AGASC 1.6 is designed to correct an error in AGASC 1.5 which
overpredicts MAG_ACA, the predicted magnitude seen by the ACA,
for red stars (BT-VT > 1.0). We have also taken this
opportunity to improve the prediction of MAG_ACA_ERR by
including an error term for the ACA response as well as
calculating errors for stars which were previously given
truncated values. The purpose of this report is to present the
changes from AGASC 1.5 and to verify expected performance in
star selection through SAUSAGE. For details on the calibration
of AGASC 1.6 please refer to the AGASC
1.6 Calibration page.
General AGASC comparisons
AGASC 1.6 only effect stars from the Tycho-2 catalog. Below is a scatter plot of MAG_ACA all the Tycho-2 stars in the AGASC
against BT-VT color. All plots on this page are from this data
set.
Now we present the MAG_ACA_ERR for the new and old AGASC. The
most noticeable change is the addition of calculated errors for
COLOR2 > 1.76 (1.5/0.85), where AGASC 1.5 uses a truncated
error. Also notable is that the bluest stars have reduced
errors as a result of eliminating the extended polynomial fit.
The below distributions shows very little overall change in
MAG_ACA. We do not expect AGASC 1.6 to have a dramatic effect on
star selection, only to give a better representation of reality.
Below we see the distribution of errors for all Tycho-2 stars
in AGASC 1.5 (top) and AGASC 1.6. The artificially errors for
stars with COLOR2 > 1.76 has been eliminated, allowing these
stars to be included in early stages of star selection. The
errors for very good (small error) stars has increased slightly
in AGASC 1.6. This is due to the addition of a component to
account for errors in the ACA response. SAUSAUGE testing below
shows that this increased error has not greatly impacted the
stage in which stars are selected. The second series of plots
is the same data as the first, plotted on a semi-log scale.
SAUSAGE TESTING
Catalogs were generated using the current flight version of
SAUSAGE (2004251) and
then re-generated using an identical testing version linked to AGASC 1.6.
The loads used for testing were: MAY3104, JUN0704, JUN1404,
JUN2104, JUN2804, JUL0504, JUL1204, JUL1904, JUL2604, and AUG0204.
The distribution of magnitudes ( MAG_ACA ) is shown below. A
KS-Test was performed to compare the distributions and yielded a
statistic of 0.167 and a 99.1% probability that the two
distributions are from the same parent.
The SAUSAGE stage selection. The overall number of stars is
greater in AGASC 1.5 due to a slightly smaller number of combined
acq/guide stars.
The number of combined acq/gui stars per catalog.
The number of acquisition only stars per catalog.
The number of guide only stars per catalog.
The faintest magnitude star in each catalog. Notice that this
distribution is broader in AGASC 1.6. This is a result of the
decrease in predicted brightness for very red stars and is an
expected result.
The average star magnitude in each catalog. This distribution
has also broadened as expected.
Aspect Information main page
Comments or questions: Aspect Help
Last modified:12/27/13
|