Cyclotron Modeling Phase-Resolved Infrared Spectroscopy of Polars III: AM Herculis and ST Leo Minoris

To Appear in ApJ, 2008

Cyclotron Modeling Phase-Resolved Infrared Spectroscopy of
Polars III: AM Herculis and ST Leo Minoris

- - -, Ryan K. Campbell^'^ and Thomas E. Harrison^'^

O.

O . Astronomy Department, New Mexico State University, Las Cruces, NM 88003

(N

Stella Kafka

t-^ ■ Spitzer Science Center / Caltech, MS 220-6, 1200 E.California Blvd, Pasadena, CA 91125

Ph' abstract

o

^,

c/3 , We present phase-resolved low resolution infrared spectra of AM Her and ST

LMi, two low-field polars that we observed with SPEX on the IRTF. Optical/NIR
lightcurves are also published to help constrain the viewing geometry and bright-

^ \ ness of the objects at the time they were observed. Currently, only limited IR

■>!::j- \ spectra have been published for these objects, and none with the phase-coverage

l>

presented here. In both cases, the resulting spectra are dominated by emission
from the secondary star in the NIR. However, the emission regions are also self-

O \ eclipsed, allowing us to isolate the cyclotron emission through subtraction of the

(^ \ dim-phase spectrum. We use a "Constant Lambda" prescription to model the

changing cyclotron features seen in the resulting data. For AM Her, we find a
"O ; best fit model of: B = 13.6 MG, kT = 4.0 keV, and logA = 5.0. The cyclotron

H \ derived accretion geometry is consistent with i = 50° and j3 = 85°. For ST LMi,

B = 12.1 MG, kT = 3.3 keV, and logA = 5.7 with i = 55° and p = 128°.

Subject headings: Cataclysmic Variables: general — Polars: AM Her, VV Pup

^Visiting Astronomer, Infrared Telescope Facility. IRTF is operated by the National Aeronautic and
Space Administration

^Supported by New Mexico State Space Grant

•^Visiting Astronomer, Kitt Peak National Observatory

-2-

Introduction

Polars are interacting binary systems containing a primary white dwarf (WD) and a
late type secondary star. Material flows from the secondary, through the Li point and falls
ballistically toward the WD. The WDs in polars are highly magnetized with magnetic field
strengths that range between 10 - 240 MG. Eventually, the accreting material couples to the
field lines of the WD and is transported to the magnetic pole(s) of the star where a dense,
standing-shock is formed, with nominal temperatures of 2 - 20 keV, which cools by emitting
bremsstrahlung and cyclotron radiation .

AM Her is the prototype polar (see Tapia, 1977a). It has an orbital period of 3.094
hrs and is nearby (78 pc; Thorensten et al., 2003). Despite being extensively studied many
characteristics of the system remain uncertain. One point of discrepancy is the exact mass
of the primary WD with estimates ranging from 0.39 M0 (Young et al., 1981) to 1.22 M©
(Cropper et al., 1998). Additionally, the geometry of the system remains unclear with
orbital inclination estimates ranging from i = 35° (Brainerd & Lamb, 1985) to i = 60° - 80°
(Watson et al., 2003), although the self-eclipse observed in the X-Ray and UV requires that
i + (3 > 90°, where (3 is the magnetic co-latitude. Indeed, both Sirk & Howell (1998) and
Gansicke et al. (1998) found that i + f3 = 105°. Later, Gansicke et al.(2001) determined that
combinations of i and j3 between {i = 50°, f3 = 55°) and {i = 35°, (3 = 70°) best modeled the
high-state optical lightcurves. In other ways, however, AM Her is well characterized. The
temperature of the primary WD has been well constrained. Gansicke et al. (2006) modeled
low-state FUSE and STIS spectra from AM Her with TLUSTY/SYNSPEC (Hubeny & Lanz,
1995), finding that T^^^ = 19800 ± 700 K. Also, the secondary star is spectroscopically
determined to be M4 ± 1 (Kafka et al., 2005b), although there is some evidence that it is
irradiated by the WD and thus the spectral type of the secondary is orbitally modulated
(Davey & Smith, 1992). The published photometry of AM Her is exhaustive. Orbital
lightcurves show variability in every pass-band from the UV (Gansicke et al. 1998) out to
the if -band (see below). Extensive AAVSO and automatic photometric telescope, "APT"
(e.g. RoboScope) monitoring from 1990 - 2004, has revealed that AM Her is usually in
one of two states: a "high-state", with large intrinsic variability: 13.0 < V^ < 14.0, and a
"low-state" with V ~ 15.5 (Kafka et al., 2005a). Finally, Bailey, Ferrario, & Wickramasinghe
(1991; henceforth BFW91) used a Constant Lambda ("CL") code to model the NIR low-state
(V ~ 15.0) cyclotron spectrum, which were binned into bright-phase (0 = 0.46 - 0.88) and
dim-phase (0 = 0.46 - 0.88) spectra. The dim-phase showed only emission from the secondary
star, while strong cyclotron emission was observed for the duration of the bright phase. After
subtracting a 3250 K model atmosphere with log g = 4.75 to mimic the secondary spectrum,
BFW91 found B ~ 14.5 MG, and a shock temperature of kT = 8.5 keV.

-3-

ST LMi (= CW1103 +254) is a short period polar {Porb = 114 min) containing a 0.7
Mq primary (Ramsay et al. 2004) with a hkely temperature of 11000 K (Araujo-Betancor
et al, 2005; Sion, 1999) and a M5 - M6 secondary (Knigge, 2006; Harrison et al., 2005;
Howell et al., 2000; Warner, 1995) at a distance of 115 - 138 pc (Araujo-Betancor et al.,
2005; Kafka et al., 2007). It was classified as an AM Her object by Stockman et al. (1983)
on the basis of its highly variable polarization. For 70 % of the orbit the object shows no
significant polarization. Subsequently, a strong pulse is observed peaking near = 0.00 at 12
% and -20 % in linear and circular light, respectively (Cropper et al., 1986). This observed
bi-modality is echoed in optical and IR orbital lightcurves, which show a quiescent "dim-
phase" for most of the orbit in each band that is followed by a significant jump in brightness
(the "bright-phase") coincident to the peak in polarization. Peacock et al. (1992) obtained
multi-band photometry of ST LMi, showing that Am ~ 0.6, 1.4, 2.0, 1.5, and 1.3 for the
BRIJH bands, respectively. Long-term \/-band lightcurves were obtained with RoboScope
from 1990 to 2003 (Kafka et al., 2005a). From 1992 - 1997, the system was in a protracted
"low-state" with <V> = 17.5 ± 0.2. From 1997 - 2003, a more variable, slightly higher state
was observed with <V> = 16.0 ± 1.5. Additionally, instances of "extreme low-states" have
been observed. In Ciardi et al. (1998) the K-hand spectra of ST LMi, showed no obvious
emission lines and were modeled successfully with a ^ 3000 K atmosphere suggesting that
accretion had almost completely shut off. Kafka et al. (2007) has presented photometry of
a similar extreme low-state showing that the system can be as faint as V = 18.5. JHK
cyclotron spectra were previously modeled by Ferrario, Bailey & Wickramasinghe (1993;
henceforth FBW93) finding that two spots were necessary to fully model their spectra: a
primary region with B = 12.0 and kT = 12 keV, and a secondary, "cool-spot" with kT =
5.0 keV. The accretion geometry of the primary emission region has also been previously
determined, with 55° < i < 64° and 140° < (3 < 150° (Schmidt et al., 1983; Potter, 2000).

Limited IR spectroscopy exists for these two low-field polars. Below, we present and
model new phase-resolved low-state infrared spectra as well as JHK lightcurves for AM Her
and ST LMi. In both instances, we show that variable cyclotron emission over the orbit
is responsible for the spectroscopic and photometric behavior. Additionally, we present
a second epoch data set for ST LMi that shows no cyclotron emission and must be in an
extreme low-state similar to that seen by Kafka et al. (2007). In the next section, we describe
the observations of each object, in section 3 we fit these data with cyclotron models, discuss
our results in section 4, and draw our conclusions in section 5.

-4-

2. Observations

AM Her and ST LMi were observed using SPEX (c.f. Rayner et al., 2003) on tlie Infrared
Telescope Facility (IRTF). AM Her was observed once on 2005 September 1, whereas ST
LMi was observed on two different epochs: 2005 Feb 7 and 2006 Feb 2. Both AM Her and
ST LMi were found to be in low-states, although as we discuss below, the 2005 Feb 7 data
found ST LMi in an extreme low-state. SPEX was used in low-resolution "prism" mode
with a 0.3" x 15" slit. To remove background, each object was nodded along the slit. In its
low-resolution mode SPEX produces R(=A/AA) ~ 250 spectra, with short enough exposure
times to obtain phase resolved spectra of polars with K < 16.0. For ST LMi, we used 240
second exposure times, where shorter, 120 s, integration times were adequate for AM Her.
Each of these spectra were then median combined with 2-3 other spectra to allow for cosmic-
ray removal and to improve the S/N ratio. The spectra were reduced using the SPEXTOOL
package (Vacca et al., 2003). A telluric correction was applied using an AOV star of similar
airmass to our program objects. We use the Kafka et al. (2005b) ephemeris to phase all
observations of AM Her. For ST LMi because of the large phase uncertainty (A</) ~ 0.10)
in the Howell et al. (2000) ephemeris, we phased our observations to the J-band minimum
found in the photometry presented in this paper, which worked out to a phase-shift of A0
= 0.15 from that ephemeris.

Because of the narrow slit size on IRTF/SPEX (0.3") infrared photometry is required
to calibrate the fluxes of the spectra. The JHK photometry for each object was obtained
with SQIID on the KPNO 2.1-m telescope (Ellis et al., 1992). AM Her was observed on
2002 September 26, and ST LMi on 2003 April 9. In addition, we obtained simultaneous
BVRIJHK photometry on 2005 May 20 for AM Her. The JHK photometry was obtained
with NIC-FPqll on the Apache Point 3.5-m, while the optical data set was obtained with
the NMSU 1-m (see Harrison et al., 2003). To aid the reader, we have collated all the
observational specifics in Table 1.

3. Modeling

To produce our cyclotron models, we use a Constant-Lambda ("CL") cyclotron code
first developed by Schwope (1990). In Campbell et al. (2008a; hereafter paper I), we
presented a theoretical synopsis of CL modeling which will not be repeated here. The model
spectra depend on four global parameters: B (the magnetic field strength), kT (the plasma

^see http://www.apo.ninsu.edu/arc35in/Instriiments/NICFPS/nicfpsusersguicie.html

-5-

temperature), G (the viewing angle to the magnetic "pole"), and A (the "size parameter"),
which is closely tied to the column density along the line of sight through the accretion region.
In paper I, we found that we could adequately model the data for EF Eri as cyclotron +
WD. In Campbell et al. (2008b; hereafter paper II), we found that in many polars there
are other sources of non-stellar continuum radiation (e.g., Bremsstrahlung emission) which
contaminate the spectra and need to be taken into account. For each object in paper II, the
accretion column was self-eclipsed. In this case, cyclotron emission is only seen for the part
of the orbital cycle when the accretion column is in view (the "bright-phase") although it
is contaminated by other sources. To subtract these away, we assume that the "dim-phase"
spectra represent all the additional components of radiation which obfuscate the cyclotron
emission. The dim-phase spectrum is then subtracted from the spectra at other phases
where accretion column is in view, thus yielding uncontaminated cyclotron spectra over the
orbit. We refer to the dim-phase subtraction method as "Stream- Emission Subtraction"
(SE-subtraction) for consistency with Schwope et al. (2002). An additional contaminant for
the objects in the current work is the irradiated secondary star whose spectral type slowly
changes with phase and can not be completely subtracted. Thus, when the SE-subtraction
technique was applied features due to the secondary star remained.

3.1. AM Her

In Figure 1, we show the JHK lightcurves taken with the SQIID on the KPNO 2.1-m
on 2002 September 26 , at a time when the system was at V = 15.5. This is a typical low-
state magnitude identical to that of our SPEX spectroscopy, as shown in Fig. 2. Overlaid
in each band, are binary star models computed using WD200q3 with a M5 secondary at an
orbital inclination oii = 50°. The J-band morphology is well explained by classic ellipsoidal
variations except the lightcurve minimum at = 0.00 is somewhat deeper than predicted
and residual structure appears at the AJ = 0.05 mag level. The derived inclination should
be considered a lower limit because other dilution sources may be present. Both the H +
K lightcurves show a large cyclotron component folded-in with the ellipsoidal variations.
In Fig. 3, we include additional BVRIJHK photometry obtained 2.5 years later on 2005
May 20, at a time when the system was again at similar brightness. We find that the same
ellipsoidal models provide excellent fits to the JHK data at this epoch. Because of the
narrow slit size on SPEX (0.3"), at each orbital phase we flux calibrate our spectra to the
"cyclotron-free" J-band lightcurve.

^WD2005 is an updated version of WD98, and can be obtained at this website maintained by J. Kallrath:

http://josef-kallrath.orlando.co.nz/HOMEPAGE/wd2002.htm

-6-

The IRTF phase-resolved spectra from 2005 September 1 are dominated by emission
from the secondary star. To remove this component, we subtracted the spectrum at
= 0.42 from every other phase. The subtraction spectrum is near to both the elhpsoidal
minimum (0 = 0.50) and because of the ongoing self-echpse, is free of cyclotron emission.
To approximate the effect of the ellipsoidal variability, we scaled the subtraction spectrum
at each phase to match the magnitude expected from our ellipsoidal models. The underlying
continuum SED and intrinsic water vapor features at 1.35 and 1.85 //m in the residual spectra
were orbitally variable producing a small blue excess and apparent water vapor emission at
ellipsoidal maxima (0 = 0.25, 0.75), and a red excess with apparent water vapor absorption at
ellipsoidal minima (0 = 0.00, 0.50), resulting from a changing spectral type as the distorted
secondary star changed orientation. From M4 to M6 water vapor absorption becomes ever
more pronounced. Thus, even small differences of the secondary temperature are apparent
in our data and residuals from the water vapor features remain in our final spectra (shown
in black in Fig. 4). The final cyclotron models are overlaid in green. No cyclotron emission
was observed over the interval 0.27 < < 0.74 due to the self-eclipse of the emission region.
Thus, these phases are not shown in Fig. 4 to aid in the presentation of data. Between 0.92
< < 0.09, the spectra appear to show a single strong cyclotron harmonic {n = 4) near 2.0
fim. The n = 4, 5, and 6 harmonics are obvious between 0.20 < < 0.26 and again from
0.75 < < 0.86.

For our best fit cyclotron models at each phase, see Table 2. The average parameter
values are: B = 13.61^;^ MG, kT = A.OtH keV, and logA = b.Otol with an average xl =
2.42. The statistical limits were derived by finding where the value of xl changed by 50 %
over its fiducial value. We find that i + P = 135°, with i = 50°, P = 85°, (pmin = 0.01, where
0mm is defined as the bluest position of the cyclotron harmonics.

3.2. ST LMi

In Fig. 5, we present JHK lightcurves of ST LMi taken with the KPNO 2.1-m during
the normal high accretion state of the system. In each band the morphology is similar: the
dim-phase lasts from 0.00 < < 0.55 with mean magnitudes of 14.4, 13.9, and 13.9 in the
J, H , and i^-bands, respectively. During the subsequent bright-phase, the object brightens
significantly (AJ = 1.5 mag). Like AM Her, a WD2005 ellipsoidal model was fit to the NIR
lightcurves, finding i = 55°. The models well approximate the dim-phase of ST LMi in the
J and iif-bands, while the fit in the ii'-band is more uncertain due to the larger scatter in
the photometry.

ST LMi was observed spectroscopically over its entire orbit once on 2005 February 7

-7-

and once on 2006 February 2. On both occasions, photometry is available within one month
of our phase-resolved spectroscopy. Fig. 6 shows the high/low states of ST LMi over the
entire history of RoboScope and also over a shorter, ~ 1.5 year baseline surrounding our 2005
and 2006 IRTF observations. For the first dataset, the nearest RoboScope data (February
28) shows 17.1 <V< 17.8, similar in brightness to its 1992 - 1997 protracted "low-state".
However, as will be discussed below, our bright-phase spectroscopy show a conspicuous lack
of cyclotron emission during this epoch. For this reason, we believe the object was in fact in
an "extreme low-state" at the time of observation, similar to that observed on 2006 Feb 12
by Kafka et al. (2007) which found 18.0 < V^ < 18.4. Curiously, our 2006 data was obtained
only 10 days prior to that epoch, but shows clear evidence of cyclotron emission and thus
must have been in a normal low-state. We note, however, photometric extreme low-states
like that observed by Kafka et al. are short lived, as normal low states were observed within
a month both before and after it.

The spectra of ST LMi are also strongly contaminated by its secondary star at every
phase. During the extreme low-state we found no evidence for cyclotron emission, with the
secondary contributing all of the NIR flux at each orbital phase. In Fig. 7, we show the
observed SPEX spectrum at = 0.02 of the 2005 dataset with the best-fitting secondary
template overlaid (M6).

For the 2006 low-state, we present SE-Subtracted data in Fig. 8a. The SE spectrum was
produced by averaging the dim-phase spectra together. Since the relative uncertainty in the
Howell et al. (2000) ephemeris is rather large (A0o — 0.10), we phased our data by defining
ellipsoidal minimum in our 2003 KPNO lightcurve as = 0.50, and then using the Howell et
al. (2000) period. We found that averaging three dim-phase spectra together produced the
best SE-subtraction spectra, with faint cyclotron features visible during the bright orbital
phase at ~ 2.25, 1.85, 1.53 and perhaps 1.30 /im, corresponding to the n = 4 - 7 harmonics
in a field with B = 12 MG. Like AM Her, the tidally distorted nature of the secondary
star imparts spectral type and overall flux changes that are orbitally modulated allowing
residual emission/absorption from the secondary star to remain even after SE subtraction
has been performed. Indeed, the upturn in the J-band SED as well as the strong water vapor
absorption at ~ 0.73 are indicative of subtraction of too cool a secondary star at those
phases. In Fig. 8b, we display the phase-resolved spectra from the 2005 extreme low-state.
Like the 2006 low-state the SE spectrum was found by averaging three dim-phase spectra
together. While weak cyclotron emission was observed during the 2006 low-state none was
seen during the 2005 extreme low-state.

Previous work has determined that ST LMi is a one pole accretor with 55° < i < 64°
and 140° < (3 < 150° (Schmidt et al., 1983; Potter, 2000). The primary accretion region

also appears to have some structure. FBW93 computed cyclotron models for ST LMi in a
high state, finding that two separate emission regions were needed to adequately model the
observed spectra from the primary pole: the first is a high temperature/high density region
located between magnetic longitudes {ip) 130 and 170 that has kT = 12.0 keV, and logA
= 7.6. The second is located between 170 < ip < 250 with kT = 5 keV and logA = 4.4.
Both regions had magnetic field strengths of ^ 12.0 MG, a result which is consistent with
the values previously published. The relevancy of these models to our SPEX data is unclear
since they were determined when the object was in a high state (J = 13.8) more similar
to that observed in our SQIID photometry than our low-state spectroscopy. In addition,
Peacock et al. (1992) reported seeing a second pole in ST LMi during a high-state {B ~
16.7). Their iJ-band photometry showed a sudden increase of AH = 0.3 mag at = 0.35
was observed, with the J-band showing a smaller increase. Simultaneous polarization curves
were published along with their photometry that found V/I = 15 - 20% over the duration
of the alleged secondary pole. The errors, however, were extremely large and the data are
in fact, consistent with zero polarization. Because both the phasing and amplitude of the
lightcurve variations are consistent with the WD2005 models found in this study (Fig. 5), we
believe the excess i^-band feature may be due to ellipsoidal variability from the secondary
star.

We used the published values i and j3 to constrain the orbital variation of O and thus,
effectively limit the possible parameter space to three dimensions, B, kT and logA. Because
of the very low rate of accretion, the models presented here have low temperatures. We
found 12.0 < B(MG) < 12.2, 3.2 < kT(keV) < 3.4, and 5.5 < logA < 6.1, and 6 which is
well described with i = 55°, f3 = 128°, (pn^ax = 0.22± 0.05. The average xl = 2.70. As for
AM Her , we find the following uncertainties on the parameter values: B ± 0.5 MG, kT ±
1.8 keV, and logA ± 0.6. Table 3 lists the phase-resolved parameters.

4. Discussion

4.1. Analysis of Results

We have presented the first phase-resolved IR spectroscopy published for these two
sources allowing for enhanced leverage over the parameters modeled. The data presented in
this paper proved the most challenging to model in this series. In Paper I, we covered the
basics of cyclotron modeling and applied our technique to NIR spectra of EF Eri finding
that it could be modeled using only cyclotron and WD emission. In Paper II, we found that
other sources of contamination could be eliminated by subtracting the dim phase spectrum
from each phase for which cyclotron emission was observed. In the present work, however.

-9-

the spectra of both AM Her and ST LMi are dominated by emission from the secondary
star in the NIR. Because of the orbitally modulated nature of both the spectral type and
brightness of the secondary, additional issues arose. Naive subtraction of only a single dim-
phase spectrum from every other phase is manifested in two ways. First, variable water
vapor absorption/emission is seen at at 1.35, 1.85 /xm. Second, the underlying SED cycles
twice between a red and blue excess over the orbit. Both artifacts result from the changing
spectral type of the secondary star. Because we subtract the same phase from each spectrum,
the secondary imprint in the SE-spectrum alternates between being too cool and too hot
when compared to the features seen at other phases. To assuage the situation, we median
combined three dim-phase spectra separated by A0 = 0.25, to produce the final SE spectrum,
thus smearing out the effects of a changing secondary star.

In AM Her, we found that for the bright-phase of the 2006 low-state (V ~ 15.5) SPEX
data: B = 13.7 ± 1.0 MG, kT = 4.2± 1.0 keV, and logA = 5.0± 0.5. The resuh is in
dramatic contrast with that found in BFW91 at a time when the system was also in a low,
though perhaps slightly higher state (V ~ 15.0): B = 14.5 ± 0.3 MG, kT = 8.5 ± 0.5 keV,
and logA = 3.3 ± 0.3. While the magnetic field strengths for the two epochs agree to within
their errors, the same cannot be said for both the temperatures and values of logA. Our
SPEX data show a system with a cooler plasma than that inferred by BFW91. To quantify
this difference, we evoke the results of Fischer & Beuermann (2001), which found kT max
oc 7hB~^'^, which we rewrite as: mi/m2 = {kTi/kT2) {Bi/ B2f"^ . Plugging in the average
bright-phase parameters from the two epochs yields that the rh for the modeled SPEX data
must be a factor of 2.34 lower than that active during the BFW91 observations. The higher
temperatures found in BFW91 may be an artifact of medianing together moving cyclotron
harmonics over ~ 40 % of the orbit, which artifically broadens each feature.

The final spectra and models allow us to understand the changing morphology of AM
Her's NIR light-curves. In Fig. 3b, the J-band is well explained by ellipsoidal variations
alone, while in the H + i^-bands the cyclotron emission component is substantial, with
a maximum contribution of 0.25 mag in both bands and disappearing at = 0.00. The
i^-band cyclotron component is relatively constant over the phases 0.75 < < 0.25. Such
behavior is explained by the cyclotron models shown in Fig. 4. Near = 0.00 the n = 4
harmonic dominates due to the low viewing angle (0 ~ 35°) at that time. Consequently, no
emission is seen in the ff-band. Later, (0.25 < < 0.75) the viewing angle is larger causing
the higher harmonics (n = 5 and 6) to be excited and thus a peak in the i^-band emission
is observed. Because the n = 4 harmonic is mostly optically thick, however, the cyclotron
emission in the i^-band remains relatively constant over the entire bright-phase.

The simultaneous optical photometry are also interesting. In the R and / bands, the

-10-

data were entirely explained by the ellipsoidal models alone for the first full orbit of phase
coverage, in line with expectations from our 13.7 MG cyclotron models that have very little
emission shortward of 1.2 fira. Subsequently, a brightening event seems to have occurred with
Am = 0.20 in both bands. The B and V^-bands are more complex and cannot be explained
by ellipsoidal models alone. Additional modulation was observed at the level of Am = 0.22
and 0.08 for B and V, respectively. Gansicke et al. (1998) modeled similar UV lightcurves as
a hot-spot. In that work, three UV bands covering wavelength regions of 1150 - 1167A, 1254
- 1286A, and 1412 - 1427A were found to be consistent with a 47000 K hotspot centered on
ip = 0° and covering 9 % of the WD surface. We find the similarity of our 5-band lightcurve
to Gansicke et al.'s UV lightcurves to be striking. Both lightcurves show identical phasing
and have amplitudes that are consistent. We note, however, the large value of -B - V^ ~ 0.15
during the bright phase. For any reasonable hotspot, B - V should be closer to 0.00. Because
(a) the UV lightcurves predict a very similar geometry {i, j3, and phase) to our cyclotron
emission region and (b) the limited wavelength coverage of the UV lightcurves, we find that
this excess emission could be caused by a partially saturated high-field cyclotron harmonic
(n = 3 or 4) that falls off toward the blue end of the l^-band and extends through the bluest
UV band. Such a broad harmonic (AA = 0.4 /im) is expected in cyclotron emission (see AM
Her's n=4 harmonic in Fig. 4, which is more or less fiat from 1.9 fira to 2.4 /im.) If the
emission were from the n = 4 harmonic it would imply a ~ 90 MG field. We also speculated
on the presence of a similar secondary high-field pole for EF Eri in Paper I.

During the bright phase of the 2006 low-state, ST LMi displayed cyclotron emission
with the following properties: B = 12.1 ± 0.5 M, kT = 3.3 ±1.8 keV and logA = 5.7 ± 0.6
similar to the "cool spot" found in FBW93. In addition, our accretion region appears to be
in a similar location on the WD surface: at magnetic longitude ip = 120°, lagging behind
the onset of the secondary accretion region found in FBW93 by A0 ~ 0.13, likely due to
the accumulation in phase-error between 1991 and 2005. Unfortunately, no errors are given
for any of the cyclotron parameters in FBW93 and thus the significance of the difference
in results is hard to assess. However, both the FBW93 magnetic field strength, and the
plasma temperature agree to within our errors. More interesting is the non-detection of
their primary accretion region which should trail the observed "secondary" emission region
by about 0.10 in phase, implying an onset at ^ 0.60 which is not seen.

4.2. Ellipsoidal Versus Cyclotron Derived Inclinations

In both cases, the geometry of the emission region was consistent with a simple single
spot model having a constant orbital inclination and magnetic colatitude. In Fig. 9, we plot

-11-

(black) the cyclotron derived values of the viewing angle against the orbital phase for both
AM Her (top) and ST LMi(bottom). In red are the simple geometrical models, with the blue
shading indicating phases for which the cyclotron regions are self eclipsed. The models fit the
AM Her data well for nearly all phases, only deviating near self-eclipse ingress and egress,
where the viewing angle is changing rapidly compared to the cadence of our spectroscopy.
For ST LMi, the bright phase is relatively short, lasting ~ 40 % of the orbit. Consequently,
few data were available to constrain its geometry. We thus used published values of i and f3
for an additional constraint, finding that we could match the data with models quite similar
to those found in the literature.

For AM Her, the cyclotron models imply an orbital inclination oi i = 50°, identical to
that found in the ellipsoidal modeling effort, although higher angles are possible if additional
sources of dilution remain in the IR light curves. Agreement was also found for the ellipsoidal
and cyclotron inclinations for ST LMi finding values of i = 55° and 40°, respectively. The
later value represents the lowest inclination ellipsoidal model, and values oii = 55° are more
consistent with the lightcurve (see Fig. 3). Some caution, however, should be given to the
fact that the spectroscopy and lightcurves of ST LMi were taken at different epochs when
the object was in different states, which could affect the accretion geometry. In Schwope et
al. (1993), the authors found that the polar MR Serpentis appeared to show longitudinal
migration of the accretion spot by ~ 30°, as well as a 10° shift in the magnetic colatitude
between its high and low states.

4.3. ST LMi in an Extreme Low-State

In Fig. 8b, we present SE-Subtracted SPEX data obtained during our 2005 February
7 observing run, which show a conspicuous lack of cyclotron emission throughout the entire
orbital cycle. The extreme low-state of ST LMi is corroborated by near-epoch (2006 February
12) optical lightcurves obtained with the WIYN 0.9-m, showing the system in a deep low-
state. Despite the poor telluric correction (the spectra were faint), the only strong feature
in the bright-phase (0.50 < < 0.85) is small "bump" longward of 2.2 /im, caused by under
subtraction of the secondary star at those phases. In this series of papers (see Papers I, II,
and Szkody et al. 2008, submitted) we have modeled seven polars, representing ~ 10%
of all known mCVs. Included in this sample were EQ Get, the prototype "Low-Accretion
Rate Polar", MQ Dra (= SDSS 1553), the prototype "pre-polar", and EF Eri, well known
for its protracted low-state. Intriguingly, in each of these objects strong cyclotron emission
was observed, while the extreme low-state dataset of ST LMi is the only example where
cyclotron emission completely disappeared. A similar situation probably explains the 2006

-12-

Feb 12 photometric low-state found by Kafka et al. (2007). This suggests that normal polars
can have periods that appear completely devoid of detectable accretion.

5. Conclusion

We obtained a full orbit of phase-resolved IR spectra for both AM Her and ST LMi. We
found both objects to be dominated by emission from the secondary in the IR. To remove
this component, we utilized the fact that emission regions for both stars were self-eclipsed.
Thus, at each phase we subtract a dim-phase or "Stream-Emission" spectrum. Because of
the changing spectral type of the secondary, we found that medianing dim-phase spectra
over ~ 25 % of an orbit produced a better subtraction. For AM Her, we found a phase
averaged model of : B = 13.6^'^og MG, kT = 4.0lJ'o ^^^ (^^^ Table 2 for specifics at each
phase). Additionally, we found that the viewing angle varied in a manner consistent with
expectations from a system with i = 50° and j3 = 85°. For ST LMi, we collected two
datasets. The first had ST LMi in a low-state with V ~ 17.4, and displayed weak cyclotron
harmonics that were difficult to decouple from the water vapor signatures leftover after SE-
subtraction. We found a phase averaged model of : B= 12.1 ± 0.5 MG, kT = 3.3 ±1.8
keV in an accretion region consistent with i = 55° and (3 = 128°. For ST LMi, we include a
second data set, taken when the object was in an "extreme low-state" showing no substantial
cyclotron emission. The non-detection of cyclotron emission contrasts with our earlier data
from both EQ Get and MQ Dra both of which show cyclotron emission, even while being in
extremely low-states.

REFERENCES

Araujo-Betancor, S., Gansicke, B.T., Long, K.S., Beuermann, K., de Martino, D., Sion,
E.M., Szkody, R, 2005, ApJ, 622, 589

Bailey,J., Ferrario,L. & Wickramasinghe, D.T., 1991, MNRAS, 251, 37

Brainerd, J.J & Lamb, D.Q., 1985, Proceeding of the 7th North American Workshop on
Gataclysmic Variables and Low-Mass X-Ray Binaries, 247

Gampbell, R.K., Harrison, T.E., Schwope, A.D., Howell, S.B., 2008, 672, 531

Gampbell, R.K,, Harrison, T.E, Mason, E. Howell, S.B., Schwope, A.D., 2008, ApJ, in press.

Giardi, D., Howell, S.B., Dhillon, V.S., Wagner, R.M., Hauschildt, RH, Allard, F., 1998,
PASP, 110, 1007

-13-

Cropper, M., 1986, MNRAS, 222, 853

Cropper, M., Ramsay, G. & Wu, K., 1998, MNRAS, 293, 222

Davey S. & Smith, R.C., 1992, MNRAS, 257, 476

Ellis, T., Drake, R., Fowler, A.M., Gatley, F., Heim, J., Luce, R., Merrill, K.M., Probst, R.,
Buchholz, N., 1992, Proc. SPIE, 1765, 94

Ferrario, L., Bailey, J., & Wickramasinghe, D.T., 1993, MNRAS, 262, 285

Fischer, A. & Beuermann, K., 2001, A&A, 373, 211

Gansicke, B.T, Hoard, D.W., Beuermann, K., Sion, E.M, Szkody, P., 1995, A&A, 338, 933

Gansicke, B.T., Fischer, A., Silvotti, R., and de Martino, D., 2001, A&A, 372, 557

Gansicke, B.T, Long, K.S. Barstow, M.A. & Hubeny, I., 2006, ApJ, 639, 1039

Harrison, T.E., Howell, S.B., Huber, M.E., Osborne, H.L., Holtzman, J. A., Cash, J.L.,
Gelino, D.M., 2003, AJ, 125, 2609

Harrison, T.E., Howell, S.B., Szkody, P., Cordova, F., 2005, ApJ, 632, 123

Howell, S.B, Ciardi, D.R., Dhillon, V.S., Skidmore, W., 2000, ApJ, 530, 904

Hubeny, I. & Lanz, T., 1995, ApJ, 439, 875

Kafka, S. & Honeycutt, R.K., 2005a, AJ, 130, 742

Kaflca, S. & Honeycutt, R.K., S.B., Harrison, T.E., 2005b, AJ, 130, 2852

Kaflca, S., S.B., Honeycutt, R.K., Robertson, J.W., 2005, AJ, 133, 1645

Knigge, C, 2006, MNRAS, 373, 484

Peacock, T., Cropper, M., Bailey, J., Hough, J.H., Wickramasinghe, D.T., 1992, MNRAS,
259, 583

Potter, S.B., 2000, MNRAS, 314, 672

Priedhorsky, W.C. & Krzeminski, W., ApJ, 219, 597

Ramsay, G., Cropper, M., Wu, K., Mason, K. O., Cordova, F. A., Priedhorsky, W., 2004.
MNRAS, 350, 1373

-14-

Rayner, J.T., Toomey, D.W., Onaka, P.M., Denault, A.J., Stahlberger, W.E., Vacca, W.D.,
Gushing, M.C., Wang, S., 2003, PASP, 115, 362

Schmidt, G.D., Stockman, H.S., & Grandi, S.A., 1983, ApJ, 271, 735

Schwope, A.D. 1990, Reviews In Modern Astronomy, 3, 44

Schwope, A.D., Beuermann, K., Jordan, S., Thomas, H.-G., 1993, A&A, 278, 487

Schwope, A.D., Brunner, H., Buckley, D., Greiner, J., Heyden, K. v. d., Neizvestny, S.,
Potter, S., Schwarz, R., 2002, A&A, 396, 895

Sion, E., 1999, PASP, 111, 532

Sirk, M.M. & Howell, S.B., 1998, ApJ, 506, 173

Southwell, J. A., Still, M.D., Gonnon-Smith, R., Martin, J.S., 1995, A&A, 302, 90

Stockman, H.S., Foltz, G.B., Schmidt, G.D., Tapia, S., 1983, 271, 725

Szkody, P. & Gapps, R.W., 1980, AJ, 85, 882

Szkody,P., Linnell, A. P., Gampbell, R.K., Plotkin, R.M., Harrison, T.E., Holtzman, J.,
Seibert, M., Howell, S.B., 2008, ApJ, submitted, 73798

Tapia, S. (1977), ApJ, 212, L125

Thorstensen, J.R, 2003, AJ, 126, 3017

Vacca, W.D., Gushing, M.G., Rayner, J.T., 2003, PASP, 115, 389

Warner, B., 1995, Astrophysics & Space Science, 232, 89

Watson, G.A., Dhillon, V.S., Rutten, R.G.M., Schwope, A.D., 2003, MNRAS, 341, 129

Wilson, R.E. & Devinney, E.J., 1971, ApJ, 166, 605

Young, P., Schneider, D.P., Shectman, S.A., 1981, ApJ, 245, 1043

This preprint was prepared with the AAS I^TJtjX macros v5.2.

-15-

Table 1. Observing Log

Date

Objeet

Instrument

Obs.

Type

I(sec)

State

2005 September 1

AM Her

IRTF/SPEX

Spec.

120

Low

2002 September 26

AM Her

KPNO

Phot.

Low

2005 May 20

AM Her

APO

Phot.

240

Low

2005 February 7

ST LMi

IRTF/SPEX

Spec.

Extreme Low

2006 February 2

ST LMi

IRTF/SPEX

Spec.

Low

2003 April 9

ST LMi

KPNO

Phot.

240

High

-16-

Table 2. Cyclotron Modeling Parameters for AM Her

Phase B(MG)

kT(kcV)

e

ogA xt

0.03 13.3

3.9 35.0

5.3 3.53

0.09 13.1

4.1 42.0

5.3 3.90

0.20 14.0

3.9 62.0

4.8 1.11

0.26 14.1

3.9 72.0

4.6 2.12

0.31

0.37

0.48

0.52

0.59

0.64

0.70

0.75 1

3.8

4.3 7

5.0

5.(

) 1.78

0.80 1

3.6

4.3 e

7.0

5.^

I 1.30

0.86 1

3.6

4.2 £

5.0

5.1

L 3.30

0.92 1

3.4

4.0 4

6.0

5.1

L 2.28

0.97 1

3.3

3.9 a

7.0

5.^

t 2.48

-17-

Table 3. Cyclotron Modeling Parameters for ST LMi

Phase

B{MG)

T(keV)

e

logA

v2

0.14

0.26

0.38

0.50

0.57

12.0

3.2

85.0

5

5

2.19

0.68

12.0

3.2

78.0

5

5

2.61

0.73

12.2

3.4

75.0

6

1

3.75

0.85

12.2

3.4

82.0

5

9

2.24

0.91

0.99

-18-

Figure 1. JHK photometry of AM Her obtained with the KPNO 2.1-m on September
26, 2002 when the object was in an faint-state {V ~ 15.5). The J and H bands show
strong elhpsoidal variations, while the i^-band morphology is the result of a combination
of ellipsoidal and cyclotron emission. The lightcurves were phased using the Kafka et al.
(2005b) ephemeris. The overplotted lines are ellipsoidal models for i = 50°.

Figure 2. Long-term RoboScope lightcurves of AM Her complimented by AAVSO
data. The "P" denotes the times of our our SQHD and NIC-FPS/NMSU 1-m photometry,
while the " S" indicates our spectroscopic observations, top: V^-band photometry following
AM Her from 1991 through 2006. bottom-left: Zoom-in of the year surrounding our SQHD
photometric measurements, bottom-right: Zoom-in of the year surrounding our IRTF spec-
troscopy.

Figure 3. BVRIJHK photometry of AM Her obtained with the APO 3.5-m/NMSU
1-m on May 20, 2005 when the object was in a faint-state (V = 15.3) similar to the KPNO
lightcurves (see Fig. 1). An identical ellipsoidal model to that used for the KPNO pho-
tometry is overplotted here, matching well from R to K, and although a small flare event
is evident in R and / during the second cycle of observation almost no cyclotron emission
should be present in these bands. Humps reminiscent of cyclotron emission reappear in the
V and especially B bands suggesting that a higher field is active on AM Her. (a) The optical
bands (b) the NIR bands

Figure 4. (a)IRTF/SPEX phase-resolved spectra of AM Her plotted (black) as a
stacked series, with a constant flux increment of XFx = l.lOx 10~^^ erg s~^ cm~^ and covering
the orbital phases for which the cyclotron emission region is in view. At each phase, the
1.22 /xm flux is normalized to the J-band lightcurve ensuring proper flux calibration with
the narrow 0.3" slit and a dim-phase spectrum (0 = 0.42) was subtracted. Because of the
variability of the secondary's spectral type over the orbit, the underlying continuum as well
as intrinsic water features changed over the orbit. Remnant intrinsic water features, however,
are still apparent at ~ 1.35 and 1.8 /xm. No cyclotron emission was observed from = 0.27
to = 0.74.

Figure 5. JHK photometry of ST LMi obtained with the KPNO 2.1-m/SQIID on the
April 9, 2003 high-state. The dim-phase lasts from = 0.00 to 0.55, while from = 0.60 to
0.95 the bright-phase is observed. The plotted lines are ellipsoidal models for i = 55°.

Figure 6. Long-term lightcurves of ST LMi. top: V-band photometry following ST
LMi from late 1991 to early 2006. bottom: Zoom-in of the year surrounding our IRTF/SPEX
spectroscopy. The "P" denotes the epoch of our photometry, while "S" marks our spectro-
scopic data during the 2005 extreme low-state and the 2006 low-state, respectively.

-19-

Figure 7. IRTF data of ST LMi in an extreme low-state (2005). An M6 template
spectrum is plotted during the dim-phase (0 = 0.02) confirming the spectral classification.

Figure 8. (a) Phase- Resolved SE-subtracted spectroscopy of ST LMi, taken in February
2006 during a low-state. The IRTF/SPEX data are plotted (black) as a stacked series - a
constant increment of XFx = 1.2x10^^^ erg s^^ cm~^ is added to each spectrum to offset it
from the spectrum below it. The SE spectrum subtracted from each phase was a median of
three dim-phase spectra. The best fit cyclotron model for each of the bright-phase spectra
are shown in green, (b) the same, but for the February 2005 extreme low-state.

Figure 9. (a) The derived value of the viewing angle (0) for AM Her is plotted vs.
the orbital phase (0) in black, with the best fit geometry {i = 50°, j3 = 85°) overlaid in red.
The blue shading indicates phases for which AM Her is self-eclipsed, (b) the same, but for
ST LMi and with a geometrical model of {i = 55°, (3 = 128°).

-20-

0.5

1.5

11.4

11.6

11.8

12

0.5

1.5

11.6

0.5

1
Orbital Phase

1.5

-21-

.1. *■

-

13

" . .;■' :*'". ■■■■■'■ : :-.-.'■.■■ -!|.- > Pv|

FS \

14

" ". i-" ■ '': "■ ■'■ '" '?;■.-■ ;-^'' ■ 1 :'' -

-

)
i

15

- -■ ■':■■■ ■ ; : ' r > ■:;■;;• ^ .

; -'y/^.- '■ * W' f ■ '^ '4^; Ki

t||^

1 1 1 1 1 1

1 1

13

14

15

4200

2000

4000

JD - 2448200

:

D

:

* _

»

P 1\- s

-

-

-

\. --J ^-T -

-».*' -"

il %-»«v %'■**..;.»

^.- -. ■ _ ■ „■ .

r.:(V:

1

1 1 1 1 1 1 1 1 1 1 ~

~l 1 1

1 1 1 1 1 1 1 1 1 1 1 1 ~

4300 4400
JD - 2448200

5300 5350 5400 5450
JD - 2448200

-22-

15.4
15.5

g'

CQ

15.7

15.2
15.3
15.4
15.5

h

'Hi

d^i

iiji

^.gi

15.6 :_____Miyi|iii^

h

.1^5

Jjji^

'^Mlmih^ — :

0.5

1

Phase

1.5

0.5

1
Phase

1.5

Phase

13.6
13.8

14

0.5

1
Phase

1.5

-23-

12.1
12.2
12.3
12.4

12.5 '*

12.6

1 1 1 1 1

1

1 1 1 1 1

'^

^i

1 1 1 1 1 1 1

•/ \ •

J -^

I

••

•

w

•
•
•

7 *\' 'v

1

1

1

0.5

1.5

—I 1 r-

11.2

11.4

11.6

11.8

12

0.5

1.5

10.8

Orbital Phase

-24-

AM Her <t>= 0.00 - 1.00

10-10 .

8x10-" -

i- 6x10-" I- _,/<A^^-^,

^-^''>^,v^^-W;^>^''* ----^^ ^ ^ g

00

4x10-"

8x10-"

^/VAjMk,

n = 4

-.i>^
v

0.97

0.92

0.86

0.80

0.75

0.26

0.20

0.09

1.5
Wavelength (/xm)

-25-

12.5

13.5

12.5

«

13.5

Phase

-26-

14 -

15

^ 16

17

1 -

-

p

S

s

-

1 *

•»

■

■

■r

1'

••
'i.

■B

* V

'■ .

-

-

■;«■

■•:%

f>*\

^•:n-

•■-V

•

/

'•

k

■■■ 1

1

-

-

1000

2000 3000
JD - 2448600

4000

5000

Of)

a

-

14

-

S

s

-

15

-

-

16

-

«

-

17

-

i

-

-

«• .

.1

^

•

1

-

18

^

1 1

1

1 1

1 1

!

1

4800

5000
JD - 2448600

5200

-27-

8xl0-'2

1 1.3

1.4 1.6 1.!

Wavelength (^m)

ff>iiiiii

i_ 0.02

2.2 2.4

-28-

ST LMi 0= 0.00 - 1.00

10-"

8x10-"

' Mil

6xl0-'«

BO

0)

4x10-"

2xl0-i«

rH.1. J^^^.

■f «AuWA/A(V^^ ^ ^ ./^^'^'^

ff

n = 6 n = 5 n = 4

0.99

0.91

0.80

0.73

0.68

0.57

0.50

0.38

0.14

1 1.2 1.4 1.6 1.8 Z 2.2 2.4

Wavelength (/im)

-29-

ST LMi 2005 Extreme Low-State

2x10-"

4xl0-'2

Ml

'lpV\f jMJ^I

vyw'K

nW

Y^vVIa

V^^M/WA

1 i.a

1.4 1.6 I.a

Wavelength (/.irri)

V^

0.91

0.83

0.72

0.60

_0.48

0.37

0.25

0.14

0.02

2. a 2,4

-30-

0.5 1 1.5

Orbital Phase (0)