From: IN%"offiong@irg.usask.ca" "JJ Offiong" 13-NOV-1996 21:20:28.27 To: IN%"STEELE@DANSAS.USASK.CA" CC: Subj: RE: Please submit text for proposal sections Return-path: Received: from irg.usask.ca by DANSAS.USASK.CA (PMDF V5.0-6 #15020) id <01IBT88FVQGW000LXE@DANSAS.USASK.CA> for STEELE@DANSAS.USASK.CA; Wed, 13 Nov 1996 21:20:26 +0000 (GMT) Received: by irg.usask.ca from localhost (router,SLmailNT V2.0); Wed, 13 Nov 1996 15:20:20 -0800 (PST) Received: by irg.usask.ca from frey.usask.ca (128.233.17.27::mail daemon,SLmailNT V2.0); Wed, 13 Nov 1996 15:20:20 -0800 (PST) Date: Wed, 13 Nov 1996 15:20:19 -0600 From: JJ Offiong Subject: Re: Please submit text for proposal sections To: STEELE@DANSAS.USASK.CA Message-id: <19961113152020.1042745e.in@irg.usask.ca> Organization: University of Saskatchewan MIME-version: 1.0 X-Mailer: Pegasus Mail for Windows (v2.42a) Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Priority: normal Comments: Authenticated sender is Hi David, John and I have started working on the Ground-Based Data Handling Team report. This is a rough outline, with gaps and sections missing, but you may want to include it for the sake of completeness. Sorry it's late, but I assumed John would send you a copy and he assumed I would. J ***** Ground-Based Data Handling team report ====================================== From our perspective, there are numerous tasks to be grappled with. There is: data decompression from the telemetry (picture reconstitution) spectral inversion (from picture to ozone absorption values) tomographic inversion (from ozone absorption values to O3 distribution reconstruction) Data Decompression ------------------ Compression algorithm Decompression algorithm Associated data losses Spectral Inversion ------------------ Error Sources The reconstituted images captured by the CCD will be the result of more than just O3 absorption. Other absorption phenomena attenuating the starlight reaching the CCD will include Rayleigh scattering, NO2 absorption, and Aerosol scattering. In addition, instrumental noise and dark current phenomena will be present to some extent. Each of these are addressed below. -Rayleigh scattering Rayleigh scattering, caused by the presence of molecules between the star and our instrument, will be present to some extent in our 'unocculted' views above the ozone layer, and will increase as the instrument descends. This phenomenon varies as ... (formulae, variables, etc.) (estimate of significance in CCD images) (removability from the data) -NO3 absorption NO3 is also present in significant amounts above 20km altitude. (concentration by altitude, citations) (absorption levels expected) (estimate of significance in CCD images) (removability from data?) -Aerosol scattering Probably the most unpredictable influence on the CCD images is the effect of airborne particles in all levels of the atmosphere. (determining variables) (variation by altitude) (expected significance) -Instrumental sources of error In addition, instrumental optical phenomena, noise from the intensifier and ... and dark currents will all be present to some expect. (expected optical phenomena and significance) (expected noise sources and levels) (expected dark currents and significance) -Cumulative estimate of non-removable error (transfer function magnitude, character, etc.) Spectral Inversion Method -algorithm -effectiveness -speed estimate Tomographic Inversion --------------------- There are a variety of criteria and approaches to consider in designing the reconstruction process once spectral inversion has been completed. To understand the process, we will make some simpli- fying assumptions, and assess the validity of each of these assumptions. Simplifying assumptions - vertical linear instrument trajectory (True as long as terminal velocity > expected horizontal wind speeds.) - an appropriate distribution of stars of appropriate brightness and constancy (Validity determined with statistical analysis of stellar sphere) (Optimizable by choosing best zeniths and therefore dates and times.) - stationary globe wrt stellar sphere (Not true but completely predictable. Can be added later.) First level of processing The most rudimentary information available from the occultation data is a one dimesional variation in the ozone column concentrations within the plane defined by the trajectory of the instrument and each star viewed by the instrument. Each star viewed by the instrument both above the ozone and through some part of the ozone provides this data in its own instrument-trajectory--star-position plane. The collection of the data in these planes can then be reconstructed into 3-D ozone distributions using other techniques. A 2-D Reconstruction option If a purely vertical variation of ozone concentration were assumed with each star plane, Abel's equation could be used and a lower triangular matrix generated in order to find calculate what that vertical ozone vari- ation is. This vertical variation would be different for each star plane. The vertical ozone variations could then be interpolated between adjacent star planes to give the variation in vertical ozone distributions with res- pect to azimuth. 3-D Reconstruction options -Fourier reconstruction The collection of star planes for which 1-D ozone variations is might yield a crude three dimensional reconstruction by finding the Fourier coefficients best approximating the variation in ozone concentration around a circle of a given altitude and radius centered on the instrument trajectory. -Moving Sector reconstruction If we pick a small angle and assume that the variation in the ozone distribution within a sector of that size does not vary significantly across its width, then we can treat stars falling within an azimuthal sector of that width as though they were all in the same plane. This permits us to get two or more look-directions within the plane affording a significant foundation for reconstruction of the ozone as it varies with distance from the instrument. Doing this repeatedly, moving the sector boundaries a few degrees each time around the domain should enable us to reconstruct the entire domain in three dimensions. -Other reconstruction techniques __ "You're not going crazy, Arthur. You're going __ sane in a crazy world." - The Tick J Offiong Phone: (306)242-1972 E-Mail: