DATA_SET_DESCRIPTION |
Data Set Overview
=================
This data set contains Calibrated data taken by New Horizons
Long Range Reconnaissance Imager
instrument during the LAUNCH mission phase.
LORRI is a narrow angle (Field Of View, FOV = 0.29 degree square), high
resolution (5 microradian/pixel), telescope. A two-dimensional (2-D) CCD
detector, with 1024x1024 pixels (optically active region) operates in
standard frame-transfer mode. LORRI can also perform on-chip 4x4 binning
to produce images of 256x256 pixels. LORRI has no color filters and so
provides panchromatic imaging over a wide bandpass extending
approximately from 350 nm to 850 nm. The common data product is a 2-D
image of brightnesses that can be calibrated to physical units once
color spectrum information is known. Refer to DOCUMENT/SOC_INST_ICD.*
for more detail.
During the Launch Phase, LORRI conducted a variety of instrument
commissioning activities in preparation for science operations. An
extensive series of bias images was collected during the spring
and summer of 2006, before the LORRI aperture door was opened.
Images were also collected of LORRI's internal lamps. The LORRI
aperture door was opened on 2006 August 29, and an extensive
series of calibration observations of stars in the open cluster M7
was performed during the next several days. Commissioning test
observations were also performed on the planetary targets Jupiter,
Uranus, Neptune, and Pluto. The Jovian observations were conducted
specifically to test LORRI's ability to perform imaging using
exposure times that were short compared to the frame transfer time
(i.e., shorter than 13 ms). Some histogram-only solar scattered
light observations were performed during a Ralph commissioning
test.
Every observation provided in this data set was taken as a part of a
particular sequence. A list of these sequences has been provided in
file DOCUMENT/SEQ_LORRI_LAUNCH.TAB.
N.B. Some sequences provided may have no corresponding observations.
For a list of observations, refer to the data set index table. This
is typically INDEX.TAB initially in the INDEX/ area of the data set.
There is also a file SLIMINDX.TAB in INDEX/ that summarizes key
information relevant to each observation, including which sequence
was in effect and what target was likely intended for the
observation.
Version
=======
This is VERSION 3.0 of this data set.
The pipeline (see Processing below) was re-run on these data for each
version since the first (V1.0). As a result, ancillary information,
such as observational geometry and time (SPICE), may be updated.
This will affect, for example, the calibration of the data if parameters
such as the velocity or orientation of the target relative to the
instrument, or the recorded target itself, have changed.
See the following sections for details of what has changed over each
version since the first (V1.0). Note that even if this is not a
calibrated data set, the calibration changes are listed as the data
will have been re-run and there will be updates to the calibration
files, to the documentation (Science Operations Center - Instrument
Interface Control Document: SOC_INST_ICD) and to the steps required
to calibrate the data.
LORRI updates for Data Sets V3.0
================================
LORRI V3.0 provides a few updates from LORRI V2.0. The lossy
images were recalibrated, including expanding the 'bad' pixel
designation of 8x8 boxes affected by the first 34 pixels of header
information in the calibrated quality map. Also, updates were
made to the documentation and catalog files, primarily to resolve
liens from the V2.0 peer review. No new observations were added
with Version 3.0.
LORRI updates for Data Sets V2.0
================================
Improvements to geometry & timing in data header sections
---------------------------------------------------------
The data pipeline corrected some FITS header keywords
(SPCEMEN, SPCPAXIN, SPCPAYIN, PA_XINST, PA_YINST) that were
in error by 90 degrees and that express the orientation of the
LORRI FOV around the boresight; the correction did not affect
the quaternion, which was and is correct.
The data pipeline re-ran the data with updated SPICE files;
SPICE files contain information relating to geometry and
timing of observations. This may have changed values in
the data PDS labels and the data files' FITS header section
(referred to as OBJECT *_HEADER in the PDS labels), but not
the image data in those files.
World Coordinate System (WCS) keywords have been added to the
FITS headers.
Approximate target and target location in line and sample have
been added in PDS mission-specific keywords
NEWHORIZONS:APPROX_TARGET_NAME
NEWHORIZONS:APPROX_TARGET_LINE
NEWHORIZONS:APPROX_TARGET_SAMPLE
APPROX_TARGET_LINE and APPROX_TARGET_SAMPLE quantities are
one-based and consistent with line and sample as defined by PDS
keywords LINES, LINE_SAMPLES, LINE_DISPLAY_DIRECTION, and
SAMPLE_DISPLAY_DIRECTION.
CAVEATS:
These were derived from other information in the labels.
These should not be used whenever better information, e.g.
SPICE, is available.
The project provides these values but makes no claim as to
their accuracy beyond that of keywords from which they were
derived.
These values should only be used as a rough guide to the
location of the selected target; if no target is seen near the
specified location, and/or the calculated position is out of the
field of view, the user should consider other possibilites (the
target specified is too faint, or it is the wrong target) and
methods (e.g. SPICE).
Improvements and corrections to ancillary files
-----------------------------------------------
Some of the supporting files, mainly in the /DOCUMENT/
subdirectory hierarchy, have been updated for content
and for readability. The primary file for which this
was done was the SOC Instrument Interface Control
Document (often referred to as the SOC Instrument ICD).
The PDS labels of the dead- and hot-pixel calibration files
were corrected to state that the data are in integer format
instead of real.
Processing
==========
The data in this data set were created by a software data
processing pipeline on the Science Operations Center (SOC) at
the Southwest Research Institute (SwRI), Department of Space Operations.
This SOC pipeline assembled data as FITS files from raw telemetry
packets sent down by the spacecraft and populated the data labels
with housekeeping and engineering values, and computed geometry
parameters using SPICE kernels. The pipeline did not resample
the data.
Calibration
===========
Detailed information about calibration of LORRI images is available
in the SOC Instrument Interface Control Document (ICD) in the
DOCUMENT section of this data set. The LORRI calibration will only
be briefly summarized here; refer to the ICD for details about
what is summarized here.
N.B. The units of the RDR image data are calibrated Data Number (DN);
responsivity factors are provided in the PDS label and FITS
headers to convert the calibrated DNs to physical units; the
factor to use is dependent on the target scene spectrum. Refer
to the ICD and other LORRI documentation [CHENGETAL2008],
[MORGANETAL2005] for more detail. Note also that some versions
of [CHENGETAL2008], including the published version, have an
error in the units of its Figure 9 ordinate.
The PDS keyword PROCESSING_HISTORY_TEXT has been provided in each
PDS label for calibrated data file with details of the parameters
used or calculated and of the calibration files used in the
calibration process. The responsivity factors mentioned in the
previous paragraph are included there.
The calibration of LORRI images involves all of the following steps
in order:
1) Bias subtraction
2) Signal linearization
3) Charge transfer inefficiency (CTI) correction
4) Dark subtraction
5) Smear removal
6) Flat-fielding
7) Absolute calibration (DN with scene-dependent radiance divisors)
Ground testing has demonstrated that the linearization, CTI and
dark subtraction steps are not necessary i.e. the output from the
Bias subtraction step may be passed directly to Smear removal step.
In addition, the calibration procedure calculates the error and
a data quality flag for each pixel and includes those results
in the calibrated data product as additional PDS OBJECTs (FITS
extensions) appended to the main OBJECT with the data image.
The quality flag PDS OBJECT is an image of values of the same
size as the main IMAGE product, with each quality flag pixel
mapped to the corresponding pixel in the main product. A quality
flag value of zero indicates a valid pixel; a non-zero value
indicates an invalid pixel. Each quality extension pixel value is
an accumulated sum of individual quality flag values. The list below
contains the quality flag value associated with each condition:
Quality Flag Value Quality Flag Description
0 Good pixel
1 Defect in reference deltabias image (0 or NaN)
2 Defect in reference flatfield image (0 or NaN)
4 Permanent CCD defect (e.g. dead pixel) *
8 Hot Pixel identified in hotpixel map *
16 Saturated pixel in level1 data (A/D value of 4095)
32 Missing level1 data ( assume fill value of 0 )
64 unused at present
Note that for windowed products, all pixels in an image are not
returned in the downlink telemetry. In the raw data, the pipeline
sets such pixels to zero DN (Data Number); the calibration processes
those zero-DN pixels as if they were real raw values, but also flags
them as missing data in the quality flag PDS OBJECT (FITS extension).
Displaying such images using an automatic stretch (contrast
enhancement) may result in a confusing result with the majority of
the displayed image appearing as an inverse of the calibration
(calibration of zero values); therefore the quality flag PDS OBJECT
should always be checked when looking at these data.
Ongoing in-flight calibration observations will be analyzed to
assess the long term stability of the calibration, including whether
the currently unused steps may need to be implemented in the future.
* As of late 2016, there are no known dead or hot pixels on the
LORRI detector, so all hot and dead pixel map calibration files
contain all zeroes. From the current flat-field calibration file
it can be seen that there are many pixels with relative
sensitivities up to six times the mean (unity), those called warm
pixels. Those pixels are calibrated in the flat-field step.
The first 34 pixels (51 bytes) of each LORRI image contain housekeeping
information. These bytes are corrupted by LOSSY compression. Please
refer to the DOCUMENT/SOC_INST_ICD.* file for details and implications
in raw and calibrated datasets.
Data
====
The observations in this data set are stored in data files using
standard Flexible Image Transport System (FITS) format. Each FITS
file has a corresponding detached PDS label file, named according
to a common convention. The FITS files may have image and/or table
extensions. See the PDS label plus the DOCUMENT files for a
description of these extensions and their contents.
This Data section comprises the following sub-topics:
- Filename/Product IDs
- Instrument description
- Other sources of information useful in interpreting these Data
- Visit Description, Visit Number, and Target in the Data Labels
Filename/Product IDs
--------------------
The filenames and product IDs of observations adhere to a
common convention e.g.
LOR_0123456789_0X630_ENG.FIT
^^^ ^^^^^^^^^^ ^^^^^ ^^^\__/
| | | | ^^
| | | | |
| | | | +--File type (includes dot)
| | | | - .FIT for FITS file
| | | | - .LBL for PDS label
| | | | - not part of product ID
| | | |
| | | +--ENG for CODMAC Level 2 data
| | | SCI for CODMAC Level 3 data
| | |
| | +--Application ID (ApID) of the telemetry data
| | packet from which the data come
| | N.B. ApIDs are case-insensitive
| |
| +--MET (Mission Event Time) i.e. Spacecraft Clock
|
+--Instrument designator
Note that, depending on the observation, the MET in the data filename
and in the Product ID may be similar to the Mission Event Time (MET)
of the actual observation acquisition, but should not be used as an
analog for the acquisition time. The MET is the time that the data are
transferred from the instrument to spacecraft memory and is therefore
not a reliable indicator of the actual observation time. The PDS label
and the index tables are better sources to use for the actual timing of
any observation. The specific keywords and index table column names for
which to look are
* START_TIME
* STOP_TIME
* SPACECRAFT_CLOCK_START_COUNT
* SPACECRAFT_CLOCK_STOP_COUNT
Instrument Instrument designators ApIDs **
=========== ================================== =============
LORRI LOR 0X630 - 0X63B *
* Not all values in this range are in this data set
** ApIDs are case insensitive
There are other ApIDs that contain housekeeping values and
other values. See SOC Instrument ICD (/DOCUMENT/SOC_INST_ICD.*)
for more details.
Here is a summary of the types of files generated by each ApID
(N.B. ApIDs are case-insensitive) along with the instrument
designator that go with each ApID:
ApIDs Data product description/Prefix(es)
===== ===================================
0x630 - LORRI High-res Lossless (CDH 1)/LOR
0x636 - LORRI High-res Lossless (CDH 2)/LOR
0x632 - LORRI High-res Lossy (CDH 1)/LOR
0x638 - LORRI High-res Lossy (CDH 2)/LOR
0x631 - LORRI High-res Packetized (CDH 1)/LOR
0x637 - LORRI High-res Packetized (CDH 2)/LOR
0x633 - LORRI 4x4 Binned Lossless (CDH 1)/LOR
0x639 - LORRI 4x4 Binned Lossless (CDH 2)/LOR
0x635 - LORRI 4x4 Binned Lossy (CDH 1)/LOR
0x63B - LORRI 4x4 Binned Lossy (CDH 2)/LOR
0x634 - LORRI 4x4 Binned Packetized (CDH 1)/LOR
0x63A - LORRI 4x4 Binned Packetized (CDH 2)/LOR
Instrument description
----------------------
Refer to the following files for a description of this instrument.
CATALOG
LORRI.CAT
DOCUMENTS
LORRI_SSR.*
SOC_INST_ICD.*
NH_LORRI_V###_TI.TXT (### is a version number)
Other sources of information useful in interpreting these Data
--------------------------------------------------------------
Refer to the following files for more information about these data
NH Trajectory tables:
/DOCUMENT/NH_MISSION_TRAJECTORY.* - Heliocentric
LORRI Field Of View definitions:
/DOCUMENT/NH_FOV.*
/DOCUMENT/NH_LORRI_V###_TI.TXT
Visit Description, Visit Number, and Target in the Data Labels
---------------------------------------------------------------
The observation sequences were defined in Science Activity Planning
(SAP) documents, and grouped by Visit Description and Visit Number.
The SAPs are spreadsheets with one Visit Description & Number per row.
A nominal target is also included on each row and included in the data
labels, but does not always match with the TARGET_NAME field's value in
the data labels. In some cases, the target was designated as RA,DEC
pointing values in the form ``RADEC=123.45,-12.34'' indicating Right
Ascension and Declination, in degrees, of the target from the
spacecraft in the Earth Equatorial J2000 inertial reference frame.
This indicates either that the target was either a star, or that the
target's ephemeris was not loaded into the spacecraft's attitude and
control system which in turn meant the spacecraft could not be pointed
at the target by a body identifier and an inertial pointing value had
to be specified as Right Ascension and Declination values. PDS-SBN
practices do not allow putting a value like RADEC=... in the PDS
TARGET_NAME keyword's value. In those cases the PDS TARGET_NAME value
is set to CALIBRATION. TARGET_NAME may be N/A (Not Available or Not
Applicable) for a few observations in this data set; typically that
means the observation is a functional test so N/A is an appropriate
entry for those targets, but the PDS user should also check the
NEWHORIZONS:OBSERVATION_DESC and NEWHORIZONS:SEQUENCE_ID keywords in
the PDS label, plus the provided sequence list (see Ancillary Data
below) to assess the possibility that there was an intended target.
Ancillary Data
==============
The geometry items included in the data labels were computed
using the SPICE kernels archived in the New Horizons SPICE
data set, NH-X-SPICE-6-LAUNCH-V1.0.
Every observation provided in this data set was taken as a part of a
particular sequence. A list of these sequences has been provided in
file DOCUMENT/SEQ_LORRI_LAUNCH.TAB. In addition, the
sequence identifier (ID) and description are included in the PDS label
for every observation. N.B. While every observation has an associated
sequence, every sequence may not have associated observations. Some
sequences may have failed to execute due to spacecraft events (e.g.
safing). No attempt has been made during the preparation of this data
set to identify such empty sequences, so it is up to the user to
compare the times of the sequences to the times of the available
observations from INDEX/INDEX.TAB to identify such sequences.
Time
====
There are several time systems, or units, in use in this dataset:
New Horizons spacecraft MET (Mission Event Time or Mission Elapsed
Time), UTC (Coordinated Universal Time), and TDB Barycentric
Dynamical Time.
This section will give a summary description of the relationship
between these time systems. For a complete explanation of these
time systems the reader is referred to the documentation
distributed with the Navigation and Ancillary Information
Facility (NAIF) SPICE toolkit from the PDS NAIF node, (see
http://naif.jpl.nasa.gov/).
The most common time unit associated with the data is the spacecraft
MET. MET is a 32-bit counter on the New Horizons spacecraft that
runs at a rate of about one increment per second starting from a
value of zero at
19.January, 2006 18:08:02 UTC
or
JD2453755.256337 TDB.
The leapsecond adjustment (DELTA_ET = ET - UTC) was 65.184s at
NH launch, and the first three additional leapseconds occured
in at the ends of December, 2009, June, 2012 and June, 2015.
Refer to the NH SPICE data set, NH-J/P/SS-SPICE-6-V1.0, and the
SPICE toolkit docmentation, for more details about leapseconds.
The data labels for any given product in this dataset usually
contain at least one pair of common UTC and MET representations
of the time at the middle of the observation. Other portions
of the products, for example tables of data taken over periods
of up to a day or more, will only have the MET time associated
with a given row of the table.
For the data user's use in interpreting these times, a reasonable
approximation (+/- 1s) of the conversion between Julian Day (TDB)
and MET is as follows:
JD TDB = 2453755.256337 + ( MET / 86399.9998693 )
For more accurate calculations the reader is referred to the
NAIF/SPICE documentation as mentioned above.
Reference Frame
===============
Geometric Parameter Reference Frame
-----------------------------------
Earth Mean Equator and Vernal Equinox of J2000 (EMEJ2000) is the
inertial reference frame used to specify observational geometry items
provided in the data labels. Geometric parameters are based on best
available SPICE data at time of data creation.
Epoch of Geometric Parameters
-----------------------------
All geometric parameters provided in the data labels were computed at
the epoch midway between the START_TIME and STOP_TIME label fields.
Software
========
The observations in this data set are in standard FITS format
with PDS labels, and can be viewed by a number of PDS-provided
and commercial programs. For this reason no special software is
provided with this data set.
Contact Information
===================
For any questions regarding the data format of the archive,
contact
New Horizons LORRI Principal Investigator:
Andrew Cheng, Johns Hopkins Univ., Applied Physics Lab
Andrew Cheng
Johns Hopkins University
Applied Physics Laboratory
Space Department
11100 Johns Hopkins Road
Laurel, MD 20723
USA
|
CONFIDENCE_LEVEL_NOTE |
Confidence Level Overview
=========================
During the processing of the data in preparation for
delivery with this volume, the packet data associated with each
observation were used only if they passed a rigorous verification
process including standard checksums.
In addition, raw (Level 2) observation data for which adequate
contemporary housekeeping and other ancillary data are not available
may not be reduced to calibrated (Level 3) data. This issue is raised
here to explain why some data products in the raw data set,
NH-X-LORRI-2-LAUNCH-V3.0,
may not have corresponding data products in the calibrated data set,
NH-X-LORRI-3-LAUNCH-V3.0.
Data coverage and quality
=========================
Every observation provided in this data set was taken as a part of a
particular sequence. A list of these sequences has been provided in
file DOCUMENT/SEQ_LORRI_LAUNCH.TAB. N.B. Some sequences
provided may have zero corresponding observations.
Refer to the Confidence Level Overview section above for a summary
of steps taken to assure data quality.
The LORRI instrument replaces the first 34 12-bit pixels of each
LORRI image (408 bits; 51 bytes) with encoded binary header
information, so those first 34 pixel values in the first row are not
representative of the brightness of the imaged scene at those
locations; these pixels are in the bottom-left corner of images
displayed left-to-right and bottom-to-top. Furthermore, if the image
was LOSSY-compressed before downlink (ApIDs 0x632, 0x635, 0x638,
0x63B *), the header information corrupts the first 40 pixels of the
first 8 rows of the image because of the Discrete Cosine Transform
compression algorithm. The SOC pipeline extracts these data into the
FIRST34 extension of LORRI FITS files, which is also corrupt in
LOSSY-compressed files. The SOC calibration pipeline also flags these
pixels as bad in the QUALITY_MAP extension of calibrated FITS files;
no such flags are available in the raw FITS files; the SOC pipeline
did not flag the additional corrupt pixels beyond the first 34 in
LOSSY-compressed data until the Pluto P2 delivery late in 2016.
* See ApID definitions under the Data section of the Data Set
Description, DATA_SET_DESC, in this data set catalog.
Observation descriptions in this data set catalog
=================================================
Some users will expect to find descriptions of the observations
in this data set here, in this Confidence Level Note. This data
set follows the more common convention of placing those
descriptions under the Data Set Description (above, if the user is
reading this in the DATASET.CAT file) of this data set catalog.
Caveat about TARGET_NAME in PDS labels and observational intent
===============================================================
A fundamental truth of managing data from some spacecraft missions
is that the intent of any observation is not suitable for insertion
into the command stream sent to the spacecraft to execute that
observation. As a result, re-attaching that intent to the data
that are later downlinked is problematic at best. For New Horizons
that task is made even more difficult as the only meta-data that
come down with the observation is the unpredictable time of the
observation. The task is made yet even more difficult because
uplink personnel, who generate the command sequences and initially
know the intent of each observation, are perpetually under
deadlines imposed by orbital mechanics and can rarely be spared for
the time-intensive task of resolving this issue.
To make a long story short, the downlink team on New Horizons has
created an automated system to take various uplink products, decode
things like Chebyshev polynomials in command sequences representing
celestial body ephemerides for use on the spacecraft to control
pointing, and infer from those data what the most likely intended
target was at any time during the mission. This works well during
flyby encounters and less so during cruise phases and hibernation.
The point to be made is that the user of these PDS data needs to
be cautious when using the TARGET_NAME and other target-related
parameters stored in this data set. This is less an issue for the
plasma and particle instruments, more so for pointing instruments.
To this end, the heliocentric ephemeris of the spacecraft, the
spacecraft-relative ephemeris of the inferred target, and the
inertial attitude of the instrument reference frame are provided
with all data, in the J2000 inertial reference frame, so the user
can check where that target is in the Field Of View (FOV) of the
instrument. Furthermore, for pointing instruments with one or more
spatial components to their detectors, a table has been provided
in the DOCUMENT/ area with XY (two-dimensional) positions of each
inferred target in the primary data products. If those values are
several thousand pixels off of a detector array, it is a strong
indication that the actual target of that observation is something
other than the inferred target, or no target at all e.g. dark sky.
Review
======
This dataset was peer reviewed and certified for scientific use on
12-5-2016.
|