L. Rudnick: RM2010
Hidden Rotation Measures
Lawrence RudnickUniversity of Minnesota
withDamon Farnsworth, UMN
Shea Brown, CSIRO
L. Rudnick: RM2010
Take Home Messages
• Two components with small RM differences can generate a wide range of complex behaviors.
• Polarization angle behavior over wide bands can be reasonably fit with RM* 2 – but give the WRONG answer
• RM synthesis can give the same WRONG answer as 2 fitting.
• CAVEAT: RM(2), P(2). RM determinations must be tailored to the specific scientific goals, and must utilize multiple techniques, including RM synthesis, Q, U vs. 2 and Q vs. U diagnostics.
L. Rudnick: RM2010
XRAYNVSS INVSS P800
L. Rudnick: RM2010
A little pre-history…
L. Rudnick: RM2010
Two simple components -- >polarization complexity
Q, U vs.
P vs.
0o 20o 40o 60o 80o 100o
AMP RM o
0.07 -5 90
0.05 -6 var
o,2
L. Rudnick: RM2010
3C33 Polarization
Leahy & Perley, 1991
WSRT352 MHz
Q
U
377.5 356.1 324.1 MHz MHz MHz
L. Rudnick: RM2010
3C33 Polarization
-7 rad/m2
L. Rudnick: RM2010
3C33 Polarization
-7 rad/m2
L. Rudnick: RM2010
3C33 Polarization
-7 rad/m2
-7 -5
-3 -1 +1
-9
-15 -13 -11
-7 rad/m2
RM synthesis
L. Rudnick: RM2010
3C33 Polarization:
?
-7 rad/m2
L. Rudnick: RM2010
3C33 Polarization
AMP RM o
0.07 -3.1 93
0.05 -0 46
L. Rudnick: RM2010
3C33 Polarization
AMP RM o
0.07 -3.1 93
0.05 -0 46
RM (2)
L. Rudnick: RM2010
3C33 Polarization – the range of fits
-10 -5 0 5 10
-10
-
5
0
5
10
RM stronger component
RM
wea
ker
com
pone
nt
2
L. Rudnick: RM2010
Other 3C33 fits…Exclude 20cm
Note: FACTOR of 2-3 in RMs
-10 -5 0 5 10
-10
-
5
0
5
10
RM
s:
-3.1
, -
0.52
RM
s: -7.3, -9.4
L. Rudnick: RM2010
RM synthesis and two components
-10,-2[b]
L. Rudnick: RM2010
RM synthesis and two components
• Faraday spectrum peaks (P) are NOT at the input RMs
• Whether two RMs are resolved depends on the PHASES of the components
L. Rudnick: RM2010
Addressing the challenge…
• Very broad 2 coverage
• Check for RM(2) and P(2)
• Use ALL of the available diagnostics
L. Rudnick: RM2010
Take Home Messages
• Two components with small RM differences can generate a wide range of complex behaviors.
• Polarization angle behavior over wide bands can be reasonably fit with RM* 2 – but give the WRONG answer
• RM synthesis can give the same WRONG answer as 2 fitting.
• CAVEAT: RM(2), P(2). RM determinations must be tailored to the specific scientific goals, and must utilize multiple techniques, including RM synthesis, Q, U vs. 2 and Q vs. U diagnostics.
Partial support for this work was provided by U.S. National Science Foundation GrantbAST-0908668 to the University of Minnesota.
L. Rudnick: RM2010
Take Home Messages
# COROLLARY: When looking at trends – e.g. redshift, cluster impact parameter, size, etc. systematic errors in RM can be fatal.
Partial support for this work was provided by U.S. National Science Foundation GrantbAST-0908668 to the University of Minnesota.
• Two components with small RM differences can generate a wide range of complex behaviors.
• Polarization angle behavior over wide bands can be reasonably fit with RM* 2 – but give the WRONG answer
• RM synthesis can give the same WRONG answer as 2 fitting.
• CAVEAT: RM(2), P(2). RM determinations must be tailored to the specific scientific goals, and must utilize multiple techniques, including RM synthesis, Q, U vs. 2 and Q vs. U diagnostics.
L. Rudnick: RM2010
L. Rudnick: RM2010
