The MISR Science Team met August 12-14, 1996, at the Jet Propulsion Laboratory in Pasadena, CA. As in previous years, the MISR annual meeting was an opportunity for all members of the team to obtain the latest information regarding the MISR instrument status. This year was also an opportunity to ratify MISR's existing data analysis algorithms and to discuss the science research possible, given the emphasis on the Level 2 and Level 3 products. The following text briefly describes the focus of the meeting and highlights the most important aspects of the proceedings.
Introduction
The Principal Investigator, Dave Diner, opened the session, bidding welcome to the team and summarizing resolved issues from the previous science team meeting. He then gave a brief overview of the meeting's objectives: 1) to update the team on instrument and software development status, 2) to identify critical time-sensitive issues, 3) to develop a research plan, thereby honing the MISR science emphasis, 4) to review test data requirements, and 5) to refine MISR's validation strategy.
Instrument Status, Software Development Status, and In-flight Calibration
Terrance Reilly gave the instrument and project status report. The construction of the MISR instrument is approaching completion with all flight subsystems, except the digital electronics, ready for integration. The previous problems with the digital electronics, namely the field-programmable gate arrays (FPGAs), are fixed, but have left little flexibility in the schedule. However, the MISR project expects to deliver an excellent scientific instrument by the scheduled delivery date while staying within budget. Francesco Bordi, EOS Project Office, continued by placing MISR's development in context with a description of the overall progress of the AM-1 project.
The science data system status was presented by Graham Bothwell. In the last year, MISR activities in this area have expanded, with staffing levels being increased to nearly full complement and system specification and development work progressing on defining the overall software system for Version 1. This delivery of the software system is expected to be as fully functional as possible, building upon the successful delivery of the earlier Beta Version in March, 1996. Most of the software analysis and design has been completed for Level 1, which includes the raw data handling and geometric and radiometric calibration procedures. For the Level 2 software, which performs the science data processing, final requirements analyses, based upon the completed ATBDs, are well underway. The production of MISR's ancillary data sets is expected to be completed in the next few months, while other issues, such as browse data and quality assessment, are being further defined. Progress also continues with the MISR home page on the "web" at http://www.misr.jpl.nasa.gov.
Carol Bruegge continued the meeting by describing the status of the MISR cameras and plans for in-flight radiance scaling and conditioning. Dynamical, thermal, radiometric, and spectral testing of 10 cameras (nine flight and one spare) has been successfully completed. With some additional ground processing to reduce out-of-band errors and to perform a point-spread-function deconvolution, the cameras meet all requirements. Once in orbit, the MISR cameras will continue to be calibrated using the on-board calibrator, vicarious calibration relying on field measurements, histogram equalization, and trend analyses which incorporate the preflight calibration data. Veljko Jovanovic continued the discussion on in-flight calibration by detailing MISR's plans to perform geocalibra-tion and coregistration using ground targets.
MISR Algorithm Development and Associated Science
As opposed to previous years, this section of the meeting focused on ratifying the existing MISR algorithms and discussing how a better link could be forged between the MISR products and proposed research topics.
Top of Atmosphere/Cloud
This session began with an overview presented by Roger Davies. In the last year, MISR cloud detection and classification algorithms were finalized, and significant progress was made on the design of the Level 2 software. New cloud masks, using stereoscopic and radiometric techniques, were also defined. These methods have been successfully tested on simulated images of Mexico and Hawaii, projected to the view angles of the MISR cameras. Peter Muller showed examples of applying MISR stereo retrievals to Along Track Scanning Radiometer (ATSR-2) data. Davies then described the methodology used by MISR to determine local albedos from which are calculated expansive and restrictive albedos.
Chris Borel and Sig Gerstl presented their work on calculating clear sky albedos. Their method utilizes a semi-empirical model which is fit to the observed radiances and then used to extrapolate to sun and view angles not observed by the MISR cameras. Comparisons with simulated MISR data, generated from multiple-stream radiative transfer codes, suggest that retrieval errors can be constrained to less than 1.5%.
Aerosol/Surface
John Martonchik started his briefing regarding MISR Aerosol/Surface issues by reviewing the MISR aerosol products. He then described the data sets crucial for determining MISR aerosol parameters, namely the Simulated MISR Ancillary Radiative Transfer (SMART) data set and the Aerosol Climatology Product (ACP) data set. The SMART data set contains the radiative transfer parameters for ten aerosol pure particle types. These types are combined, using a modified linear mixing theory, to reproduce the scattering properties of particle mixtures. The ACP is a collection of three files describing aerosol optical properties and the likelihood of an aerosol being observed. The SMART data set has been successfully built and is in the process of being tested while the ACP is under construction. Having reviewed MISR surface products, Martonchik continued by describing the algorithms and models used in their determination.
The MISR aerosol retrieval calculation uses a modified linear mixing method, which assumes that the scattering properties of an aerosol mixture can be estimated by linearly adding the contributions of the individual aerosol components. The standard linear mixing method fails if the mixture's components differ in their absorption properties. Wedad Abdou presented results showing that the modified method is a substantial improvement over the standard technique.
Since MISR aerosol retrievals are performed over land as well as ocean, it is necessary to determine what terrestrial surfaces are suitable for aerosol retrieval, and to construct simple Bidirectional Reflectance Function (BRF) models. Michel Verstraete presented work carried out with Bernard Pinty and Ola Engelsen regarding the most suitable land type--dense, dark vegetation (DDV). Their parametric BRF model adequately reproduces observations of reflectance factors from wheat fields to hardwood forest canopies over a range of sun zenith angles.
Ralph Kahn presented his summary of the sensitivity of the MISR instrument to aerosol properties. Preliminary studies indicate that column extinction optical depth can be retrieved to about 0.05 or 10%, whichever is larger, under a wide range of sky conditions. The aerosol retrieval will also be able to distinguish among many common particle types, which represent constraints on a combination of particle shape, size distribution, and composition. For example, MISR can distinguish spherical from non-spherical particles over calm ocean.
Tom Ackerman discussed results from the Atmospheric Radiation Measurement (ARM) Enhanced Shortwave Experiment (ARESE) field campaign. Comparison of observed-to-computed irradiance suggests that calculations overestimate the diffuse component by 5-10%. The only way to match all the observations is to include an unknown gaseous absorber that absorbs predominantly towards shorter wavelengths. This absorption could possibly be due to dissociation and/or fluorescence of trace gases such as NO 2. The implications of these results for MISR are yet to be determined.
Ranga Myneni continued the science discussion, describing results from AVHRR studies of the global vegetation index. Evidence suggests that the photosynthetic activity of global vegetation increased from 1981 to 1991 in a manner consistent with a prolonged growing season. The correlation with the increase in atmospheric carbon dioxide suggests that the carbon cycle has responded with fluctuations in temperature which are small on the global scale but have great regional significance. Ranga further described some of the MISR/MODIS surface product synergy.
Test Data
Test data can be considered as having two basic functions: 1) to test the scientific validity of a certain algorithm, and 2) to demonstrate that software produces expected results in accordance with specified requirements. Robert Ando described the methodology being used by the MISR team, emphasizing how test data can be used to ratify software. MISR tests range from unit tests which verify the functionality of individual executable components to full system-wide tests. Dave Diner led the discussion to decide upon a plan to improve upon the existing test data sets from both a software and scientific perspective. In addition to simulations, data from AirMISR, the airborne version of MISR, may give the closest approximation to MISR observations in the pre-launch time frame.
Level 3 Products
Dave Diner opened the discussion by describing the "at launch" Level 3 products. MISR products will be spatially-and-temporally-binned Level 2 parameters expressed monthly on the global equal-angle-1 grid adopted by the AM-1 instruments.
"Post launch" products will be much more mature, being more in number and possessing more intermediate spatial and temporal resolution. They may also be reported on a more-advanced grid. Jon Kimerling presented an alternative gridding scheme using hexagonal and pentagonal cells of equal area. Such a grid has the advantage of requiring no special projection for the poles, having no geographical singularities, and being nestable over many orders of magnitude in spatial scale. However, this gridding scheme has not yet been adopted since several implementation details need to be resolved.
Ralph Kahn continued in the "post-launch" theme by leading the discussion regarding strengthening the link between the team's specific science goals and the Level 2 and Level 3 products. These issues ranged from establishing and comparing climatologies to regional "process" studies. Each team member was charged with outlining specific research topics and describing which products would be necessary for the research.
Validation
Pre-launch validation efforts concentrate on algorithm validation and technique development, while post-launch efforts focus on MISR product validation and vicarious calibration. Jim Conel discussed possible field campaign agendas and how they may be combined with instrument networks. Since MISR cloud validation is necessarily difficult and as yet in the nascent stages, it was suggested that future planning efforts be concentrated in this area. Tom Ackerman suggested that MISR cloud validation efforts emphasize climatologies by comparing retrievals with long-term databases at existing measurement sites. Such sites should possess radar and lidar instrumentation, as these methods are the most preferable ways of quantifying cloud heights and motions.
A major advance in the last year, relevant to field instruments, has been the development of an airborne version of MISR (AirMISR). Tom Chrien discussed AirMISR's design, its abilities, and its relationship to MISR itself. Constructed from a spare MISR camera, AirMISR yields pushbroom images with views ranging from 70deg. forward to 70deg. aft by utilizing a gimbal system. Flying aboard an ER-2 aircraft at an altitude of 20 km, it will produce nadir-view images about 10 km on a side with 7 m resolution. Construction and engineering tests are expected to be completed in 1996. AirMISR's primary mission is to collect MISR-like data sets to support development and validation of MISR parameters. AirMISR will also provide an additional radiometric calibration path to assist with in-flight instrument characterization.