Processing MODIS data with MODISL1DB will result in
identical products to those produced by the OBPG if the same calibration
LUTs are used, as outlined in the SeaDAS Processing Version Chart.
Notable changes in this release are:
Updated to the Collection 6 l1bgen_modisa and l1bgen_modist programs:
There is now only one Linux distribution of MODISL1DB. This release has
been tested on various versions of CentOS, Fedora, and Ubuntu using both
Intel and AMD processors, and will likely run on most other Linuxes.
If you encounter problems please let us know.
Added support for Intel Mac OS X 10.6.
Dropped support for Intel and PowerPC Mac OS X 10.4.
Dropped support for Solaris 8 (SunOS 5.8)
MODIS scripts can now be executed concurrently in the same directory
allowing for processing multiple files at once in one directory.
Added the l1asubset_modis binary that strips non-ocean bands out of
Level-1A files for oceans processing. This greatly reduces filesize and
speeds up processing.
Please be aware that if the SeaDAS software package is already installed
on your computer, you will already have full access to ALL of the
functionality of MODISL1DB. If you do decide to also install MODISL1DB,
SeaDAS and MODISL1DB should NOT both be enabled at the same time. To
disable SeaDAS, simply comment-out any lines in your shell initialization
file that set up the SeaDAS environment.
MODISL1DB can be downloaded and installed using the
online
installation page. The online installer will do everything for you,
including downloading files, installing the software, and setting up your
UNIX environment for MODISL1DB. Or, you can
manually download
MODISL1DB, and follow the installation instructions.
Using MODISL1DB
MODISL1DB is command-line driven, with processing controlled by four main
wrapper scripts: modis_L0_to_L1A_GEO.csh,
modis_L1A_extract.csh,
modis_L1A_to_GEO.csh, and modis_L1A_to_L1B.csh.
Two utility scripts (modis_update_luts.csh and
modis_update_utcpole_leapsec.csh) are also included to automate updating
of calibration LUTs and utcpole.dat/leapsec.dat. It is suggested these two
scripts be set up as cron jobs, though they can also be run manually. All
MODISL1DB scripts reside in the top-level 'scripts' directory. For usage
instructions please consult each script's associated README file:
Unlike MODISL1DB 1.5, the MODISL1DB benchmark script provided is now the same one
used for SeaDAS benchmarks. The script is now called seadas_benchmark.bash and is
available along with a MODIS Aqua Level-0 5 minute granule from our ftp site in
the seadas_benchmarks.tar.gz
file. When this script is executed, if a MODISL1DB environment is detected the script
will simply process the Level-0 file to Level-1B and report the processing times.
The SeaDAS Benchmarks Table
lists processing times from various platforms for comparison to your computer's
performance. If you perform the benchmark, please email results along with your
detailed system specs to
seadas at seadas.gsfc.nasa.gov
for inclusion in the table.
Source Code
Source code is not distributed with the MODISL1DB software package.
However, the complete source code and build environment for all MODISL1DB
binaries is available within the
SeaWiFS
Data Analysis System (SeaDAS) software package.
To build custom MODISL1DB binaries, users can install SeaDAS and then
rebuild any of the MODISL1DB binaries within their SeaDAS environment.
These custom binaries can then be substituted directly into an existing
MODISL1DB installation. Or, if the processing component of SeaDAS is
installed (distributed in the file seadas_processing.tar.gz), the
following SeaDAS script can be run to create a complete MODISL1DB package
(created in $SEADAS/modisl1db/) containing the new custom binaries:
