Just under 4 weeks to go until the world’s spectroscopy experts convene right here at the Talavera Space Hut (my backyard!). Progress is being made. I was able to identify a suitable target for the session. The program organizers suggested I use a “Be” star as my target. That is a star with a B spectral type that has a hydrogen emission line in its spectrum, not necessarily just an absorption line although they can have both. That is a unique subset of the spectral type and an active area of research. These stars are not that difficult to find. There is a handy online catalog I found several years ago. Unfortunately the physics of this I have to defer to perhaps another post and is beyond the scope of this for now. My job here is to demonstrate how to capture a high resolution spectra with the Lhires III spectrograph. We will leave the analysis of the spectra for later.
The strategy is as follows:
- Capture your “calibration star” spectra. This is typically a bright type A spectral class star which has a very strong H alpha absorption line. Later you compare that spectrum with one from a professional database and adjust your spectrum to match that. This will eliminate errors due to for example water molecules in the atmosphere etc
- Capture your target star. We need to figure out what exposure time will yield about 80% saturation of the camera sensor. We take around 15 or so images
- Do all the calibration frames. The difference here with regular imaging is that all of the calibration can be done with the telescope parked and the cover on. Flat frames are done inside the spectrograph with a tungsten lamp you simply turn on. You do have to take a calibration lamp image which is an image of 3 neon lamp emission lines also done inside the instrument simply turning on a switch.
That’s basically it! Both the target star and calibration star are pretty bright so I found that exposure times were extremely short. The first run was with 40 sec exposures. The entire project takes about an hour! Compare that to a typical 30 hour regular imaging project! Probably as I get further into this I might do something that takes a lot longer but for the purposes of demonstration, this is ideal.
I was very confused about this guide field shown above. Apparently the center of the field shown by the reticle is NOT the center of the spectrograph slit. That is fairly common and I was told not a major issue. In order to center the spectrum on the sensor after exposing the main camera, I have to move the star to the location shown by the black arrow. I discovered this pretty much by trial and error. That isn’t that hard to do using the mount’s joystick. Centering the star on the slit is fairly easily done as well. Once the star is centered on the slit, I click on a small star in the same guide field and start the guiding. You could make the argument that guiding for such bright stars is not really necessary however you are taking multiple exposures and during the course of the 10 minutes or so that is happening the star could move off the slit center. The other issue is this weird field illumination that I mentioned in an earlier post. That is not actually the edge of the spectrograph slit (white arrow) but could very possibly be a shadow from the internal lamp as I was told. So again, not anything to worry about. Focusing is also interesting since I am just using the knob of the primary mirror on the Celestron. That is not particularly fine focusing! After several sessions though you learn how to improve it and even with a fine focuser I can assure you there wouldn’t be a huge difference with this optical system.
And now, the results from our first run:
Flats and darks were also obtained. We took the images at -10C and a series of 40 sec darks to be used for all of the images. The software can extrapolate for shorter exposures but not longer. The flat images are easy to take because all you do is turn on a tungsten lamp in the spectrograph and use that.
I did a rough processing of the spectra to see if we’re on the right track:
Awesome! A high resolution absorption spectrum! Unfortunately something is wrong. The H alpha absorption line is well known and occurs exactly at 6563 A. NOT where it is shown here.
Ok so what now? I reviewed everything we did and looked at the calibration lamp lines again. I compared the ones I took to a reference I found on Christian Buil’s website and began to realize something!
The longest wavelength line 6599 which is off by itself is supposed to be on the far right. You can see here it is reversed. Doh! My camera was upside down!! To confirm that I went back to the telescope during the day, took the cover off and pointed to a random region of sky. Doing that you should be able to see absorption lines from the Sun.
Now if you turn the micrometer counterclockwise increasing the numerical values, you should be moving the spectrograph’s “window” toward longer wavelengths which means the absorption lines should be moving to the left. By convention a spectrum has blue or shorter wavelength on the LEFT and the longer wavelengths are to the RIGHT. This is in fact the exact opposite of what I observed! My suspicions were correct. The camera needs to be flipped 180 degrees!
Back to round 2. We need to reorient the camera correctly and adjust the spectrum as it is recorded so that the spectral line is horizontal. Stay tuned….and thanks for reading.