From: IN%"huber@isas11.usask.ca" "Matthias Huber" 12-NOV-1996 20:39:10.50 To: IN%"steele@DANSAS.USASK.CA" CC: Subj: summary of my work Return-path: Received: from isas11.usask.ca (isas11.usask.ca) by DANSAS.USASK.CA (PMDF V5.0-6 #15020) id <01IBRSHWM1OG000KXI@DANSAS.USASK.CA> for steele@DANSAS.USASK.CA; Tue, 12 Nov 1996 20:39:08 +0000 (GMT) Received: from localhost (huber@localhost) by isas11.usask.ca (8.7.6/8.7.3) with SMTP id OAA21403 for ; Tue, 12 Nov 1996 14:36:34 -0500 Date: Tue, 12 Nov 1996 14:36:33 -0500 (EST) From: Matthias Huber Subject: summary of my work To: steele@DANSAS.USASK.CA Message-id: MIME-version: 1.0 Content-type: TEXT/PLAIN; charset=US-ASCII Content-transfer-encoding: 7BIT Ozone: We have to expect a Ozone and Temperature distribution at Churchill as follows: (Data taken from Air Force Cambridge Research Laboratories: Ozonesonde Observations Over North America, Volumne 4, Dec 1967) Jan 16th 1965: peak at 20 km altitude: partial pressure of ozone: 0.25 mmb, at 30 km " : " : 0.07 mmb Nov 24th 1965: peak between 18 an 25 km: almost a plateau with 0.16 mmb at 25 km: 0.09 mmb Oct 21st 1965: peak at 18 km " : " : 0.18 mmb, at 30 km " : " : 0.05 mmb For all three measurements the partial ozone pressure ranges between 0.03 and 0.01 for altitude < 10 km. Temperature: The temperature distribution looks almost the same for the three measurements: It decreases from the ground temperature (typically 260 Kelvin) to about 215 Kelvin at 10 km altitude. In the 10 to 30 km range the temperature doesn't change significantly. We therefore can expect a constant temperature in the region of interest. The instruments have to function at least down to a temperature of 200 K. If we want to measure down to the ground we need to know the temp dependence of all our components in advance (because of the possibility of loosing or damaging the instrument). I think one can safely say that the range > 40km can be used for calibration of the instrument. The second ozone peek that exists is neglectible due to the very low pressure at that altitude and therefore the very low absolute number of ozone molecules. Airglow/Auroral contribution: A launch from Churchill automatically involves an auroral contribution because Churchill is located right in the auroral zone: 68 degrees north (AACGM-model). Therefore an increased intensity at 557.7 nm and 630.0 nm is expected. The following approximations to solve this problem were discussed: 1) only use the wavelength between 557.7 and 630.0 nm. Problem: we limit ourselves to a quite narrow wavelength band. And we cannot use the whole Chapuis-absorption band. 2) use notch filters to minimize the intensity at these wavelengths. Problem: The characteristic of notch filters includes higher harmonics that absorb up to 30% at wavelengths different from the ones we want to exclude. They also have a 'large' halfwidth. Therefore we exclude a whole range of wavelengths from our spectrum. 3) include the whole spectrum and don't bother about the increased intensity. Problem: Background noise on CCD: Actual dark areas are recognized as being green (and/or red). The pixel that recognizes the aural wavelengths can not necessarily be used because intensity of aurora can change. Solution: One can use a single pixel as a kind of notch filter: always exclude the pixel that registers the auroral wavelength(s). Therefore a high resolution would be better. We decided to go with 3) because we want to use the whole feature of the Chapuis-band and we want to keep our option open to detect other species. The Soulution to 3) just came to my mind now, maybe we discussed something like this before and I just forgot about it. Optics: Our aim is to obtain 0th and 1st order on one CCD. This means that we should go with a grating that hasn't a lot of gratings/mm. With a 40 deg. field of view, an 80 slits/mm grating and a wavelength band between 450 nm and 750nm we have to cover a total field of view of 43.7 deg. 100 slits/mm would still keep us under 45 deg. used formula: sin(theta)= lambda*1000*n+sin(alpha) where n: slits/mm lambda: wavelength alpha: angle of incidence (max. 20 deg) theta: outcoming angle Because stars are not very bright, we should possibly use all the light efficiently. Therefore we need a lens that doesn't absorb a lot of light, i.e. it has to have a small f-number: f# = f/d, where f is the focal length and d is the diameter of the aperture. In principle only a Double Gauss lens is possible for our conditions (45 deg. field of view and small f#). (see Warren J. Smith: Modern Lens Design page 20). The analysis of different graphs of lenses showed that on always has to compromise. One lense with an f# < 1.0 has a very bad aberration. The same argument holds vice versa: good aberration but bad f#. The conclusion is that we use a lense with an f# of 1.4 or 1.5. They can be choosen depending on our exact field of view we want to detect. Clearly we have to focus on a Double Gauss lense. A ray-tracing should be done with a advanced ray-tracing program, where on can change parameters easily or where one can even program parameters and then the program gives a design. Matthias Huber