A Map of the Ionized Component of the Intermediate Velocity Cloud Complex K

o
o

m

>$

A Map of the Ionized Component
of the Intermediate Velocity Cloud Complex K

L. M. Haffner, R. J. Reynolds

Department of Astronomy, University of Wisconsin, Madison, WI 53706

haf f nerOastro . wise . edu, reynoldsOastro . wise . edu

and

S. L. Tufte

Department of Physics, Lewis and Clark College, Portland, OR 97219

tuf teOlclarke . edu

>

O : ABSTRACT

(N

o

The Wisconsin H-Alpha Mapper (WHAM) Northern Sky Survey is reveahng that
many intermediate- velocity (|vlsr| ^ 100 km s~^) neutral clouds and complexes have
an associated ionized component. We present the first map of the Ha emission from an
intermediate- or high-velocity structure: Complex K. This large, high-latitude feature
stretches from £ Ri 10° to 70°, b ^ +30° to +70° and peaks in velocity over t^lsr ~ —60

^ • to —80 km s ^. The neutral and ionized gas generally trace each other quite well in

^ • the complex, but the detailed structure is not identical. In particular, the Ha emission

peaks in brightness at slightly higher Galactic longitudes than corresponding 21 cm

features. The ionized gas has a peak Ha intensity of 0.5 Rayleighs, corresponding to
an emission measure of 1.1 cm~^ pc. Structures in the complex are traced by WHAM
C^ I down to about 0.1 Rayleighs (0.2 cm~^ pc). Typical line widths of the Ha emission

are ~ 30 km s~^, limiting temperatures in the ionized gas to < 20,000 K. If radiation
is the primary ionizing mechanism, the Lyman continuum flux required to sustain the
most strongly emitting ionized regions is 1.2 x 10^ photons cm^^ s^^. There appears
to be no local, stellar source capable of maintaining the ionization of the gas; however,
the required ionizing flux is consistent with current models of the escape of Lyman
continuum radiation from OB stars in the disk and of ionizing radiation produced by
cooling supernova remnants.

Subject headings: Galaxy: halo — ISM: clouds — ISM: individual (IVC-K) — ISM:
structure — ISM: atoms — HII regions

1. Introduction

Optical emission lines are a new tool for the exploration of high- and intermediate- velocity
clouds (HVCs and IVCs), traditionally discovered and studied in the 21 cm line of H I or through
absorption line observations. The Ha intensity from such structures is typically a few tenths of
a Rayleigh (1 R= 10^/47r photons cm^^ s^^ sr^"*^ = 2.4 x 10^^ ergs cm^^ s~^ sr""*^) and has only
recently been accessible through modern detectors. With such faint emission, previous studies
of emission lines such as Ha, [S II], and [N II] from HVCs and IVCs have been limited to single
pointed observations, although they have gleaned some of the first information about the ionization,
temperature, and metallicity of these gas structures (Weiner & Williams 1996; Tufte, Reynolds, &
Haffner 1998; Bland-Hawthorn et al. 1998; Wakker et al. 1999). Since most of the known HVCs and
IVCs are at high latitudes and many are distant (d > 500 pc) structures, studies of their ionized
component reveal information not only about the clouds themselves but also about the environment
of the Galactic halo and the escape of radiation from the Galactic plane.

With high sensitivity and velocity resolution, the recently completed Wisconsin H- Alpha Map-
per^ (WHAM) Northern Sky Survey is yielding an enormous amount of new information about the
ionized nature of IVCs. Wakker (2001) has recently separated out the large {A£ ~ 40°, A6 ~ 40°)
H I intermediate- velocity Complex K (previously classified with Complex C) centered near i = 50°,
b = +50°, i^LSR ~ —60 to —80 km s^^. Here we present the first map of the ionized component of
Complex K.

2. Observations

The WHAM Northern Sky Survey provides the first velocity-resolved survey of Ha emission
from the Galaxy. Each of the 37,300 pointings in the northern survey captures a spectrum over
^LSR ~ —100 to +100 km s~^ from a one-degree patch of sky. Since WHAM provides 12 km s~^
spectral resolution, we are able to cleanly remove the bright and variable geocoronal Ha emission
(/ rv^ 2 — 13 R) and numerous fainter (/ ~ 0.1 — 0.3 R) atmospheric lines (N. R. Hansen et al., in
preparation), allowing detections of faint Galactic emission below 0.1 R. The survey achieves nearly
full coverage of the sky with beam spacings of Aft = 0?85 and Ai = 0?98/cos6 above 6 > —30°.
More information about the instrument and survey strategy can be found in Tufte (1997), Haffner
(1999), and L. M. Haffner et al. (in preparation).

^More information about WHAM can be found at http://www.astro.wisc.edu/whani/.

3. Results

Figure 1 shows an image of the Ha emission from WHAM and contours of 21 cm emission from
the Leiden-Dwingeloo H I survey (Hartmann & Burton 1997) toward Complex K. Note that the
image and contours reflect a strict integration limit between —95 and —50 km s^^. As a result, the
colorbar legend underestimates the total Ha line intensity by about 20% {e.g. Figure 2, see below).
Since the distance to Complex K is still relatively uncertain, we have not applied an extinction
correction to the data presented here. However, since the total Nhi (—100 < vlsr < +100) over
the face of the cloud only varies from about 1 to 4 x 10^'^ cm"^, such a correction to the observed
Ha flux would be at most about 25% if all the neutral material is between us and Complex K.

The ionized emission is distributed diffusely over the neutral extent of the complex; it does
not appear to arise from numerous point sources within the complex. The integrated Ha intensity
in the region displayed in Figure 1 ranges from 0.5 R down to a conservative detection limit of
0.1 R or EM ~ 1.1 to 0.2 cm"^ pc (1 R R^ 2.25 cm"^ pc for T = 8000 K). The extent of the
Ha emission in the main part of the complex {i = 30° to 65°, b = +30° to +70°) matches the
Nff = 7 X 10^^ cm~^ contour quite closely; however, the peaks in each emission component do
not correspond as well. A detailed quantitative analysis of the Ha and H I intensities within each
one-degree WHAM pointing reveals that there is little correlation between the integrated intensities
of the IVC components. The linear correlation coefficient for pointings within the main part of
the complex is 0.4. Although the spatial distribution and velocity profiles correspond extremely
well, a simple relationship relating the Ha intensity to the H I column does not exist. In fact, for
many of the largest concentrations of H I emission {e.g., i = 48°, b = +36°; i = 54°, b = +52°;
i = 45°, b = +44°), the brighter Ha emission is consistently at higher longitudes than the regions
of brightest 21 cm emission. We created Ha and H I latitude profiles from the images to quantify
the extent of this offset. By examining the two latitude profiles simultaneously and by computing
correlation coefficients between the profiles in their natural and offset positions, we find that the
Ha emission is consistently shifted or extended to higher Galactic longitudes in nearly all latitude
slices by a 2 to 4 degrees.

Figures 2 and 3 present representative Ha spectra (top panels) from Complex K with the one-
degree WHAM beam centered at £ = 57?02, b = +49?22 and £ = 48?03, b = +36?49, respectively.
Corresponding H I 21 cm spectra centered at £ = 57?0, b = +49?0 and £ = 48?0, b = +36?5
from the Leiden-Dwingeloo H I survey are displayed in the bottom panel of each figure. These
two sightlines sample some of the brighter emission from Complex K. The dashed vertical lines
denote the integration range used to construct the Ha image and 21 cm contours in Figure 1.
Since hydrogen has a thermal width of ~ 22 km s~^ at T = 10^ K, we have chosen a restricted
integration range to avoid contamination in the image from the local emission component centered
near wlsr = km s~^. As noted above, the values in the color legend of Figure 1 should be
increased by about 20% to estimate the total intensity of a typical Complex K emission line.

The solid lines in the top panels of Figures 2 and 3 represent a Gaussian component fit con-

volved with the WHAM instrument profile (approximately 12 km s^^ FWHM). We also performed
a multi-component fit on the H I spectrum. Only the IVC component is shown clearly in the
bottom panels of Figures 2 and 3 since little effort was spent trying to accurately model the bright,
local 21 cm emission. Figure 3 clearly shows at least two components in the intermediate- velocity
neutral gas. A single Gaussian fit to the Ha emission in this direction results in a large (> 40
km s^^) width, atypical of diffuse gas; a free-fit two Gaussian profile is not well defined due to
the intrinsic ionized gas line widths. Instead, we have chosen to fix the mean velocity of the two
components in the Ha fit to be the same as those clearly defined in the H I profile. The best pa-
rameters for our fits are given in Table 1. Channel maps and profile analyses show some variation
in the mean velocity of the Ha and H I emission, with values that fall roughly between —50 and
—75 km s^^. There appears to be no systematic velocity gradients in the complex, although the
higher velocity H I component at —75 km s^^ is limited to the two regions of the complex centered
at £ = 48°, b = -F37° and i = 38°, b = -h62°.

4. Discussion

As reviewed by Wakker (2001), little absorption line data directly toward Complex K exists
at this time, and thus we have little information about its distance, only that it lies between about
0.3 and 7.7 kpc. There is a firm upper limit since the cloud has been seen in absorption (de Boer
& Savage 1983; Bates et al. 1995; Shaw et al. 1996) toward M 13 (£ = 59°, b = +41°). Carretta
et al. (2000) have recently calculated the distance modulus of M 13 to be 14.44 mag, which sets
an upper limit of 7.7 kpc on the distance to that portion of Complex K toward M 13. A lower
limit to the distance comes from the work of Bates et al. (1995) who included three local stars in
their absorption line study of stars in M 13. Two of their targets, HD 151749 and HD 150462,
have reliably measured parallaxes (HIPPARCOS, Ferryman et al. 1997), which place them at 150
and 175 pc, respectively. Their third target, HD 149802, has an estimated distance of about 270
pc given that it and HD 150462 have both been classified as AO but differ in V brightnesses by 0.9
mag. They did not detect Complex K absorption in either Ca II or Na I toward any of these stars.

If the entire complex is at a common distance then its angular extent, from b = -|-30° to
b = -|-70°, corresponds to a length of 0.68d kpc and a maximum vertical extent above the midplane
of 0.94(i kpc, where d is the distance to the complex in units of 1 kpc. If instead the complex is a
column of gas perpendicular to the Galactic plane at a constant radial distance (distance along the
Galactic plane) from the sun, its length is 2.2r kpc, where r is the radial distance of the complex
from the sun in units of 1 kpc. Gas participating in Galactic rotation toward the center of the
complex near £ = 55°, b = +45° would be expected to have a radial velocity of wlsr = +10 km s^^
at 1 kpc up to a maximum (at the tangent point) of t;LSR = +30 km s^"*^ at 6.5 kpc (Clemens 1985).
As a result, the deviation velocity (Wakker 1991) for Complex K is approximately —100 km s^^
from that expected by circular rotation.

With few observations of the ionized component of IVCs and HVCs, it is difficult to dis-

criminate among possible candidates for the source of ionization. Since the clouds are moving at
anomalous velocities, processes such as ram-pressure heating (Weiner & Williams 1996) and radia-
tion produced by a cloud-halo shocked interface have been explored (Tufte et al. 1998) in addition
to photoionization by an ambient Lyman continuum flux (Tufte et al. 1998). In the case of Complex
K, the relatively diffuse Ha emission over the extent of the neutral gas suggests the latter ionization
mechanism. However, at its lower distance limit (300 pc), the vertical extent of the complex is only
z = 150 to z = 300 pc, fully within the WIM layer, and its anomalous velocity would likely cause
an interaction with the ambient medium. But unless Complex K is falling nearly directly towards
us, we find little evidence for an Ha enhancement in the morphology due to cloud motion. In
support of photoionization by an external or ambient flux, note that while the total H I column
densities presented in Table 1 for the IVC components vary by a factor eight, the Ha intensities
are relatively constant. For the cloud as a whole, we find that the neutral gas displays much higher
intensity contrast than the ionized gas in Complex K (see also Figures 2 and 3). Furthermore, as
noted above in §3, the point-to-point correspondence of the total IVC intensity of the two emission
lines is poor. These observations are consistent with photoionizing the skin of a neutral cloud. In
this case, the cloud becomes self-shielding from ionizing radiation and Nm is independent from the
Ha intensity, which becomes mostly a function of incident Lyman continuum fiux and geometry.

If a cloud is ionized by an external photoionizing source, the incident Lyman continuum flux
on the cloud surface can be estimated (see Tufte et al. 1998). Ignoring the complexity of geometry
but applying the maximum possible extinction correction (25%; see §3) to the sample ionized IVC
component in Table 1, we find the incident flux of ionizing photons is (/> ~ 1.2 x 10® photons
cm~^ s^^ for the brightest Ha regions in Complex K. With the small lower limit on its distance,
there is some chance that Complex K is a relatively local feature that could be ionized by nearby
sources. A search within 16° of £ = 55°, h = +45° (the rough center of the brightest regions of
the complex) reveals no early spectral type objects with sufficient luminosity to ionize the region,
even at its minimum distance of ~ 300 pc. As noted above, the ionized regions with higher Ha
intensities appear to be offset a few degrees from the larger column density neutral regions. To test
whether the ionization of the complex could be sustained by a brighter, more distant halo source,
we have explored a large region at higher Galactic longitudes than Complex K (80° < i < 180°;
+30° < b < +80°) for possible ionizing sources. In this case, the only potential source appears to be
the 09.5 V star HD 93521 {£ = 183°, b = +62°, d = 1.7 kpc; Irvine 1989). Given the direction and
range of distances possible for Complex K, it is closest to the star when placed about 700 pc away
{z « 500 pc); here they are separated by about 1.5 kpc. Using a value of logQo = 48.38, where
Qq is the number of hydrogen ionizing photons per second emitted from an 09.5 V star (Vacca,
Garmany, & Shull 1996) and taking into account only the geometrical dilution of the radiation
(i.e. no intervening gas or dust absorption), we find that the ionizing fiux at Complex K from HD
93521 is (p Ri 8.4 x 10^ photons cm^^ s^^, far short of that needed to maintain the ionization of the
complex.

More likely, a scenario similar to those used to explain the ionization of the Warm Ionized

Medium (Miller k Cox 1993; Dove & Shull 1994; Dove, Shull, k Ferrara 2000; Slavin, McKee, k
Hollenbach 2000) is also providing the ionizing flux for Complex K. Dove et al. (2000) cite upward
halo ionizing fluxes for their various models of the propagation of OB association radiation through
the surrounding superbubbles. Their range of calculated values, (p ~ 1-2 x 10^ photons cm~^
s~^, agree well with what is required to maintain the ionization of Complex K. Bland-Hawthorn
k Maloney (1999) have produced a model of the propagation of the ionizing flux in the Galactic
halo and the resulting Ha emission from halo clouds. We find that our observed Hq intensity is
consistent with that expected from their model in the direction of Complex K for distances < 8
kpc from the sun (0.15-0.30 R; J. Bland-Hawthorn 2000, private communication). Cooling hot gas
generated in supernova remnants has also been shown by Slavin et al. (2000) to provide a mean
value of = 5.4 x 10^ photons cm^^ s^^ to the halo ionizing flux, nearly half that required for the
brightest emitting regions in the complex and sufficient alone to sustain the faintest.

With WHAM, we have finally begun to explore the structure of the ionized component of
IVCs and HVCs with the Ha line in the same manner as has been done for the neutral gas with
the 21 cm line. Furthermore, maps in other optical emission lines such as [S II] and [N II] will
provide additional information about the temperature, ionization, and metallicity of the ionized
gas (Haffner et al. 1999; Wakker et al. 1999). These future observations have the potential to
provide insight not only about the nature of these high latitude structures but also about the
ambient Galactic halo gas and radiation field.

We would like to thank Bart Wakker for the first H I-Hq comparisons in this region. Joss
Bland-Hawthorn for providing model results along the Complex K sight line, and Bob Benjamin for
continuing halo consultation and Galactic rotation code. Source searches were conducted through
the SIMBAD database operated at CDS, Strasbourg, France. WHAM is supported by the National
Science Foundation through grant AST 96-19424.

REFERENCES

Bates, B., Shaw, C. R., Kemp, S. N., Keenan, F. R, k Davies, R. D. 1995, ApJ, 444, 672

Bland-Hawthorn, J. k Maloney, P. R. 1999, ApJ, 510, L33

Bland-Hawthorn, J., Veilleux, S., Cecil, G. N., Putman, M. E., Gibson, B. K., k Maloney, P. R.
1998, MNRAS, 299, 611

Carretta, E., Gratton, R. G., Clementini, G., k Fusi Pecci, F. 2000, ApJ, 533, 215

Clemens, D. P. 1985, ApJ, 295, 422

de Boer, K. S. k Savage, B. D. 1983, ApJ, 265, 210

Dove, J. B. k Shull, J. M. 1994, ApJ, 430, 222

-7-

Dove, J. B., Shull, J. M., k Ferrara, A. 2000, ApJ, 531, 846

Haffner, L. M. 1999, PhD thesis. University of Wisconsin-Madison

Haffner, L. M., Reynolds, R. J., & Tufte, S. L. 1999, ApJ, 523, 223

Hartmann, D. & Burton, W. B. 1997, Atlas of Galactic Neutral Hydrogen (Cambridge; New York:
Cambridge University Press, ISBN 0521471117)

Irvine, N. J. 1989, ApJ, 337, L33

Miller, W.W., I. & Cox, D. P. 1993, ApJ, 417, 579

Perryman, M. A. C, Lindegren, L., Kovalevsky, J., Hoeg, E., Bastian, U., Bernacca, P. L., Creze,
M., Donati, F., Grenon, M., van Leeuwen, F., van der Marel, H., Mignard, F., Murray,
C. A., Le Poole, R. S., Schrijver, H., Turon, C, Arenou, F., Froeschle, M., & Petersen, C. S.
1997, A&A, 323, L49

Shaw, C. R., Bates, B., Kemp, S. N., Keenan, F. P., Davies, R. D., & Roger, R. S. 1996, ApJ, 473,
849

Slavin, J. D., McKee, C. F., & Hollenbach, D. J. 2000, ApJ, 541, 218

Tufte, S. L. 1997, PhD thesis. University of Wisconsin-Madison

Tufte, S. L., Reynolds, R. J., & Haffner, L. M. 1998, ApJ, 504, 773

Vacca, W. D., Garmany, C. D., & Shull, J. M. 1996, ApJ, 460, 914

Wakker, B. P. 1991, A&A, 250, 499

Wakker, B. P. 2001, ApJS, in press

Wakker, B. P., Howk, J. C, Savage, B. D., van Woerden, H., Tufte, S. L., Schwarz, U. J., Benjamin,
R., Reynolds, R. J., Peletier, R. F., & Kalberla, P. M. W. 1999, Nature, 402, 388

Weiner, B. J. & Williams, T. B. 1996, AJ, 111, 1156

This preprint was prepared with the AAS lATgX macros v5.0.

Table 1. Spectral Fit Parameters

Direction

Component

Mean

FWHM

Intensity

{i,b)

[km s-i]

[km s-i]

(57?02,+49?22)

Ha Local

-4.1+0.8

27.0+2.5

0.50+0.03

R

(57?02,+49?22)

Halve

-64.0+0.8

29.0+2.3

0.48+0.02

R

(57?00,+49?00)

HIIVC

-60.4+1.0

22.0+2.5

8.9+0.9

10^^ cm~2

(48?03, +36? 49)

Ha Local

-4.5+0.5

23.9+1.4

0.78+0.02

R

(48?03, +36? 49)

Halve

-47^

22.2+10.0

0.15+0.03

R

(48?03, +36? 49)

Halve

-73^

24.7+4.6

0.27+0.03

R

(48?00,+36?50)

HIIVC

-46.8+1.1

19.1+2.2

1.9+0.3

10^^ cm-2

(48?00,+36?50)

HIIVC

-72.8+0.6

25.7+1.1

5.3+0.3

1019 cm-2

^For the fit in this direction, the mean velocities of the ionized IVC components
have been fixed to be the same as those derived from the fit to the neutral gas. See
S3 for full details.

Fig. 1. — The ionized and neutral components of Complex K. Integrated Ha emission from wlsr =
—95 to —50 kms is presented as the color image. H I column densities (integrated over the same
velocity range) of 7, 14, and 21 x 10^^ cm~^ from Hartmann & Burton (1997) are displayed as light
blue contours.

Fig. 2. — {top) WHAM Ha spectrum of the one-degree portion of the sky centered at i = 57?02, b =
+49?22. WHAM data is plotted as squares with one sigma error bars. A 3.7 R geocoronal line
centered near vlsr = +30 km s~^ has been removed from the spectrum. A simple two component fit
to the Galactic emission is shown as the solid line (see text), (bottom) H I 21 cm emission spectrum
toward i = 57?0, b = +49?0 from the Leiden-Dwingeloo H I survey. The local emission peaks off
the displayed scale near 8.2 K. The solid line denotes the IVC portion of a multi-component fit (see
text). In both panels, the dashed lines denote the integration range used for the image in Figure 1.

Fig. 3. — Same as Figure 2 except toward {top) I = 48?03, b = -|-36?49 for Ha and (bottom)
i = 48?0, b = -|-36?5 for H I. The local H I emission peaks off the displayed scale near 15.8 K.

10

80

70

60

50

40

30

20

10

ntensity [Rayleighs]/45 km s

-1

0.00 0.05 0.10 0.15 0.20 0.25 > 0.30

11

0.020

00

-50

Velocity [km s"

50

12

CO

a

X

0.025

0.020

0.015

0.010

0.005

0.000

■0.005 -

^ 1.0

E
o

CSl

0.5

0.0

J I I I I I I I I I I I I I I I I I I I I =kl I L

00

-50

Velocity [km s~ ]

50