On August 28, 1997, the EOSDIS Core System (ECS) Team demonstrated the data capture and processing functions critical for timely support of the upcoming EOS AM-1 and Landsat-7 missions using ASTER, MODIS, and Landsat-7 Instrument Team-provided data and Product Generation Executables (PGEs). Requirements were translated into 42 critical functions, all of which were successfully demonstrated to NASA and the science community. Highlighting the results of 3 months of focused integration effort following the successful ECS May Demo, the demo scenarios tested system functionality including data ingest and conversion, product generation, on-demand data acquisition, user registration, product subscription, and product order. The scenarios also emphasized instrument data handling as well as data system archive, search, and retrieval capabilities.
The ASTER scenario presented by Lynne Case (Interoperability and Data Management Subsystem lead) tested support from the Data Acquisition Request Tool (DART). This tool was developed to provide on-demand data acquisition as well as expedited requests for data access. A PGE automatically triggered processing of additional PGEs to produce the requested granules. The system also allows for backward chaining so that users may request end products without having to specify the intermediate products required to create the needed end product. Additionally, the ASTER data were used to demonstrate the ability to update Quality Assurance (QA) metadata at the DAAC using the QA graphic user interface.
Evelyn Nakamura (Science Data Server lead) demonstrated the scenario which showed the exercise of the MODIS instrument launch-critical automatic scheduling and archiving capabilities. These capabilities began with a production chain starting with a synthetic MODIS MOD_PGE 01. PGE 01 staged and destaged 1.7 Gbytes of data and upon completion automatically triggered two instances of MODIS Level 1B PGE 02, which subsequently destaged 2.0 Gbytes of data. When PGE 02 completed execution, the resulting output file served as input for PGE 08 for the generation of a MODIS Level 2 Sea Ice Product. The B0 Search and Order Tool (B0SOT) was used to search and display inventory metadata to verify that the products were properly inserted into the archive. Additionally, outputs from a failed PGE were pushed via ftp to a simulated Science Computing Facility (SCF) where the data could then be examined to determine the cause of the failure.
One of the most exciting events of the day occurred during the MODIS scenario. A MODIS ancillary product was requested, but due to an unexpected operational condition, the system could not see the file. The first request was, therefore, rejected. The system is configured automatically to retry on a pre-planned delayed basis. A second retrieval request was automatically initiated and proved successful.
Jan Dreisbach (Science Data Server lead) discussed the Landsat-7 support scenario. This scenario tested the interface protocols for the Level 0 Processing System and Image Assessment System (IAS) and ingest of both types of data. B0SOT performed directory and inventory searches of Landsat-7 data, which were archived and distributed via ftp and 8 mm tape, thus demonstrating concurrent ordering from two workstations. User login and registration, operator-assisted order status tracking, data subsetting service, and browsing orders with the EOSView data visualization tool were exercised.
The B.0 development lead, Mary Armstrong, discussed work that was done prior to the August Demonstration. ECS effectively ingested and archived MODIS Level 0 data and timed the activity to ensure that the system achieved the throughput rate required at launch. The system also supported the EDOS protocol interface, ingesting and archiving ASTER L1A and L1B granules from D3 tape. Using the Data Preparation (DPREP) PGE, ECS converted EOS AM-1 ancillary data into orbit and attitude data.
Randy Miller presented a plan for achieving required ECS throughput performance followed by a presentation on the ingest thread performance by Nick Singer. Mike Daily discussed the continuing work on the JAVA Earth Science Tool (JEST), which will provide users with World Wide Web access to search for, browse, and order data from various participating archive centers.
The August Demonstration marks a significant milestone in the validation of ECS for data insertion, production, and retrieval. This effort emphasizes the ECS team's commitment to success in providing a high-performance, extensible information system infrastructure in support of NASA's Mission to Planet Earth Program.