* Data Set Overview

Burst Ordered Data Products (BODP) are comprehensive data files that
include engineering telemetry, radar operational parameters, raw echo
data, instrument viewing geometry, and calibrated science data.  The
BODP files contain time-ordered fixed length records.  Each record
corresponds to the full set of relevant data for an individual radar
burst.  The Cassini Radar is operated in 'burst mode,' which means the
radar transmits a number of pulses in sequence then waits to receive
the return signals.  'Burst' is a descriptive term for the train of
pulses transmitted by the radar.  The term 'burst' (somewhat
unconventionally) refers to an entire measurement cycle including
transmit, receipt of echo, and radiometric (passive) measurements of
the naturally occurring radiation emitted from the surface.  In fact,
even when the transmitter is turned off and only passive measurements
are made Burst Ordered Data Products are still fixed header length,
fixed record length files.

The header is an attached PDS label.  Records are rows in a table.
Each data field is a column.  All one needs to know to read a
particular data value from a particular data field is the header
length, the record size, and the byte offset of the data field within
the record.  Since a UTC time tag is included in each record, it is a
simple matter to restrict the data one reads to a particular time
interval.

The BODP comprise four separate data sets:  the Short Burst Data Record
(SBDR), the Long Burst Data Record (LBDR), the Altimeter Burst Data
Record (ABDR), and the Altimetry Burst Data Record Summary (ABDR-SUMMARY)
products.  The only difference between the first three formats is whether
or not two data fields are included: the sampled echo data and the
altimeter profile.  The altimeter profile is an intermediate processing
result between sampled echo data and a final altitude estimate.  LBDRs
include the echo data but not the altimeter profile.  ABDRs include
the range compressed altimeter pulse profiles but not the echo data.
SBDRs include neither.  These trivial differences necessitate different
data sets because the two fields in question are much larger than all
the other data fields combined.  The majority of the bursts in a typical
Titan pass are passive measurements.  These bursts do not produce echo
data or altimeter profiles.  Of the active mode bursts most are not in
altimeter mode so no altimeter profiles are produced.  Including these
two data fields when they are invalid would ridiculously increase the
size of the archived data.  The alternative of having variable length
records was deemed to overly complicate data archiving and analysis 
procedures.  Maintaining three data sets reduces data volume while
allowing record lengths to remain fixed.The descriptions of each field in
the records can be found in the LBDR.FMT, SBDR.FMT, and ABDR.FMT files.
These files are located in the appropriate data directories (e.g.,
DATA/LBDR/LBDR.FMT).  They describe the size, type, and meaning of each
field.  The LBDR.FMT and ABDR.FMT files reference the SBDR.FMT files for
all fields held in common by the three data sets.  The ABDR-SUMMARY
product is derived from the ABDR.  It contains the estimated surface
height and related measurements derived from the average of all pulses in
a burst.

In order to further facilitate temporal segmentation of the data, the
structure and time sequence of the data is described in the sequence design
memo for the observation found in the EXTRAS directory.  Each record in a
BODP file is comprehensive and contains the acquisition time, viewing
geometry, quality flags, and radar mode of the burst, so that data
segmentation can be easily automated without resorting to the sequence
design memo.

An SBDR record is produced for every burst throughout the pass in an
observation.  An LBDR file is produced only for bursts during which the
transmitter was on.  (Sometimes it is necessary to create multiple LBDR
files in order to avoid file lengths > 2 Gbytes which are problematic for
older operating systems.)  Likewise, ABDR and ABDR-SUMMARY files are
produced only for periods in which the radar is in altimeter mode.  If
desired, bursts can be easily matched across data sets.  One data field in
each record is a burst identifier, which uniquely distinguishes a burst
from all other bursts in the mission.  Records in different data sets that
correspond to the same burst have the same burst ID.

The SBDR data record is divided into three consecutive segments from
three different levels of processing: 1) the engineering data segment,
2) the intermediate level data segment (mostly spacecraft geometry),
and 3) the science data segment (brightness temperature, backscatter,
measurement geometry, etc.).  The engineering data segment contains a
complete copy of the telemetry data downlinked from the spacecraft.
It includes temperatures, instrument instructions, operational
parameters of the radar, and raw measurements (i.e., unnormalized
radiometer counts.)

For more information about the format and content of the SBDR, LBDR, ABDR,
and ABDR-SUMMARY files, see the Cassini Radar Burst Ordered Data Product
(BODP) Software Interface Specification, JPL D-27891.  A copy of the
document is located on this volume as file BODPSIS.PDF in the DOCUMENT
directory.


* Parameters

A complete listing of the parameters can be found in the SIS.


* Processing

Cassini RADAR telemetry packets are transmitted to earth along with
other spacecraft and instrument telemetry at the conclusion of each
data take.  The radar data packets are queried from the telemetry data
system (TDS) on a computer in the radar testbed which has access to TDS.
These packets are placed sequentially into a raw data file.  The raw file
is initially processed by radar software on the testbed computer which
identifies radar science activity blocks (SAB) within the telemetry stream
and reformats the data and provides some quick look displays and limit
checking.

The reformatted data file (L0) is delivered to the radar processing group
for processing by Radar Analysis Software (RAS) and then the SAR Processor
(SP, if applicable).  Temperature telemetry files from the spacecraft
are also queried from TDS and delivered to the processing group for RAS
and SP to use.  All other ancillary data is obtained from SPICE kernel
files which are delivered by different elements of the project to an ftp
site.  These files are separately archived in the PDS system.  The RAS
pre-processor reads the radar L0 file, associated temperature telemetry
files, and the SPICE kernel files and all relevant data are placed into
the SBDR/LBDR engineering and intermediate level data segments.  The
science processors for radiometery, scatterometry, and altimetry, as well
as SAR, ingest the SBDR/LBDR files and produce mode-specific science data
products (and modify science data segment fields).

Note that measurement geometry will not be available and will be flagged
as invalid for cases in which there is no target body or the measurement
extends beyond the limb of the target body.  There is no plan to compute
the science data segment for non-Titan bodies with the exception of
radiometric observations of Saturn and its rings.  For other bodies these
fields will be flagged as invalid.  For most non-Titan icy satellite
observations, due to SNR effects, only a single brightness temperature or
backscatter measurement will be computable rather than values for each
burst.  For these observations a single backscatter value and a single
antenna temperature value will be reported in the AAREADME.TXT file in the
root directory of the volume.  Sometimes a short list of backscatter values
will be reported as a function of frequency of the returned echo.


* Data

See Parameters.


* Ancillary Data

There are no ancillary data needed to use the SBDR.


* Coordinate System

Data locations for each record are computed from Project SPICE kernels.
Measurement geometry information is available for both the active and
passive mode measurements.  Some of the active and passive mode quantities
are likely to be identical (e.g., polarization orientation angle).
However, separate data fields are reported, because the differences in the
passive and active mode measurement times can in principle cause the two
cases to differ.  Passive geometry is computed for the time corresponding
to the midpoint of the passive receiver window (summed radiometer windows).
Active mode geometry is computed for the time halfway between the midpoint
of the transmission and the midpoint of the active mode receiver window.
The full set of measurement geometry for each case includes the
polarization orientation angle, emission/incidence angle, azimuth angle,
the measurement centroid, and four points on the 3-dB gain contour of the
measurement.  The centroid and contour points are specified in latitude
and longitude, using the standard west longitude positive geodetic
(planetographic) coordinate system sanctioned by the IAU.  The geodetic
part of the definition is moot since Titan is modeled by a sphere.  The
measurement geometry will not be available and will be flagged as invalid
for cases in which there is no target body or the measurement extends
beyond the limb of the target body.


* Software

No software is provided within this volume.


* Media/Format

The data are provided on media as determined by PDS.  The main data
files are ZIPPED as described in the PDS standard.  Detached labels
are provided for the ZIPPED files.  The ZIPPED files also include
their attached labels.  Other formats are defined within the attached
labels of the files.

* Confidence Level Overview

The burst ordered data products (SBDR, LBDR, and ABDR) contain all science
data obtained by the Cassini RADAR instrument during the mission.  A data
record is obtained for each measurement cycle.  This data is calibrated
whenever a calibrated value obtained from a single measurement cycle is
physically justified.  This is always the case for close Titan flybys and
seldom the case for other observations.

* Review

The data will be validated internally by the Cassini Radar Team prior to
each release of data to the PDS.  The overall data set organization will
also be peer reviewed once by the PDS prior to the release of the first
volume.

* Data Coverage and Quality

The following observations will be included in the SBDR and LBDR archives:

1) Close (1000-200000 km) Titan flyby observations
2) Distant Titan observations
3) Observations of Saturn and its rings
4) Observations of icy satellites
5) Radiometer observation of Jupiter
6) Sun scan observations used to calibrate antenna patterns
7) Earth swingby observation

The SBDR contains one record for each active or passive measurement cycle
but does not contain the sampled radar echo data or altimeter profiles.

The LBDR includes only active mode measurement cycles.  It contains sampled
radar echo data but no altimeter profiles.

The ABDR includes only altimeter measurement cycles.  It contains altimeter
profiles but no radar echo data.

The ABDR-SUMMARY file is derived from the ABDR.  It contains the estimated
surface height and related measurements derived from the average of all
pulses in a burst when the radar is in altimeter mode.

SBDRs will be produced for all the observations mentioned above.

LBDRs will be produced for close Titan flybys, the Earth swingby, and also
some distant Titan and icy satellite observations whenever active mode
measurements are obtained.

ABDRs and ABDR-SUMMARY files will only be produced for close Titan flybys
because this is the only time altimeter mode is employed.

Three types of calibrated measurements are obtained: backscatter
(scatterometer), antenna temperature (radiometer), and surface height
(altimeter).

The calibrated scatterometer values are normalized radar cross-section, a
unitless quantity related to the slope, roughness, and composition of the
observed surface.  We currently estimate the accuracy of this quantity to
be + or - 3 dB absolute and + or - 2 dB relative.  We estimate the accuracy
of the antenna temperature to be + or - 5 percent absolute and + or - 0.2 K
relative.

We estimate the accuracy of the surface height to be approximately + or -
100-200 m.

Location error of the measurements is dominated by ephemeris and pointing
knowledge errors and is expected to be less than 2 km throughout the
mission.

* Limitations

Ephemeris error is expected to improve throughout the mission.  Since there
is no plan to recompute the ephemeris of previous observations as new
measurements are obtained, however, earlier observations may have poorer
location accuracy.

The dominant backscatter calibration error term is error in our knowledge
of the gain of the attenuators in the receiver.  Engineering tests are
currently planned to improve our knowledge of the attenuator gains.

Besides ephemeris error, the main errors in the altimeter surface height
determination come from off-nadir pointing and actual surface height
variation with the altimeter footprint of diameter typically 25-50 km.