Calculate Q and U images from a set of time-series data files CALCQU
If LSQFIT is set TRUE, then the output NDFs associated with parameters OUTQ, OUTU and OUTI
hold time series data in which the data values represent Q, U and I respectively, rather than raw
bolometer values. These time-series are much shorter in length than the supplied input time-series
files. Each input time series is split into blocks of adjacent time slices, and a single Q, U and I value is
created for each bolometer for each such block. The size of these blocks is specified by the POLBOX
configuration parameter, which is an integer giving the size of each block as a multiple of the
time taken for a complete revolution of the half-waveplate. Each (Q,U,I) triplet is found
by doing a least squares fit to the supplied input data (i.e. the analysed intensity data)
within a single block of time slices. The fitted function includes first, second, fourth and
eight harmonics of the half-waveplate, together with a linear background: ("
w"
is the
angle of the half-waveplate, and "
itime"
is the zero-based offset of the time slice into the
box):
y = Asin(4w) Bcos(4w) Csin(2w) Dcos(2w) Esin(w) Fcos(w) Gitime H Jsin(8w) Kcos(8w)
The returned Q, U and I values are then:
U = 2A Q = 2B I = 2( Gbox/2 H )
The Q and U values are specified with respect to either north or the focal plane Y axis (see parameter
NORTH). Each single pair of corresponding Q and U values in the output NDFs are created from a
single least-squares fit, and the residuals of each such fit are used to calculate a notional variance for
the corresponding pair of Q and U values. These are not "
real"
variances, but are just a scaled form of
the residuals variance using a scalaing factor that gives reasonable agreement to the visible noise in
the Q and U values measured in ten test observations. These variances are intended for
determining relative weights for the Q and U values, and should not be used as absolute variance
values.
If LSQFIT is set FALSE, then the output NDFs associated with parameters OUTQ, OUTU and OUTI
are each 2D and contain a single Q, U or I value for each bolometer. Multiple 2D images are created, as
the telescope slowly moves across the sky. Each image is created from a block of time slices over which
the sky position of each bolometer does not change significantly (see parameters ARCERROR,
MAXSIZE and MINSIZE). The resulting set of Q images can be combined subsequently using
KAPPA:WCSMOSAIC, to form a single Q image (normally, the "
I"
image should be used
as the reference image when running WCSMOSIAC). Likewise, the set of U images can
be combined in the same way. All the created Q and U images use the focal plane Y axis
as the reference direction (positive polarisation angles are in the same sense as rotation
from the focal plane X axis to the focal plane Y axis). Since this direction may vary from
block to block due to sky rotation, the idividual Q and U images should be processed using
POLPACK:POLROTREF before combining them using KAPPA:WCSMOSAIC, to ensure that they all
use the same reference direction. Q and U values are determined from the Fourier component of
each time series corresponding to the frequency of the spinning half-waveplate (6 - 12
Hz), and should be largely unaffected by signal at other frequencies. For this reason, the
cleaning specified by parameter CONFIG should usually not include any filtering that
affects frequencies in the range 2 -16 Hz. There is an option (see configuration parameter
SUBMEAN) to subtract the mean value from each time slice before using them to calculate Q and
U.
Separate Q, U and I esimates are made for each half revolution of the half-wave plate. The Data values in the returned NDFs are the mean of these estimates. If there are four or more estimates per bolometer, the output will also contain Variance values for each bolometer (these variances represent the error on the final mean value, not the variance of the individual values). A warning message will be displayed if variances cannot be created due to insufficient input data.
The current WCS Frame within the generated I, Q and U images will be SKY (e.g. Right Ascension/Declination).
"
def"
(case-insensitive) or a null (!) value is supplied, a set of default configuration parameter values will be
used.
The supplied value should be either a comma-separated list of strings or the name of a text file preceded by an
up-arrow character "
"
, containing one or more comma-separated lists of strings. Each string is either a "
keyword=value"
setting, or the name of a text file preceded by an up-arrow character "
"
. Such
text files should contain further comma-separated lists which will be read and interpreted in the same
manner (any blank lines or lines beginning with "
#"
are ignored). Within a text file, newlines can be
used as delimiters, as well as commas. Settings are applied in the order in which they
occur within the list, with later settings over-riding any earlier settings given for the same
keyword.
Each individual setting should be of the form:
keyword=value
The available parameters include the cleaning parameters used by the SC2CLEAN and
MAKEMAP commands, plus additional parameter related to the calculation of Q and U. Further
information about all these parameters is available in the file $SMURF_DIR/smurf_calcqu.def.
Default values will be used for any unspecified parameters. Assigning the value "
def"
(case insensitive) to a keyword has the effect of resetting it to its default value. Parameters not
understood will trigger an error. [!]
"
AZEL"
, then they use the elevation axis as the reference
direction, and if "
ICRS"
is supplied, they use the ICRS Declination axis. If "
TRACKING"
is
supplied, they use north in the tracking system - what ever that may be. If a null (!) value is
supplied, the Y axis of the focal plane system is used as the reference direction. Note, Stokes
parameters created with LSQFIT=FALSE always use focal plane Y as the reference direction. ["
TRACKING"
] "
OUTQ"
container file, but using "
U"
insead of "
Q"
in the NDF names. If LSQFIT is
TRUE, these will be a group of time series NDFs. No I data files are created if a null (!)
value is supplied. [!] "
Qi_s_c"
, where "
i"
is the integer one-based index of the time slice block from which the image was made, "
s"
is the name of the subarray (e.g. "
s4a"
, etc), and "
c"
is an integer one-based chunk index. If LSQFIT is TRUE, these will be a group of time
series NDFs. "
Ui_s_c"
, where "
i"
is the integer one-based index of the time slice block from which the image was made, "
s"
is the name of the subarray (e.g. "
s4a"
, etc), and "
c"
is an integer one-based chunk index. If LSQFIT is TRUE, these will be a group of
time series NDFs. "
"
syntax. An
example can be found in $STARLINK_DIR/share/smurf/resist.cfg [$STARLINK_DIR/share/smurf/resist.cfg]