DATA_SET_DESCRIPTION |
Data Set Overview
=================
Data Set MARS RECONNAISSANCE ORBITER MARS SHARAD REDUCED DATA
RECORD V1.0 (Level 1B data) consists of received echoes that have
been Doppler filtered, range compressed, and converted to complex
voltages, correlated with the auxiliary information needed to
locate observations in space and time and to process data further.
Level 1B data users are expected to be mainly geologists
interested in determining the structure of the shallow Martian
subsurface. Data users must be aware that processed echoes may
contain artifacts due to off-nadir surface reflections, the so-
called clutter, reaching the radar after nadir surface echoes, and
thus appearing as subsurface reflections.
Processing
==========
SHARAD RDR Data Products are generated at the SHARAD operation
center in Rome, Italy, under the responsibility of the Team Leader
Institution (INFOCOM Department, University of Rome 'La
Sapienza').
In the processing producing RDRs, data are range- and Doppler-
processed. The software used to produce RDR data products requires
a configuration file containing values for parameters used by the
processing algorithms. Once the parameters are set, data
processing proceeds in a fully automated way. Level 1B processing
consists of the following functions:
1) Auxiliary data loading
2) Data selection
3) Data healing
4) Phase distortion compensation
5) Doppler parameters estimation
6) Range-Doppler processing
7) Multi-look processing
8) Output generation
Auxiliary data loading consists in the loading of files needed in
the processing. Such files are part of the SHARAD Level 1A (EDR)
archive, which thus constitutes the sole input for the production
of RDR data products. They are the antenna pattern file,
describing the SHARAD antenna relative gain as a function of the
spacecraft high-gain antenna and solar panels position; the
reference function file, that is the complex conjugate of the FFT
of the discretely-sampled transmitted signal; and the spurious
frequencies file, reporting the spectrum of coherent noise
produced by the spacecraft electronics within the SHARAD operation
bandwidth.
Data selection consists in the loading of the Level 1A data
products to be processed and of the configuration file setting
parameters required by processing algorithms. During this phase,
each sample of each raw echo, compressed on-board into a 4-, 6- or
8- bit integer, is decompressed according to two different
algorithms, depending on the selection of either static
compression or dynamic compression for on-board processing. If a
static compression has been adopted, then the uncompressed value
is given by
S32 = S4,6,8 / 2N
where S32 is the sample in 32-bit floating point notation, S4,6,8
is the compressed value of the sample, and N is the number of pre-
summed echoes. If dynamic compression has been adopted, then
S32 = S4,6,8 / 2p
where p is the SDI Bit-Field parameter reported into the Science
Ancillary Data.
Data healing consists of the compensation of antenna pattern
variations caused by the movement of the spacecraft high-gain
antenna and solar panels during the data take, and of the removal
of spurious signals introduced by the spacecraft electronics.
Phase distortions may occur due to interaction between the radar
signal and the Martian ionosphere plasma, whose refraction index
is a function of frequency. Phase distortion compensation is a
fully automated function implementing the Phase Gradient Autofocus
(PGA) algorithm.
Doppler parameters estimation produces estimates of the Doppler
centroid frequency and of the Doppler bandwidth for the raw data.
Doppler centroid estimation determines the average Doppler
frequency of received echoes in order to perform an optimum
azimuth filtering. The centroid frequency is the frequency
dividing the energy spectra of averaged signals in two equal
parts. The Doppler bandwidth determines the maximum possible
azimuth resolution of the output echoes, and the amount of data
overlapping from one synthetic aperture to the next.
Range-Doppler processing consists of range compression followed by
Doppler processing. Range compression consists in the matched or
inverse filtering of each raw echo. The filtering is performed by
multiplying the raw signal FFT spectrum by the reference function
(matched filtering) or by its inverse (inverse filtering). Doppler
processing is performed through the Chirp Scaling Algorithm (CSA)
for SAR processing, which allows also for Range Cell Migration
Compensation (RCMC). CSA guaranties good performances in terms of
computational speed and achievable azimuth resolution.
Multi-look processing aims at reducing speckle: the random
fluctuations of received power due to coherent reflections, by
averaging the values of neighboring range-Doppler pixels within a
given window around each pixel. Such filtering will increase the
radiometric resolution, but the spatial resolution will decrease
as a result. The window size is fixed, while the window weight is
selectable by means of the configuration file.
In the generation of the output files, the processed echoes, now a
complex quantity because of the processing taking place in the
Fourier spectral domain, are complemented with the original
auxiliary data contained in the scientific telemetry of the
instrument, with parameters characterizing the ground processing
of the echoes, with geometric quantities generated on-ground from
spacecraft navigation data, and with parameters extracted from
instrument and spacecraft housekeeping telemetry. All this
auxiliary information, with the exception of processing
parameters, is copied from the input Level 1A files used to
produce the RDR data product.
Data
====
Each SHARAD RDR data product is the result of the processing all
echoes acquired continuously in time using the same operation
mode, instrument status and on-board processing scheme. There is
one RDR data product for every SHARAD Experiment Data Record data
product acquired in subsurface sounding mode, which in fact
constitutes the input for the RDR product generation.
The content of each SHARAD RDR data product is highly variable in
terms of number of processed echoes, and depends on how operations
for the instrument were planned during a given data collection
period.
Each processed radar observation in an RDR data product is the
result of range and azimuth processing of a variable number of raw
echoes. Thus, although there is a one-to-one correspondence
between EDRs and RDRs, such correspondence does not hold between
individual raw and processed echoes: in general, many raw echoes
(of the order of several tens or a few hundred) result in the
production of a single processed echo.
Because each processed echo is a sequence of time samples of a
received signal, complemented by ancillary information, the
natural organization for processed echoes within a Data Product is
a table, in which each line contains data from a single processed
echo, and each column contains the value of a single parameter or
time sample across different processed echoes.
Each Data Product consists of two files:
1) A binary file containing the scientific data of the instrument:
a sequence of processed echoes, each of which is preceded by a
header containing information on the instrument setting and on-
board processing of the data, and followed by parameters
characterizing the ground processing of the echoes, by geometric
quantities generated on-ground from spacecraft navigation data,
and by parameters extracted from instrument housekeeping
telemetry.
2) A detached ASCII PDS label file describing the content of the
data product.
Ancillary Data
==============
Ancillary data describe the instrument settings during data
acquisition and report parameters used in on-board processing.
Scientific data consist of the processed complex echo, expressed
as a vector for the real part of the complex echo, followed by a
second vector of the same length containing the imaginary part of
the echo. Processing parameters have been computed by the ground
processing software, and contain information such as the Doppler
centroid and Doppler bandwidth of original raw echoes. Geometric
parameters have been computed on ground from spacecraft trajectory
and attitude data, and allow the location of each processed echo
in space and time. Instrument engineering parameters are extracted
from SHARAD housekeeping telemetry and report quantities such as
currents and temperatures within the instrument.
Coordinate System
=================
SHARAD RDR data products conform to a Project-determined set of
cartographic standards. All map-projected data use planetocentric
coordinates and east-positive longitudes in the range 0 to 360
degrees, computed w.r.t. the IAU 2000 reference ellipsoid. Vector
quantities such as spacecraft positions are expressed in a
Cartesian planetocentric reference frame.
Media/Format
============
Each SHARAD RDR data product consists of a binary file in fixed
record-length format, and a detached PDS label containing
information on source data, production process, relation between
stored bytes and physical quantities, product identification,
storing and organizing of ancillary data and descriptive
information needed to interpret and process the data. The
structure of data contained in the binary file is that of a PDS
Table object. The PDS label contains pointers to a file containing
definitions of the columns of such Table object.
Each record in a binary file is a processed echo: the result of
the Doppler filtering and range compressing of a variable number
of raw echoes, and is expressed as a time series of complex
voltages. Each record also contains a number of parameters
describing the operation of the instrument during data
acquisition, together with engineering and spacecraft information.
Specifically, each record is subdivided into five parts, each of
which contains a different type of information: ancillary data,
scientific data, processing parameters, geometric parameters and
instrument engineering parameters.
PDS labels are written in Object Description Language (ODL). PDS
label statements have the form of 'keyword = value'. Each label
statement is terminated with a carriage return character (ASCII
13) and a line feed character (ASCII 10) sequence to allow the
label to be read by many operating systems. The labels contained
in SHARAD EDR files conform to the general structure used for all
PDS detached labels.
|