas density too.
So, this is a very rough resume of my experience with UAP spectroscopy, i.e. something that is 10 times more difficult than star/galaxy spectroscopy. But it is feasible if you are a panzer, so to say, especially if you find the patience to stay all night awake and do manually the entire thing. If you want to do that automatically, then you need a lot of software or even A.I. But, be sure: in my humble opinion, photos that show both the light source and the spectrum are almost totally useless (unless strong well separated lines are seen).
Here below I list some of my tech papers where I discuss also spectroscopy of UAPs:
( pp. 33-34 and pp. 42-43 )
( pp. 6-11 )
( here and there inside the article )
( here and there inside the article )
PS – An ideal discovery from a scientific idealist
High-Resolution is strongly needed if we want to search for a Zeeman Effect in spectral lines (line splitting) due to a strong magnetic field in the plasma (plain plasma if it is a natural phenomenon or plasma surrounding a flying machine). We can measure the magnetic field intensity from line splitting, assuming that the magnetic field is sufficiently high to produce line splitting (not less than 10.000 Tesla, I think). In such a case, assuming that observations are done at night of a luminous object we would not probably need a magnetometer, as we would measure the magnetic field intensity at zero distance from the source. And, above all, we could carry out a sort of “dynamic back-engineering”, if that superstrong magnetic field is a consequence of some propulsion mechanism involving very high currents and superconductors.