Contact Stansberry - ";
echo " there is a problem w/ your write-up.\n";
return ;
}
$row = mysql_fetch_array($result);
$title = $row["title"];
$princ = $row["principal"];
$deputy= $row["deputy"];
$campn0 = $row["campn0"];
$aorkeys = $row["aorkeys"];
// get real name of principal, deputies
$princ = ioc_get_person($princ);
$princ = $princ[0];
$deps = explode(",",$deputy);
foreach ($deps as $depty) {
$depty = trim($depty);
$depty = ioc_get_person($depty);
$depty = $depty[0];
$depty = explode(",",$depty);
$depty = $depty[0]; // last names only
$deplist[] = $depty;
}
$deplist = implode(", ", $deplist);
$caid = sprintf("%03d",$caid);
$file = "mips-".$caid.$campn.".analysis.php";
// END PHP.
?>
<? echo "MIPS-$caid, Campaign $campn IOC/SV Analysis"; ?>
Principal:
Deputy:
Analyst:
AORKEYS:
Last Updated:
Task Outcome Summary
- DATA STATUS: Incomplete
- TASK OUTCOME: Incomplete
Abstract
The D2 execution of this task used an outdated IER which
resulted in an incomplete data set. Because of this, the comparison
between RAW and SUR exposures was done for only 4 second exposure
times and a non-optimal stim setting. Despite these problems,
the data set shows that the SUR data compression algorithm works
as determined by ground testing. The first DCE commanded in an
exposure sequence (DCE0) continues to produce a lower on-board
slope value than fitting of a comparable RAW ramp.
Analysis
Twenty DCEs were obtained in both the RAW and SUR modes with the 24 micron
stimulator set to "24FLD" "LOW". The CSMM was positioned at the 24 micron
dark position, which eliminated all but a few DN/sample-time from scattered
light at the 31 K temperatures of D2. The RAW data were processed using the
SUR processing emulation software "sursimslope" written by Frank Masci.
This software was compared against the DAT slope fitting algorithm
written by James Muzerolle and was found to produce virtually the same
result. These algorithms emulate the on-board SUR processor by making
a linear, least-squares fit to the samples in the RAW ramp after discarding
the specified number of "ignore frames" at the beginning of the exposure.
The "ignore frames" parameter is a patchable constant. There is a different
"ignore frames" parameter for the first DCE in a commanded exposure
sequence (DCE0) and all subsequent DCEs produced by a single CEMIPSUR
command. The "sursimslope" program was run using ignore frame = 3
for DCE0 and ignore frames = 1 for all subsequent DCEs. The simulated SUR
images were compared with actual SUR images of the same exposure time
and commanded order. The comparison utilized the IDL "dostats" program
written by Susan Stolovy.
Results
A table of the comparison between slopes derived from 4 second RAW and SUR images
of the 24FLD_LOW stim is below:
For DCE0, there is a large (9%) disparity between the predicted median slope in
DN/sample-time derived from the RAW DCE and the actual SUR slope. Subsequent
DCEs rapidly converge on extremely similar values for the RAW and SUR slopes.
The DCE1 SUR slope is low by 2%, DCE2 is low by 0.7%, and by DCE5, the values
differ by less than 0.3%.
The results found in campaign D2 are very similar to the results obtained by
an identical analysis of the LF1 ground test (9/14/02). The table of results
for that test (which used the identical stimulator setting) is below:
MODE
|
AOR/EXP/DCENUM
|
MEAN
|
MEDIAN
|
EXPTIME
|
SUR
|
12032.0.0
|
1085.15
|
868.00
|
2.62
|
RAW
|
12032.1.0
|
1236.73
|
998.50
|
2.62
|
SUR
|
12032.0.1
|
1194.96
|
963.00
|
3.67
|
RAW
|
12032.1.1
|
1218.40
|
983.29
|
3.67
|
SUR
|
12032.0.2
|
1211.57
|
976.00
|
3.67
|
RAW
|
12032.1.2
|
1221.35
|
985.68
|
3.67
|
SUR
|
12032.0.3
|
1215.64
|
980.00
|
3.67
|
RAW
|
12032.1.3
|
1222.62
|
986.32
|
3.67
|
SUR
|
12032.0.4
|
1217.95
|
982.00
|
3.67
|
RAW
|
12032.1.4
|
1223.22
|
987.39
|
3.67
|
SUR
|
12032.0.5
|
1219.39
|
983.00
|
3.67
|
RAW
|
12032.1.5
|
1223.50
|
987.96
|
3.67
|
SUR
|
12032.0.6
|
1220.47
|
983.00
|
3.67
|
RAW
|
12032.1.6
|
1223.97
|
987.68
|
3.67
|
SUR
|
12032.0.7
|
1221.21
|
984.00
|
3.67
|
RAW
|
12032.1.7 |
1224.10
|
988.21
|
3.67
|
SUR
|
12032.0.8
|
1221.94
|
985.00
|
3.67
|
RAW
|
12032.1.8
|
1224.15
|
988.25
|
3.67
|
SUR
|
12032.0.9
|
1222.41
|
985.00
|
3.67
|
RAW
|
12032.1.9
|
1224.40
|
987.86
|
3.67
|
SUR
|
12032.0.10
|
1222.99
|
986.00
|
3.67
|
RAW
|
12032.1.10
|
1224.39
|
988.61
|
3.67
|
SUR
|
12032.0.11
|
1223.39
|
986.00
|
3.67
|
RAW
|
12032.1.11
|
1224.58
|
988.82
|
3.67
|
SUR
|
12032.0.12
|
1223.72
|
987.00
|
3.67
|
RAW
|
12032.1.12
|
1224.66
|
988.69
|
3.67
|
SUR
|
12032.0.13
|
1224.05
|
987.00
|
3.67
|
RAW
|
12032.1.13
|
1224.48
|
988.57
|
3.67
|
SUR
|
12032.0.14
|
1224.13
|
987.00
|
3.67
|
RAW
|
12032.1.14
|
1224.60
|
988.36
|
3.67
|
SUR
|
12032.0.15
|
1224.44
|
987.00
|
3.67
|
RAW
|
12032.1.15
|
1224.71
|
989.11
|
3.67
|
SUR
|
12032.0.16
|
1224.53
|
988.00
|
3.67
|
RAW
|
12032.1.16
|
1224.57
|
989.21
|
3.67
|
SUR
|
12032.0.17
|
1224.68
|
988.00
|
3.67
|
RAW
|
12032.1.17
|
1224.65
|
989.43
|
3.67
|
SUR
|
12032.0.18
|
1224.79
|
988.00
|
3.67
|
RAW
|
12032.1.18
|
1224.67
|
988.00
|
3.67
|
SUR
|
12032.0.19
|
1224.91
|
988.00
|
3.67
|
RAW
|
12032.1.19
|
1224.54
|
989.25
|
3.67
|
In this test, the DCE0 SUR
slope is lower than the derived RAW slope by 15%. The differences between the
subsequent DCE RAW and SUR slopes is very similar to the on-orbit data.
Thus, we have demonstrated that for 4 second exposures, the on-board SUR
algorithm produces results very similar to ground testing. For DCE0, the
on-orbit SUR results are better than the ground results (although we are
only comparing 2 DCEs). Excluding the first DCE taken in an exposure sequence,
the SUR algorithm produces slopes within 1% of slopes derived from RAW DCEs.
Conclusions
For the first DCE commanded by each exposure sequence (single execution of CEMIPSUR),
the slope produced by the on-board SUR algorithm is systematically 9% lower than
slopes fit to comparable RAW mode data. This effect was seen (at an even worse level)
in ground testing. For other DCEs, the SUR algorithm produces slopes for unsaturated
pixels which are within 1% of slopes derived by linear, least-squares fitting of
RAW ramps. We conclude that until and if the DCE0 SUR disparity can be calibrated,
the first DCE taken by a command sequence should be discarded.
Output and Deliverable Products
None
Actions Following Analysis
The first DCE taken in each exposure sequence (DCE0000) should not be used
in deriving important calibration products or scientific results.
The IER used for this task in D2 was outdated, resulting in no useful comparison
data for 10 second and 30 second exposure times, and a non-optimal stimulator setting.
The 30 second RAW data would be especially useful for confirming that the 24um nonlinearity
has not changed on-orbit. The task should be repeated with an updated IER if there
is time in IOC/SV.