On January 17 and 18, 1996 an algorithm review for the Stratospheric Aerosol and Gas Experiment (SAGE) III was conducted at the Langley Research Center (LaRC). M. Patrick McCormick, SAGE III Principal Investigator (PI) from LaRC, chaired the review.

Recently, two Russian scientists were nominated to joint the SAGE III Science Team, namely: Nikolai P. Elansky, Russian Academy of Sciences, and Albert A. Chernikov, Central Aerological Observatory. The Russian delegation at the meeting included Elansky, Yuri A. Borisov, and Oleg Postyliakov.

The SAGE III Science Manager stated that the primary objectives of this review were to: 1.) obtain the Science Team's concurrence on the algorithms and to ensure agreement on the approach taken in the Algorithm Theoretical Basis Documents (ATBD); and 2.) decide on any alternative channel approaches.

The secondary objectives of this meeting were to discuss critical spectroscopy and to provide information on data products, telemetry, mission operations, software development, data processing, configuration management, etc.

William P. Chu (LaRC) gave an overview of the development plan for the ATBDs. The ATBD schedule was presented and the formal ATBD review in November 1996 was emphasized. Two versions of the software have to be delivered: the engineering version by early 1997 and the flight version by early 1998. Ultimately, the software for the algorithms is to be delivered to the Distributed Active Archive Center (DAAC) for processing the SAGE III data.

Four of the nine ATBDs (aerosol, cloud, ozone, and OClO) have been made available to the science team for internal review. The other five will be made available for internal review by the end of March.

There are essentially four different types of algorithms: SAGE II-type, simultaneous (water vapor), global and simultaneous (oxygen), and differential retrievals (mostly lunar). The general steps in the solar and lunar retrievals were described.

The SAGE III Standard Data Products Table taken from the NASA EOS Execution Phase Project Plan was presented. This table is scheduled to be baselined within the next few months. Several comments were made that clarification was needed for some of the columns and could be provided as notes at the bottom of the table. Derek M. Cunnold (Georgia Tech) said that the temperatures (2 K at 70 km) on the table seemed to be too optimistic. Joseph M. Zawodny (LaRC) responded that the accuracy on the table is not representative of all altitudes but are the "best" that we can do in particular altitude ranges.

Benjamin M. Herman (University of Arizona) presented an overview of the prototype algorithm and described the forward problem for a solar occultation experiment.

Chu followed with a description of the solar occultation measurement geometry, contributions to extinction, typical sunrise science data output, and concluded with the refraction calculation. The refraction calculation will use the same procedure as that used with SAM II, SAGE I, and SAGE II, which is based on analytical solutions to the refraction integral for any given temperature profile. Two important areas that need attention are decoupling of gases during the retrievals and how accurately the species can be separated.

Zawodny discussed both the radiometric and positional calibration techniques. He stated that any changes in wavelength in the spectrometer could only be induced by temperature changes in the grating and the charge coupled device (CCD) detector. The current plan is to perform a readout across the sun at an altitude of 150-180 km and perform a wavelength fit to the exoatmospheric solar spectrum. Cunnold asked if we could determine the degradation of the filter function well enough to do NO2 trends and Zawodny replied that there was no reason to think that there will be degradation since the bandpass is determined by the CCD geometry. Zawodny stated that there will be no way to monitor the 1.55 micrometer channel wavelength in flight. Chu added that the 1.55 micrometer channel is broadband and slight shifts should not be too disruptive. Zawodny stated that the users of this channel must account for possible degradation of this channel, and Chu added that we will have ground-based instruments to monitor for degradation.

The scan mirror scans the sun field-of-view (FOV) in a zigzag fashion. Spectral calibration of the relative reflectivity is maintained over a range of angles by looking at the Sun from 300 down to 100 km. A small slope is both apparent and expected in the measurements at very high resolution (1-2 counts out of 4000).

Next, Zawodny explained how we determine where the instrument is looking. The global positioning system (GPS) will tell us where the instrument is; and we know the position of the Sun and Earth. Refraction can be accounted for with the National Meteorological Center (NMC) data initially. Also, the scan mirror scans at a constant rate. Given these, the tangent altitude and position of the FOV on the Sun will be calculated. Spacecraft attitude, rate, and acceleration can also be determined.

Zawodny stated that if the scan hits a cloud, the scan waits (0.1-0.2 seconds) then, if the Sun does not reappear in the FOV, the scan mirror reverses direction. Other discussions on the spacecraft maneuvers during data taking, degradation of the sunseeker over time, vertical resolutions, etc., were also held.

Mark C. Abrams (Science Applications International Corporation (SAIC)) presented a comparison of the SAGE II/SAGE III retrievals. He presented a flow diagram for each of the SAGE III solar occultation measurements and the lunar occultation measurements. David H. Rind (Goddard Institute for Space Studies (GISS)) pointed out that we should indicate that the NMC data are being used as a "first guess" for the temperature in the SAGE III algorithm and then the actual SAGE III temperature measurements will be used. Steven C. Wofsy (Harvard University) pointed out that we should make sure that we are doing things correctly before we drop the SAGE II-type inversion, which used NMC data throughout.

Abrams discussed the approach planned for the aerosol and ozone retrievals. He emphasized the importance of maintaining continuity with the SAGE record and stated that minimal changes between the SAGE II and SAGE III algorithms will be made.

Er-Woon Chiou (SAIC) presented the water vapor retrieval by describing the forward problem, calculation of the slant path transmittance, removal of the interfering species, inversion of water vapor, and simulation studies. Rind pointed to the SAGE III Standard Data Products Table again. It shows water vapor down to 3 km. He said that we should be able to measure to the ground (0 km). The science team agreed that this table needed to be updated to accurately reflect what we plan to deliver. The emissivity growth approximation (EGA) method will be used to calculate the slant path transmittance.

Michael C. Pitts (SAIC) presented the need and algorithm approach for retrieving temperature and pressure profiles. These measurements allow SAGE III constituent retrievals to be independent of external data products; allow for the presentation of data on pressure surfaces and in mixing ratios; and provide a self-calibrated, accurate temperature data set for trend studies. The retrievals utilize the oxygen A-band located in the visible region centered near 760 nm. The spectra will be measured from 740 to 780 nm with 2-nm resolution. The retrieval process utilized will reduce the solar radiance measurements into atmospheric transmittance profiles at each wavelength, calculate the slant path transmittance using the EGA method, remove the Rayleigh scattering, separate several species (aerosol, ozone, etc.), and invert the temperature and pressure. The temperature and pressure inversion approach using the Carlotti global fitting was described and simulated retrievals were shown.

Zawodny presented the lunar occultation retrieval approaches for the O3, NO2, NO3, and OClO species. He explained a typical lunar occultation event and showed transmission versus wavelength profiles for O3, O2, NO2 and H2O with continuum at a tangent altitude of 20.0 km. Retrieval error estimates based on expected instrument performance were also shown.

Geoffrey S. Kent (Science and Technology Corporation) provided a status report of the cloud ATBD and discussed the cloud algorithm and data product. The cloud science objective is to identify the presence of cloud in each event at all altitudes between 6 and 30 km. The theoretical basis for the cloud algorithm is different from the other data products because clouds are extracted from the aerosol extinction data. The algorithm proposed uses aerosol extinction data at 0.525, 1.02, and 1.55 um and relies on the wavelength variation in extinction to distinguish aerosol from cloud. Two methods which were used to separate the aerosol and cloud in the SAGE II data set were explained, and then the proposed method of separating aerosol and cloud in the SAGE III data set was explained. The proposed SAGE III algorithm performance was better than the SAGE II two-wavelength method but is questionable at times of strong volcanic activity.

Zawodny presented an alternative channel selection for the CCD, which would move channels into the Chappuis bands for ozone (differential ozone channels). Comparison charts of the current versus proposed plan were shown for the science telemetry volume and CCD utilization. Obie Bradley (LaRC) will use Zawodny's proposal to conduct a feasibility study. Phil B. Russell (NASA Ames Research Center) showed similar channel selections for his airborne system. Wofsy stated that differential measurements were what we should be doing if feasible. At this time, Wofsy added, there is a fair amount of OClO present during the twilight and that 3 additional channels for OClO might be worthwhile as well. It was decided that Wofsy's OClO computations would be used by Zawodny to do a feasibility study of a solar OClO measurement.

Eric Shettle reported on the quality of the laboratory spectroscopy for the SAGE III measurements. There is disagreement in the evaluation of Ritter and Wilkerson data for oxygen (760 nm) and the differences in ozone data will show up as "structure" in differential measurements. It was agreed that a consensus is needed for SAGE III spectroscopy measurement priorities. This priority list will be presented at the next EOS Investigators Working Group (IWG) meeting.

The consensus of the Science Team was to incorporate Zawodny's alternative channel plan (differential ozone and relocated aerosol channels, etc.) into the ATBDs. The solar retrieval of OClO, using three additional channels will be added into the ATBD as a "Research Product" and Zawodny will make an assessment of its feasibility.

For species which vary rapidly during sunrise or sunset, the slant-path column amounts will be a "Research Product" and should include neutral density, path length, and refraction correction angle.

Scott R. Quier (SAIC) gave an overview of the software development requirements placed on SAGE III as an EOS instrument. The software development process will consist of a series of builds. This approach has the advantage of limited risk and readily available executable code. There will be no beta release for the software.

Abrams presented the potential ancillary products such as potential temperature and geometric altitude versus geopotential height. There were discussions about derived products being archived versus giving the equations to the users. Also, some data like the NMC data is already available at the DAAC so there would be no need to store it with the SAGE III data set as well.

Mike S. Cisewski (GATS, Inc.) gave an overview of the mission, data formats, and the CCD data. For the SAGE III/Meteor-3M mission, the command station is located at the Russian Space Agency Mission Control Center. The command link is from LaRC to Moscow and the data link is from Wallops Flight Facility to LaRC. There are five basic data formats: low rate engineering, solar science, solar line calibration, lunar science, and ancillary data. During the mission, each day's science data are transmitted to both Russian and U.S. data stations twice every 24 hours. On the U.S. side, the data are transmitted to the Wallops Flight Facility, temporarily archived, data quality checked, and sent to LaRC. Cisewski also discussed the CCD data formats, channel selection, and spare channel availability. He assured the scientists that his goal was to get the maximum amount of science data down. The Mission Operation Center (LaRC) will perform Level-0 processing and send this data to the LaRC DAAC and the Science Computing Facility.

Michael W. Rowland (SAIC) presented an overview of the approach and methods planned for SAGE III Level-1 and Level-2 processing.

Patricia L. Erickson (SAIC) presented an overview of the data management test flow and the configuration management flow.

Anne C. Edwards (SAIC) has a preliminary Home Page underway. Plans are to have the ATBDs available on the Home Page so interested parties can download the documents at their home sites whenever the newest version hits the street.

[Table of Contents] [Previous] [Next]