Peeking through the picket fence: What astrophysical surprises may be present in the 100–1200 Å region?

Adv. Space Res. Vol.11, No.11. pp. (1 1)5—(l1)14. 1991 Printed in Great Britain. All rights reserved.

0273—1177/91 $0.00 +.50 Copyright (1) 1991 COSPAR

PEEKING THROUGH THE PICKET FENCE: WHAT ASTROPHYSICAL SURPRISES MAY BE PRESENT IN THE 100-1200 A REGION? Jeffrey L. Linsky and Donald G. Luttermoser Joint Institutefor Laboratory, Astrophysics, National Institute of Standards and Technology, University of Colorado, Boulder, CO 80309-0440, U.S.A.

ABSTRACT In anticipation of more sensitive EUV and FUV spectroscopic instruments, we simulate spectra, including interstellar absorption, of solar-like, RS CVn, and flare stars as folded through the instrument parameters of the EUVE, Lyman/FUSE Phase A, and a desirable next-generation spectrometer. We find that even the relatively insensitive EUVE spectrometer will be able to detect sufficient spectral lines from many active binary and dMe stars to determine their coronal emission measure distributions. Tile Lyman/FUSE or nextgeneration spectrometers are needed to study solar-type stars or flaring stars with high time resolution. The high throughput and effective area of a next-generation spectrometer is needed for Doppler imaging studies, stellar wind and downfiow measurements, and high time and spectral resolution of stellar flares. INTRODUCTION This meeting on “EUV and FUV Astrophysics of Stars” occurs at the threshold of a very rapid increase in the data anticipated in these new spectral regions. • Until now spectrometers on Voyager and EXOSAT have provided only limited glimpses of a few hot white dwarfs and other hot stars in the EUV (100—912 A). We look forward to spectra of many stars of diverse types from the HUT instrument On Astro—1, ORFEUS on ASTROSPAS, the EUVE spectrometer, and the EUV channel on Lyman/FUSE. • Until now sounding rockets have provided a few low-resolution images of the EUV sky. We look forward to the feast of data expected from the the Wide Field Camera on ROSAT and the all-sky and deep surveys with EUVE. • Until now, Copernicus has obtained FUV (912—1200 A) spectra of only a limited number of stars located within 1 kpc, which have provided valuable information on the properties of interstellar gas on the lines of sight towards these stars. We look forward to high-resolution spectra of nearly all types of stars in the Galaxy and the brighter sources in many other galaxies, which will be provided by Lyman/FUSE. We introduce the session by summarizing the important features in solar FUV and EUV spectra and then present simulated spectra of solar-type and active late-type stars that should be detected by spectrometers planned for EUVE, Lyman/FUSE, and by a hypothetical next-generation spectrometer with enhanced throughput and spectral resolution. These simulations will allow us to extrapolate our present understanding of stellar chromospheres, coronae and flares into two nearly unexplored spectral regions. Missions being planned to observe celestial sources in these new spectral regions should confirm or refute our present paradigms, which are based on solar analogy, and more importantly should discover new phenomena and new types of stars. We suggest how one might proceed to make such discoveries.

‘Staff Member, Quantum Physics Division, National Institute of Standards and Technology

(11)5

(11)6

J.L. Linsky and D.G. Luttermoser

SPECTRA OF SOLAR-TYPE STARS The solar FUV spectrum provides a convenient first approximation to spectra of late-type dwarf stars, spectroscopic binaries, and perhaps a broader range of stellar types. Table 1 lists the brightest emission lines for a solar active region /1/ in order of decreasing intensity with lines grouped according to degree of ionization. The line intensities in active regions are typically ten times larger than for quiet regions, and the emission lines are much brighter than the underlying continuum for both active and quiet regions. While La dominates the solar spectrum in this region, Lfi is comparable in flux to the lines of Silil and CIII, formed at roughly 60,000 K, and the lines of OVI, formed at 300,000 K. Higher members of the Lyman series are much fainter. In stellar spectra interstellar absorption will remove most or all of the flux in La, so that the SiTu, CIII, and OVI lines will likely dominate the observed spectra. The solar active region spectrum should be a good approximation to the spectra of dMe and RS CVn stars as these stars typically have surface fluxes 10-30 times that of the quiet Sun in the CIV lines, which are formed at 100,000 K /2,3/.

~7X~ 1216 1206 977 1025 1176 1032 1037 972 1238 1085 1242 991 949 1200 1194 937 1128 923 930 1122 1190 1157 933 1010

Table 1. Strongest Lines in Solar Active Regions FUV (912—1250 A) ________ EUV (148—350 A) _______ HI I-Il III-IV V-VT Intensity X~A) Low Medium High Flux La 61,800 304 Hell 11,490 SiIII 630 284 FeXV 2960 CIII 410 335 FeXVI 2340 L/3 314 195 FeXII 880 CIII 288 256 Hell 820 OVI 230 180 FeXI 700 OVI 151 211 FeXIV 680 Ly 94 188 FeXI 650 NV 83 193 FeXII 630 NIl 77 171 FeIX 580 NV 57 274 FeXIV 550 NIlI 52 174 FeX 520 L8 46 334 CaVI 470 NI 46 264 FeXIV 410 Sill 30 177 FeX 360 Le 25 347 SiX 350 SiIV 25 252 FeXIV 270 NIV 25 220 FeXIV 210 LC 20 SiIV 19 Sill 18 CI 13 SVI 12 CII 10

944

2SVI s~ ster’. 10

______________________

______

Intensity units: 1012 ph cm

Flux units: 106 ph cm2 ~

Emission lines formed at higher temperatures dominate the solar spectrum below the 912 A photoionization edge of hydrogen, but this same opacity source in the interstellar medium and Hel absorption below 504 A should completely absorb the spectra of nearly all stars in the 350—912 A range (see below). Table 1 lists the solar emission line fluxes in the 148—350 A region observed by OSO-5 /4/ for the very active Sun (sunspot number 194). We note that the bright lines of FeIX-XI are formed at logT = 5.9—6.1 K and the lines of FeXIVXVI are formed at logT = 6.3—6.5 K. The 90—140 A region, which is the shortest wavelength region that will be observed by the EUVE and Lyman/FUSE spectrometers, is dominated by 2322~k — 2s12p~1 transitions of FeXVIII—XXIII, which are formed over the temperature range logT = 6.9—7.5 K /5/. The 100—350 A region also contains a3 rich spectrum of line ratios (e.g. for FeXII—XV) that are sensitive to densities in the range /6,7/. 108 — 1012 cm

Peeking through the Picket Fence 3

a

5 = 1.0 cm°

~

0.2

n11

0.2

~.

0 ~ 100

160

200

250

300

‘~

0 100

~

350

..-

160

‘

200

250

300

350

N~(cm2)

::~~

i::~ 0.4

0.1 cm

SOpc

n~= .01 cm3

,

(11)7

0.4

100pc~~..

10•* ~

0.2

‘

I

0.2

kpc

ioa

—

o

100

~~~—_i

150

. I

200 250 300 Wavsl.ngth (A)

350

~ 100

150

200 250 300 Wov.Isng$h (A)

350

Figure 1: Transmission of the interstellar medium as a function of wavelength and distance for different values of the neutral hydrogen density. The lower right plot shows the transmission for different values of the hydrogen colunm density. THE LOCAL INTERSTELLAR MEDIUM In the current picture of the local interstellar medium (LISM) /8,9/ the Sun is located in an irregularly-shaped region of warm and mostly neutral gas with 7iH < 0.1 cm3. This region may only extend 10 Pc towards the galactic center, where it comes up against a high-density cloud. The LISM appears to contain a number of velocity components and its structure perpendicular to the galactic plane is not well known at this time. As shown in Figure 1, the EUV horizon depends critically on the wavelength and column density of neutral hydrogen towards a star. If the LISM were a uniform ~H = 0.1 cm3, then we might expect to observe stars beyond 50 Pc only for A < 150 A. The observed low column densities towards stars in the third quadrant (galactic longitudes 200°— 270°)to at least 200 pc indicates that hot, low density gas (flH 0.01 cm—3) must extend beyond the local cloud in this direction. In our simulated spectra we assume that n 3 along 1, = 0.1 cm the line of sight towards each star, but this is only a rough estimate. SIMULATED STELLAR SPECTRA

\Ve compute simulated spectra folded through the response functions of three instruments — the EUVE spectrometer /10/, the proposed Lyman/FUSE high-throughput EUV channel /10/, and a hypothetical highresolution spectrometer (EUVHRS). The spectral resolutions and effective areas are shown in Figure 2. We also compute simulated FUV spectra for the prime spectrometer on Lyman/FUSE, assuming a resolution = 30,000 and effective area of 50 cm2 for 900 A A 1025 A and 100 cm2 for 1025 A ~ A 1250 A. We first compute average quiet Sun spectra (Fig. 3) folded through the response functions of the three spectrometers (Fig 2). For each spectrometer we compute the intrinsic spectrum (ignoring the very bright Hell 304 A line) and the spectrum including only interstellar hydrogen absorption with the indicated neutral hydrogen column density computed from ~H = 0.1 cm3 and the indicated distance. The distances to the sources are obtained by requiring that the total count rate in the 100—350 A band (excluding the Hell 304 A line) exceed 1000 counts in 50,000 seconds. Our detection criterion of 1000 counts was set to permit the detection of some 30 lines, which should be sufficient to derive the emission measure distribution with temperature in tile stellar corona. By this criterion the detection horizon for a solar-type star is only 2.4 pc for the EUVE spectrometer, but 18 pc for the EUVHRS. Most of the emission lines in Table 1 are detected in these simulations, except that interstellar absorption precludes detection of the FeXVI 335 A line for a star like the quiet Sun located at 18 pc. Simulated spectra of the quiet and active Sun (Fig. 4) indicate that the detection horizons using tile Lyman/FUSE Prime spectrometer are 33 and 100 pc, respectively.

(11)8

IL. Linsky and D.G. Luttermoser

Spacecraft Comparison

2500

2000

.

-

- —

£UVHRS

C 0

1500

1000

500

Spacecraft Comparison

120

EUVHRS 100

C~ E U

-

-

-

80

60 U

~

LW

40

~

0 100

150

I..

200

I

250

I

300

350

Wcvs~•ngth(A)

Figure 2: Comparison of different EUV spectrometers. (top) Spectral resolution of the spectrometer on the EUVE satellite, the high-throughput channel of the spectrometer described in the Lyman/FUSE Phase A Report /10/, and a hypothetical high-resolution spectrometer (EUVHRS). (bottom) Effective areas for these three spectrometers.

Peeking through the Picket Fence

(11)9

EUVE / Solar Obs. (2.4 pe) SM ~

60

Coi,.,,,, D.n.Ity • 7.41(1.17

(o~,~.oran,,.

Qui.t Son 1. Ii 1.304 not

13.9 hr.

80 Tolol Coonts

40

1.~

U

I I~ Li

=

1207

~ hi1

•iJ.._~h1_

d

EUVHRS / Solar Obs. (18.0 pc) 500

SM

Hydrogon Coion,n D.nnO~

QI.l.t 2,0,

20

11,,,.

(upon,,,.

,,,‘

13.9 hr.

H. II 1.304 not Inoiod.d

Total Coonts

0 500

5.55(1.15

ii

Lu11

150

10$ 3

~i~I~i IU. ~

200

d

~

250

300

350

~

400

~

300

0

200

Total Counts

Ih

II

Wovoisr,gth (A)

FUSE Phase A 400

156

/

Solar Obs. (8.5 pe)

HIdr090n Colon, 0.n,lty — 2.62(115

350

Qol.t 21111 H. 111.304 not lnolod.d

300

Total Counts 1_shoot SM)

ii

4306

Ii. ~iJ~

liii a

i

I

~

(op000ro On,.

=

—

SM

100

3.9 hr.

lotol

2476

C01I,,ts

WI S 155) =

I 132

150 I

250

Wovolongth (A)

~200. 150 100 50 0 . 100

Total Coonto wI S SM) ..

.o. 150

0

1060

~

~LI,Ioi ,. 200

250

300

.1 ~

.

350

Wovelongth (A)

Figure 3: Comparison of predicted spectra for the average quiet Sun as folded through the response functions of the three spectrometers described in Fig. 2. Each panel consists of the intrinsic spectrum (ignoring the Hell 304 A line) displaced upwards, and the spectrum including absorption the indicated 3 andinterstellar the indicated distance with such that the total neutral including hydrogen interstellar column density computed for 1000 ~H =in0.1 cm seconds. (top) Calculated spectra in counts per counts absorption exceed 50,000 spectral bin for the EUVE spectrometer. (bottom) Same for the Lyman/FUSE Phase A spectrometer. (right)

Same for the EUVHRS spectrometer. FUSE Phase A

60

C

/

Solar Obs. (33.0 pc) O.nulty

H~drog.n Coton.n

—

1.02(418 C1n°

50 0.1.1

500

nr...

(opo.or.

0

13.0 hr.

40

30

20

Toga Coo, • (wltttnot

,

11,1.11

)

= 1537

L

1.w

10

0 900

Total Coons (with 151

.I.~,,. ._I..LU..1.1 950

1000

i~hIt I

, ~

I

1091

I

5050

1

II 5100 111111 £150 I

1200

Wav.l.ngth (A)

A region for the quiet Sun as folded through the Lyman/FUSE prime spectrometer parameters /10/. Plotted are the intrinsic spectrum (displaced upwards) and the spectrum3. including The source absorption distance by was interstellar obtainedhydrogen by requiring (continuum that theand total lines) counts for the exceed indicated 1000 in distance 50,000 and nil =including 0.1 cm interstellar absorption. seconds Figure 4: Comparison of predicted spectra in the 900—1200

JASR 11111—B

(11)10

J.L. Linsky and D.G. Luttermoser

The intrinsic spectrum of the RS CVn-type system AR Lac was computed using the two-temperature emission measure distribution derived from Einstein Solid State Spectrometer data /11/. The photon flux near 100 A, which is comparable to that in the soft x-ray region and contains information on the coronal emission measure over a similar range of temperatures. The predicted spectra of AR Lac for 50,000 seconds exposure are shown in Figure 5(left) both before and including interstellar absorption for the three spectrometers. In each case a high signal/noise EUV spectrum can be obtained in far less that 50,000 seconds. The importance of high throughput is illustrated in Figure 5(right), where we compute spectra for a CrB during its flare on 1983 September 29 /12/. For each spectrometer the exposure time was taken to be 1 minute at the flare peak emission measure to study the time development of the flaring plasma. In this simulation the EUVE spectrometer provides very few photons (although an acceptable number in a 10 minute integration), while the other two spectrometers provide good statistics in only 1 minute integrations. The importanceof simultaneous high spectral resolution and high throughput is illustrated in Figure 6, which shows simulated spectra of the FeXX 133 A line from AR Lac folded through the EUVHRS. For this simulation half of the stellar flux is assumed to be uniformly distributed across the surface of the 1<0 IV star and the other half within a small coronal structure located at different projected radial distances from the center of the star and co-rotating with the star whose surface equatorial rotational velocity is 70 km s~.The influence of the active region on the detected line profile indicates that the EUVHRS can locate bright regions in the corona of this star using the Doppler imaging technique /13/. In our final simulation (Figure 7) we show that the Prime spectrometer of Lyman/FUSE has sufficient spectral resolution and effective area to measure Doppler shifts in the transition region of a star similiar to a solar active region but located at the distance of the Hyades. The counting statistics are sufficient to measure a Doppler shift as small as 2-3 km s’ due either to a stellar wind (blue shift) or to downflows in magnetic flux tubes /14/. HOW MANY STARS MAY BE DETECTED IN THE EUV? The Wide Field Camera on ROSAT will soon tell us how many stars of different spectral types are detectable with broad-band photometry in the EUV. Prior to the start of their survey, we would like to hazard some guesses concerning how many may be detected spectroscopically in the off chance that we may be right. Table 2 lists the number densities of RS CVn, W UMa, and dMe stars estimated from the Einstein Medium Sensitivity Survey (EMSS) /15/ and an estimate for the cataclysmic variables /16/. Using these space densities, we list in Table 3 the number of objects out to a range of distances that lie within a hydrogen column density 2 estimated using the LISM map of Paresce /9/. This value of NH corresponds to 20% of NH 1019 cm transmission at 200 A and 80% transmission at 100 A. The detection horizons for the three spectrometers summarized in Table 4 indicate that each instrument could detect many of each class of object.

Table 2. Number Densities of Different Types of Stars Star Type Density (pc3) Reference RS CVn 2.89 ±0.62 x 10~ EMSS /15/ W UMa 6.55 ±2.07 x iO~ EMSS /15/ dMe 2.40 ±0.38 x iO~ EMSS, log(L~) 27.8 CV 6 x 106 Patterson /16/

Table 3. Number of Stars within d(pc) for NH ~ i0~cm2 d(pc) N(RS CVn) N(W UMa) N(dMe) N(CV) 10 0.6 0.14 50 0.01 20 4 1 300 0.1 30 12 3 900 0.3 50 20 6 1600 0.6 100 340 100 30,000 8 200 1600 500 150,000 40

Peeking through the Picket Fence

(11)11

20 7000

I

[II

___________________________________________ EUVE / AR Lac Obs. (50.0 pc)

—

EUVE / a CrB Obs. (23.0 pe)

______________________

~ ~

Cthns.a 0.nWt,

SM lt~d,ogonCMo,nn D.n.Ily 1.546*19 o,n’ Co (opoa,r. TIns. 0 13.9 hr.

9000

it~~____

5000 4000

•

7106.11

a,n~

Eop.our. Tb..

unt. (without 3M) • 54770

C

10 ~

~

~

80.0 ..o

0

(w~hootlaM) o 212

3 C

~3000

2000

~~t.(wfthL±ol4~

~otr

1(1(1(1

TAIJ

0

Countu (with

I I I

ISU) w 25408

FUSE_______________________________________________ Phase A / AR Lac Obs. (50.0 pc)

35000

~MHydog

30000

140

FUSE Phase A

j~nt.(w5houtl)o

05 HpMrog.n C.to,nn lon.tty

~ 1f.

25000

453531

~

10000 Total Count. (with 55) w 1369

I

._

-

ETJVHRS

/

AR

Counts (ttheut

Counts (with

0

-

o,n

Ifs)

10.0 ..u

1838

I

20

0

• 7.10(018

6C~OOO~. lb,.,, C

udal

11111 III

90 90

40

2.4

aCrB Obs. (23.0 pc) 1

120

5000

/

CMu.n,Ib.n.Ityo 1546+20

C po.

15000 20000

I

____________________________________________

0

t._.-~___

I

ISM) = 931

~

.

Lac Obs. (50.0 pa)

_________________________________________

8+5

900 I

05 Hydr.~.nCatonn

lonooy 0

1.546*19

C1n

-

—

EUVHRS / a CrB Obs. (23.0 pc)

990

2

tOM Uydrwg.n Colons., bosH, • 7.106*19 .,n~

8+5 Copo...rs TI,,,.

• 13.5 9,,

1.8

8+5

800

C

a

‘gS0

1.2

I

Cuonts (without

8+5 I I__Il

80000

40000

0 100

J

~o~o62

=

C

_______________________________

I .1 ___________________________________________

700 490

0~

Count. (wIth

I

200 100

I

___________________________________________________ 150

200

250

$ I C.un$s (lthout SM)

C

800

o

15687

1.5

300. SM)= 1233118

I.—

300

Wavei.ngth (A)

100

H

I

~.

~l Count. (with tOM) • 7958

I~ I L

.

150

200 WOv.Isr,gth (A)

250

300

Figure 5: (left panels) Comparison of predicted spectra of AR Lac for 50,000 seconds exposure. Each panel consists of the computed spectrum without interstellar absorption (displaced upwards) and including interstellar absorption with the indicated hydrogen column density — (top) for the EUVE spectrometer, (middle) for the Lyman/FUSE Phase A spectrometer in its high-throughput mode, (bottom) for the EUVHRS spectrometer. (right panels) Predicted spectra from for 1 minute peak emission 3) and temperature (9.5a xCrB io~ K) afor the 29 integration September assuming 1983 flarethe observed by EXmeasure (5.7— x(top) i0~for cm— OSAT /12/ the EUVE spectrometer, (middle) for the Lyman/FUSE Phase A spectrometer in its high-throughput mode, (bottom) for the EUVHRS spectrometer.

J.L. Linsky and D.G. Luttermoser

(1 1)12

AR Lac Active Region Simulation (50.0 pc)

3200 2800

£UVHRS D.$.ctor

F. XX ~.I33 59.1

R.tot*o.s V.tsdiy

70 km/s

c.spo.w~.Thai • 3010 mm

2400 IS~Hydn

a’ Columa D.a1t

5’

9.54(419 .m’

2000 Disk C.sstr

1800 0 U

2R

30

1200

800 400.

Wov.I.ngth (A)

Figure 6: Simulated spectra ofthe FeXX 133 A line from AR Lac as observed by the EUVHRS spectrometer with a 30 minute integration including interstellar absorption. Half of the flux is assumed to be uniformly distributed across the stellar surface and the other half in a small corotates 1).concentrated For each simulation theactive dashedregion line isthat a comparison with the (equatorial velocity of 70 km s across the stellar surface. The active region is placed profile in star which all of therotational flux is uniformly distributed at disk center (left), at the limb (second from left), and at 1 and 2 stellar radii above the limb.

200

Solar Wind Simulation (42.0 pc) Acftvi Ityadso 02 Star

Fuo. Phaso A D.tsctar

Exposure Thai

8.0 hm~

150 1511 N 6drsg.r Column DensIty — 1.30(419 cm~

100

078

TR(SolId)

.

*0 km/s (Dotted) 076.5 977 Wov.Isng$h (A)

I~I30km/e (DaIMd)

km/s (Dot—OasIs 977.5 978

Figure 7: Simulated spectra of the CIII 977 A line observed by the Lyman/FUSE Prime Spectrometer from an active G2V star located in the Hyades cluster. The spectra for the per line spectral formed in region at 1. simulated The vertical scaleare is counts bina assuming arest resolution and withofDoppler 30,000. shifts of 10, 30, and 100 km s

Peeking through the Picket Fence

(11)13

Table 4. Detection Horizon (pc) Spectrometer

Detection Criterion i0~counts in i0~seconds Quiet Sun Active Sun dMe AR Lac AS(FUV) EUVE — 2.4 6.5 10 250 — FUSE Phase A 8.5 19 30 600 100 EUVHRS 18 36 60 2000 —

____________

CONCLUSIONS We conclude that while each of the spectrometers will detect a number of stellar sources, high throughput and spectral resolution are needed for such detailed physical studies as Doppler imaging, measurements of stellar winds and downflows, and high time resolution studies of stellar flares. We wish to thank Dr. Alex Brown for providing us with the plasma spectroscopy code, Dr. Webster Cash and Mr. Erik Wilkinson for providing an early version of the code for simulating the instrumental response function, Dr. Tom Fleming for providing space density estimates for different types of stars prior to publication, and Dr. M. Shull for comments on the text. This work is supported by interagency transfer H-80531-B from NASA to the NIST and NASA grant NAG5-82 to the University of Colorado. REFERENCES 1. A.K. Dupree, M.C.E. Huber, R.W. Noyes, W.H. Parkinson, E.M. Reeves, G.L. Withbroe, The extremeultraviolet spectrum of a solar active region, Ap. J. 182, 321 (1973). 2. J.L. Linsky, P.L. Bornmann, K.G. Carpenter, R.F. Wing, MS. Giampapa, S.P. Worden, and E.K. Hege, Outer atmospheres of cool stars. XII. A survey of IUE Ultraviolet emission line spectra of cool dwarf stars, Ap. J. 260, 670 (1982). 3. P.B. Byrne, .J.G. Doyle, A. Brown, J.L. Linsky, and M. Rodonb, Rotational modulation and flares on RS CVn and BY Dra stars. VI. Physical parameters of the chromospheres/transition regions of V 711 Tau (HR 1099), II Peg and AR Lac during October 1981, Astr. Ap. 180, 172 (1987). 4. R.D. Chapman and W.M. Neupert, Slowly varying component of extreme ultraviolet solar radiation and its relation to solar radio radiation, J. Geophys. Res. 79, 4138 (1974). 5. S.O. Kastner, W.M. Neupert, and M. Swartz, Solar-flare emission lines in the range from 66 to 171 A; 2?2p” — 2sT~2pk+l transitions in highly ionized iron, Ap. J. 191, 261 (1974); C. Jordan, The Ionization Equilibrium of Elements Between Carbon and Nickel, M.N.R.A.S. 142, 501 (1969). 6. K.P. Dere, H.E. Mason, K.G. Widing, and A.K. Bhatia, XUV electron density diagnostics for solar flares, Ap. J. Suppl. 40, 341 (1979). 7. U. Feldman, The use of spectral emission lines in the diagnostics of hot solar plasmas, Physics Scripia 24, 681 (1981). 8. P.C. Frisch and D.C. York, Synthesis maps of ultraviolet observations of neutral hydrogen gas, Ap. J. (Letters) 271, L59 (1983). 9. F. Paresce, On the distribution of interstellar matter around the Sun, A. J. 89, 1022 (1984). 10. LYMAN: The Far Ultraviolet Spectroscopic Explorer, Phase A Study Final Report, July 1989. 11. J.H. Swank, N.E. White, S.S. Holt, and R.H. Becker, Two-component x-ray emission from RS CVn binaries, Ap. J. 246, 208 (1981). 12. G.H.J. van den Oord, R. Mewe, and A.C. Brinkman, An EXOSAT Observation of an X-ray Flare and Quiescent Emission from the RS CVn Binary a CrB, Astr. Ap. 205, 181 (1988).

(11)~4

J.L. Linsky and D.G. Luttermoser

13. J.E. Neff, F.M. Walter, M. Rodonb, and J.L. Linsky, Rotational Modulation and Flares on RS CVn and BY Dra Stars. XI. Ultraviolet Spectral Images of AR Lac in September 1985, Asir. Ap. 215, 79 (1989). 14. T.R. Ayres, E. Jensen, and 0. Engvold, Redshifts of High-temperature Emission Lines in the FarUltraviolet Spectra of Late-type Stars. II. New, Precise Measurements of Dwarfs and Giants, Ap. .J. Suppi. 66, 51(1988). 15. T.A. Fleming, private 958 (1988).

communication.

For previous estimates see A. J. 98, 692 (1989) and Ap. J. 331,

16. J. Patterson, The Evolution of Cataclysmic and Low-mass X-ray Binaries, Ap. J. Suppi. 54, 443 (1984).