Caii Absorption in the Circumstellar Disk of Beta Pictoris and Other A

CaI I absorption in the

circumstellar disk of Beta Pictoris

and other Atype stars

A thesis

submitted in partial fullment

of the requirements for the Degree

Masters of Science in Astronomy

in the

University of Canterbury

Orlon K L Petterson

University of Canterbury

Dedicated to my parents

George and Chew Yoong

who have always help ed

me to achieve my b est

Abstract

Presented here are the results of observations made at Mt John University Observatory

MJUO during the sp ectroscopic campaigns to observe Pictoris in

and observations conducted in to characterise the b ehaviour of the Ca II H and K

lines and to test the Falling Evaporating Bo dies scenario Using the metho d of division

by a reference sp ectrum b oth narrow and broad variable absorption features in b oth the

redshifted and blueshifted sides of the Ca II H and K lines are clearly detected The large

data set obtained allows the determination of the evolution in terms of velocity equivalent

width FWHM and timescales of variability of the variable absorption features These are

then compared with the results of LagrangeHenri et al in their pap er on the

observing campaign

LagrangeHenri et al nd that there are velocity regimes and this is conrmed in

the MJUO data The higher the redshift the smaller the variability timescales and the

smaller the absorbing cloud In contrast the low velocity features tend to b e longer lived

and to have the deep est absorptions The correlation b etween the FWHM and velocity of

the features found by LagrangeHenri et al is conrmed but with the larger set of data

the correlation is found to b e somewhat weaker Signicant activity was seen in each set

of observations with longlived absorption features at low velocity almost always b eing

present and it has b een found that of all features observed are most likely due to

more that one FEB

The eect of stellar rotation is suggested in the data of some of the strong and more

variable absorption features However conclusive evidence of the changes in equivalent

width are not forthcoming Large numbers of high velocity features are also observed and

are seen to vary in timescales no longer than the crossing time for an orbiting b o dy to pass

across the stellar disk This lends further supp ort to the FEB scenario as an explanation

for the variable absorption features The measurement of the lling factors of the clouds

of ions indicate that these clouds do indeed cover large fractions of the stellar disk and

some of the lines even exhibit pKpH less than as predicted

The FEB scenario app ears to explain many of the characteristics of the variable ab

sorption features very well simulations can repro duce many of the absorptions however

there are some cases where the FEB scenario fails to adequately explain the observations

The ability for some of the longlived features to last as long as they are observed to

would require either large numbers of b o dies on similar orbits crossing the line of sight

for many weeks or that there is some other explanation for the origin of the absorptions

The quest for know ledge eventual ly l ls al l Time and Space

Contents

Figures xiii

Tables xiv

Introduction

The history of mo dels of the Solar Systems formation

The nebular theory

The tidal theories

Accretion theories

Mo dern theories of Solar System formation and evolution

Formation of a protostar and circumstellar disk

Cloud collapse and fragmentation

Protostellar ob jects and disks

Planetary formation

Termination of formation

Pictoris

The evolutionary status of Pictoris

The circumstellar disk

Variations in the sp ectroscopic lines an overview

The Falling Evaporating Bo dies scenario

Protoplanetary disks around other stars

Aims of this Thesis

Selection of targets

Evaluation of the Falling Evaporating Bo dies scenario

Observations

Observing log

The telescop e and echelle sp ectrograph

MJUO CCD system

Observation pro cedure

Observations of Beta Pictoris

Observations of Ophiuchus

Observations of Alpha Pisces Australus v

vi Contents

Data reduction and analysis pro cedures

Munich Image Data Analysis System

The MIDAS echelle reduction package

Mo dication to MIDAS

Determination of resolving p ower

Presentation of Sp ectra

The Beta Pictoris H K lines

The observing campaign

The observations

The Ophiuchus K line

The Alpha Piscis Australus H and K lines

Characterisation of Pictoris sp ectra

Normalising the sp ectra

Fitting the absorption features

Analysis

The observing campaign

observations Beta Pictoris

Summary of observational results

Conclusion

Future work

Acknowledgements

References

A Reducing MJUO CCD Echelle Sp ectra Using ESOMIDAS

A Creating Images of the Correct Format for MIDAS

A Initializing for the MIDAS Session

A Lo cating the Echelle Orders in the Images

A Setting the Oset of the Orders in The Stellar Sp ectrum

A Calibrating the Echelle Sp ectrum for Wavelength

A Filtering the Stellar Sp ectrum

A Preparing the Scaled Smo otheld Image

A Reduction of the Stellar Sp ectrum

A Saving the Session

A Saving data as FITS format les

A Creating MIDAS Tables

A Other Useful Commands

B MIDAS command language programs

B Selectprg

B Echorlonprg

B Echorderprg

B knormprg

B orlonktprg

C Data extracted from sp ectra vii

Figures

A schematic diagram showing the formation of cores of dense gas and dust

that will serve as the seeds of new stars in a dark molecular cloud

H image of M showing one site of star formation where a blob of gas

and dust containing still forming stars is slowly b eing ero ded away by

ultraviolet light from nearly hot stars courtesy of NASA

Accretion of material onto the rotating protostar and surrounding accre

tion disk

Schematic representation of the bip olar outow from young stellar ob ject

as material continues to accrete

HST images of bip olar jets from young stars courtesy NASA

ZAMS star surrounded by a slowly ero ding circumstellar disk

Edge on protoplanetary disks seen silhouetted against the bright back

ground of the Orion Nebula courtesy NASA

Backman et al mo del for the circumstellar disk of Pictoris Not to

scale

HST coronagraphic image of the circumstellar disk of Pictoris which

reveals a thinner disk than previous ground based images reveal

HST coronagraphic image of the circumstellar disk of Pictoris which

shows an S shap ed warp which is interpreted as the signature of planets

closer to the star

Plots of the simulations of an FEB in an orbit close to simple infall and

passing close to the star The rst plot shows a dimensional view of the

Ca II cloud around the nucleus with each dot representing a Ca II ion

which is numerically followed The second plot is the synthetic sp ectrum

The third plot is a schematic representation of the situation see Beust et

The plots here are similar to Fig but the b o dy approaches the star

much closer

Theoretical diagram of the depths of the H K lines The curves are for

dierent values of the lling factor of the absorbing cloud for a non

rotating stellar disk The values of can infer the amount of the stellar

disk b eing o cculted PK and PH are the central depths of the absorption

features see LagrangeHenri et al viii

Figures ix

The exp ected evolution of the equivalent width of the K line for an FEB

crossing the line of sight from dierent directions when the rotation of

Pic is taken into account

The exp ected evolution of the equivalent width of the H line for an FEB

crossing the line of sight from dierent directions when the rotation of

Pic is taken into account

HST images of protoplanetary systems in the Orion nebula courtesy of

NASA

The exp ected evolution of an FEB crossing the line of sight

K line region smo otheld showing the atness of the CCD

Thoriumargon lines for CaI I K line region

Thoriumargon lines for CaI I H line region

Plot of airmass vs time of year for Pictoris

Plot of airmass vs time of year for Ophiuchus

Plot of airmass vs time of year for Alpha Pisces Australus

A reduced K line sp ectrum with its corresp onding cosmic ray sp ectrum

which is used for conrming the reality of any variable absorption features

Sp ectra taken on Dec

Sp ectra taken on Jan

Sp ectra taken on Apr

Sp ectra taken on Oct

Sp ectra taken on Nov

Sp ectra taken on Dec

Sp ectra taken on Nov

x Figures

Sp ectra taken on Nov

Sp ectra taken on Dec

Sp ectra taken on Nov

Sp ectra taken on Apr

Sp ectra taken on May

Sp ectra taken on Jun

Sp ectra taken on Jul

Sp ectra taken on Aug

Sp ectra taken on Sep

Sp ectra taken on Oct

Sp ectra taken on Nov

H K line sp ectra of Ophiuchus taken from MJUO in May and June

H K line sp ectra of Ophiuchus taken from MJUO in June and July

Figures xi

K line sp ectra of Ophiuchus taken from MJUO in July August and

September

K line sp ectra of Ophiuchus taken from MJUO in September

K line sp ectra of Ophiuchus taken from MJUO in September and Octob er

H K line sp ectra of Piscis Australus taken in May and June

K line sp ectra of Piscis Australus taken in June and July

K line sp ectra of Piscis Australus taken in July August and September

K line sp ectra of Piscis Australus taken in September

K line sp ectra of Piscis Australus in Octob er

K line sp ectra taken on June showing the broadened circumstellar

line and also a clear HVF

The normalised sp ectrum after b eing divided by the reference sp ectrum

Plot showing the normalised sp ectrum and tted gaussians to the variable

absorption features Note the well dened gaussian shap ed absorption at

kms

December

January

April

Octob er

November

December

November

xii Figures

November

April

May

June

July

August

September

Octob er

November

Plot of the MJUO data for showing the strong correlation in FWHM

vs velocity as found in LagrangeHenri et al More data is present

in this plot than LagrangeHenri et al found

Plot of the MJUO data for all years showing the much weaker correlation

in FWHM vs velocity

Plot of the depths of the absorption features observed vs their velocities

over all observations

Plot of the equivalent widths in mAof the absorption features observed vs

their velocities over all observations

Plot of lling factors for quasisimultaneous observations in H K Squares

represent blueshifted features crosses LVFs from kms and trian

gles for HVFs ab ove kms xiii

Tables

Stellar parameters of Pictoris

Stellar parameters of Ophuchius

Stellar parameters of Pisces Australus

Table of observations made at MJUO in

Settings on the MJUO echelle sp ectrograph

C Table of results of analysis of Pic Ca II sp ectra xiv

Chapter

Introduction

Stars are probably the most basic ob jects one can see in the sky at night One of the

most interesting and fundamental problems in mo dern astrophysics is understanding their

origins Of particular interest in this thesis is the formation and evolution of low mass

M stars which can form planetary systems capable of supp orting life in a form

as we know it With the recent announcements of planetary b o dies orbiting other solar

type stars there is a need to understand the pro cesses b ehind planet formation and the

circumstellar disks they are b elieved to form from

The history of mo dels of the Solar Systems formation

Question How did the Sun and planets come into b eing

This question was really only considered after the Cop ernican revolution The theory

that the Sun was at the centre of the Solar System with the planets and other b o dies

orbiting around it was rst advanced by Aristarchos of Samos but wasnt really accepted

until the scientic work of Kepler Galileo and Newton proved that Aristotles view of the

Solar System didnt accurately describ e the workings of the Solar System It was clear in

the b eginning that any theory advanced would have to take various factors into account

the orbits of the planets are close to the plane of the Earths orbit

the orbits are to a certain degree of approximation roughly circular

the planets all orbit in the same direction as the Suns rotation

At the b eginning of the th century theorists had an added problem to solve in the

formation of the Solar System namely that even though of the mass of the Solar

System resides in the Sun it only has of the angular momentum p ossessed by all

the planets see Mouschovias Late in the second half of the th century new

theories of stellar formation and new data ab out the ages of various b o dies have help ed

to advance our understanding of Solar System formation Finally observations by the

Hubble Space Telescope are giving Astronomers unprecedented views of stars in various

stages of formation These observations are undoubtedly helping to constrain and advance

theories of planetary formation ab out other stars as well as the formation of our own Solar

System

Chapter Introduction

The nebular theory

The concept of a primitive nebula from which b oth the Sun and its system of planets

were b orn was rst prop osed by Kant and Laplace According

to Laplace the nebula contracts under gravitation and its rotational velocity increases

until it collapses into a disk Subsequently rings of material are shed from which form

planets and satellites

This mo del had the merit of b eing able to explain all the observational phenomena

known at the time with resp ect to the motions of the planets However later two problems

arose with this mo del namely the diculty with which planets formed from the rings of

planetoids and then the distribution of angular momentum in the Solar System Under

this mo del the Sun should still have most of the angular momentum and hence b e rotating

at the rate of rotation every hours as opp osed to the day p erio d at present

During the th century Laplaces mo del has b een mo died and expanded in various

ways to answer the ob jections that have b een raised The role of the solar wind and

magnetic eld have taken on an increasing imp ortance as ways to brake the rotation of

the Sun A rotating star loses mass through the ow of particles along the lines of force

of its magnetic eld This transp orts the particles to large distances far greater than

the stellar radius and therefore leads to a large loss of angular momentum This view is

backed up by observation which shows that young stars tend to have intense magnetic

activity and high rotation rates As these stars age their magnetic activity declines and

the rotation rate slows This explanation involving magnetohydrodynamic braking of a

star removes the ob jection ab out angular momentum

The tidal theories

The rst catastrophic theory of the formation of the Solar System was prop osed by the

naturalist Buon who suggested that the Solar System was formed from the

ejected material from the Sun resulting from the collision with a comet years ago

This theory had no scientic justication but with ob jections raised with the the nebular

theory the catastrophic theories were reexamined At the b eginning of the th century

the Buons comet was replaced by a star and the planets formed by the condensation of

material lost from the Sun The tidal force at the moment of closest approach b etween the

two stars would then draw out a lament of material from one star and this would orbit

the star with considerable angular momentum However it was found that this mo del had

problems with making large planetary masses Another version prop osed that the Sun

was part of a binary system and the planets formed from this second star It was later

suggested that the Sun encountered a protostar from which the planets formed this to ok

advantage of the new determinations that the planets formed from a cold medium

Mo dern theories of Solar System formation and evolution

Accretion theories

This nal class of mo dels consider the p ossibility that the planets formed around the Sun

from accreting interstellar material In order to avoid the accreted matter from falling

into the Sun another nearby star is p ostulated which allows the material to condense

into planets Another version supp oses that the Sun encountered two dierent nebulae

one consisting of nonvolatile grains from which the inner planets evolved and another

consisting of mainly hydrogen from which the Jovian planets formed However this

hypothesis was rejected when it b ecame apparent that the collision of particles would

lead to disp ersal rather than planet formation

Mo dern theories of Solar System formation and evolution

Mo dern theories of Solar System formation tend to follow the nebular theory as prop osed

by Laplace but mo died to take advantage of new observations There are two categories

of mo dels for the evolution of protoplanetary disks which explain planetary formation

They are

The massive nebula model which considers a viscous disk of ab out M A large

fraction of this disk is removed by the action of the solar wind in a very short time

while the rest is accreted by the Sun The planets are to form directly from the

gaseous nebula by gravitational instabilities

The lowmassnebula model where the mass of the circumstellar disk after collapse

is ab out M The disk later co ols the dust accumulates in the central plane

and forms planetesimals which can combine to form planets

The current view seems to suggest that the lowmassnebula mo del is more likely and

indeed do es app ear to b e the case in the Beta Pictoris system So this is the main mo del

to b e discussed in the rest of this chapter

It should also b e noted that the current thinking on the sub ject of circumstellar disk

formation ab out stars is that planets will only b e pro duced if sucient time is available

for accretion of material This constraint therefore rules out O and B stars as sites of any

disk formation as any disk will b e blown away by the strong stellar winds of the young

star b efore any accretion can o ccur Therefore planetary formation will b e limited to

low to mid mass stars M

Formation of a protostar and circumstellar disk

At the most basic level the evolution of a low mass star from birth is driven by gravity

consequently the amount of mass a star has plays a fundamental role in its development

Chapter Introduction

Star formation o ccurs on widely varying scales ranging from relatively isolated low

mass M dark clouds to giant molecular cloud complexes containing as much as

M of gas and dust Giant molecular cloud complexes are also the sites of high

mass star formation and therefore can inuence their surroundings considerably Since

lowmass star formation is what interests us here our attention is directed to the dark

cloud complexes with masses of M In these dark cloud complexes lowmass

star formation o ccurs largely without highmass stars Dark cloud complexes contain

clusters of lo calised mass concentrations termed dense cloud cores with masses in the

range M mean densities of cm sizes on the order of p c and

temp eratures close to K These dark cloud cores dene the initial conditions for

the formation of many lowmass stars see Figs

Figure A schematic diagram

showing the formation of cores of dense

Figure H image of M showing one site of

gas and dust that will serve as the seeds

star formation where a blob of gas and dust contain

of new stars in a dark molecular cloud

ing still forming stars is slowly b eing ero ded away

by ultraviolet light from nearly hot stars courtesy

of NASA

Cloud collapse and fragmentation

The classical initial conditions for cloud collapse are an optically thin cloud in the infra

red IR allowing dust grains in the cloud to radiate away compressional energy pro duced

by the collapse As a result the initial collapse phase o ccurs isothermally T K A

pressure gradient develops in the envelope as the eects of the nite cloud size propagate

inward and this pressure gradient retards the collapse of the envelope As the density

Mo dern theories of Solar System formation and evolution

b ecomes higher at the centre of the cloud than in the surrounding envelope this region

of higher density collapses faster and the central density b egins to increase in a run away

fashion The time taken for this to o ccur is the freefall time

f f

The run away pro cess of increasing density in the core is halted when the centre of

the cloud b ecomes optically thick in the IR Then the compressional energy is trapp ed

and the rising thermal pressure halts the collapse of the core The result is the forma

tion of a protostellar core surrounded by an infalling envelope When the temp erature

rises suciently to disso ciate molecular hydrogen in the outer core another collapse

is initiated allowing the collapsing gas to reach stellar densities and the protostar b egins

fusion reactions in its core The collapsing cloud also imparts rotation to the still forming

protostar and forms a surrounding accretion disk The time for the prehydrogen burning

evolution is initially given by the KelvinHelmholtz time

R L

Figure Accretion of material onto the rotating protostar and sur

rounding accretion disk

Protostellar ob jects and disks

At this p oint b oth the protostar and circumstellar disk are still gaining mass through

accretion of infalling gas from the clouds envelope also the protostellar ob ject may b e

accreting material from the adjoining disk see Fig The embedded protostellar ob ject

and its disk app ear as a strong source of IR emission with an eective temp erature of

K At some time during this stage a bip olar outow develops from the developing protostar

which is seen as a highly collimated ow of material see Section The bip olar ows

are generally aligned p erp endicular to the ma jor axis of the disk see Figs

suggesting that the disk may in some way help to collimate the ow

Chapter Introduction

Figure Schematic representation

Figure HST images of bip olar jets from young

of the bip olar outow from young stel

stars courtesy NASA

lar ob ject as material continues to ac

crete

Planetary formation

In this stage of formation grains of dust and ice form in the circumstellar disk in Keplerian

orbits and the dust grains rapidly grow in size This is b elieved to o ccur once turbulence

in the disk ceases so allowing dust grains to coagulate to sizes of the order of cm in

timescales yrs Collisions of dust grains within the disk are b elieved to pro duce the

km sized planetesimals the precursors to planets due to dierential motion caused by

gas drag in as little as yrs Calculations of the subsequent phases of planet formation

typically assume the existence of a swarm of ab out planetesimals in the terrestrial

planet region alone Subsequent planetesimal growth requires collisions of these km

sized b o dies to pro duce km sized b o dies in circular orbits This is probably a run

away pro cess with the largest b o dies growing fastest as the gravitation of these b o dies

increases their eective crosssection Formation of planets o ccurs from collisions of these

more widely spaced b o dies through p erturbations of the orbits from close encounters

resulting in increasing orbital eccentricity So the terrestrial planets probably formed

from the collisions of b o dies of similar sizes indeed the formation of the Mo on around

the Earth is b elieved to b e the result of a glancing blow by a Marssized b o dy

Termination of formation

For the rst time the centre of the disk is dominated by the presence of a hydrogenburning

star arriving on the mainsequence see Figs The young star is observed to b e

rapidly rotating with a great deal of magnetic activity and via its interaction with the

circumstellar disk a form of magnetohydrodynamic braking o ccurs The bip olar outow

Mo dern theories of Solar System formation and evolution

Figure ZAMS star surrounded by

a slowly ero ding circumstellar disk

Figure Edge on protoplanetary disks seen sil

houetted against the bright background of the Orion

Nebula courtesy NASA

previously found with the star fades away and a stellar wind b egins to blow away the

gas and dust in the inner disk This represents the nal stage of formation which is also

known as the clearing out phase The now mainsequence star blows away the left over

material with a stellar wind that strips away the ma jority dust and gas leaving b ehind

the planetary system of planets asteroids and comets Eventually the circumstellar disk

will dissipate and fade from view app ear much as our Solar System do es now

Chapter

Pictoris

The circumstellar disk around Pictoris was rst detected by the IRAS satellite in obser

vations made in when the star was found to have a large infrared excess The disk

was then imaged for the rst time by Smith and Terrille in showing a disk of dust

surrounding the star out to arcsec in a nearly edgeon orientation As a consequence of

this observation Pictoris has b een sub ject to a great deal of scrutiny from Astronomers

around the world The search for other ob jects similar to Pictoris also started at this

time

The evolutionary status of Pictoris

The evolutionary status of Pictoris p oses some interesting problems when the number

of unique characteristics this star are considered It exhibits a prominent and in the

IRAS observations a barely spatially resolved IR excess Aumann et al Telesco

et al a attened nearly edgeon disk of visible and near IR scattering mate

rial extending out hundreds of AUs Smith and Terrille Paresce and Burrows

and a circumstellar gaseous shell detectable as a strong narrow absorption fea

ture in the Ca II K H and Na I D photospheric lines Slettebak Slettebak

and Carp enter Hobbs et al VidalMadjar et al In more

recent times variable absorption features have b een seen in the UV resonance lines of Al

I I I Mg I I Fe II as well as in the Ca II and Na I lines which are sp oradic in app earance

To truly understand this bizarre ob ject we need to have a prop er physical character

isation of the prominent circumstellar disk Are we observing a young planetary system

in formation or is it the remnant of a dusty protostellar accretion disk in its clearing

out phase Or could it b e a bip olar nebula emanating from an older evolved p ostmain

sequence star as Herbig has sp eculated on the optical app earance and an AIV

classication of the central star by Slettebak Using all available observational

data and fully selfconsistent mo dels of the disk the current view is that the disk is es

sentially a attened swarm of orbiting particles viewed in a nearly edgeon disk and the

inner region of the disk may have ob jects akin to stargrazing comets causing the variable

absorption features seen

As Smith and Terrille rst p ointed out the absolute brightness of Pictoris is

ab out magnitudes fainter than a typical AV star and even magnitudes fainter than

The evolutionary status of Pictoris

the ZAMS The determination of sp ectral type came from Kondo and Bruhweiler

who found that the IUE sp ectra of Pictoris matched p erfectly the IUE archival sp ectrum

of the unreddened AV comparison star U Ma More sp ecically the argument is as

follows Pictoris is near enough parallax arcsec for an E B V

Its observed B V with V Hoeit and Jaschek This colour

corrob orates Kondo and Bruhweilers result that its UV sp ectrum is that of an

A star for which T is in the range K for the V to III luminosity classes

and for which the b olometric correction is These data translate directly into

an absolute magnitude M and an absolute b olometric magnitude M

V bol

Since a normal galactic A dwarf has an M in the range dep ending on how

far it has evolved o the ZAMS Allen SchmidtKaler its supp osed

underluminosity is indeed in the range to magnitudes It should b e noted that

this discrepancy rises to magnitudes for a typical subgiant of luminosity class IV This

result raises the question of whether circumstellar material might b e absorbing some of

the light from the star hence its underluminosity If this is due to gray extinction from

the disk then the particles must b e quite large however this would then give a very small

eective disk size in the near IR atly contradicting the observed tens of arcsecond disk

claimed by Telesco et al and Backman and Gillett Therefore large particles

blo cking light from Pictoris cannot account for its apparent underluminosity

A thorough analysis of Pictoris sp ectrum and the calibration of narrow band pho

tometry to the main physical stellar parameters allows the reasonably accurate determi

nation of fundermental parameters Because of the complete sp ectral coverage from IUE

data Kondo and Bruhweiler visible and IR data from UBVRIJHK and L

band uxes Hoeit and Jaschek and Glass and IRAS and

uxes Gillett and its well known distance the luminosity of Pictoris is

L Paresce Thus the observed b olometric luminosity is M

bol

The Stromgren indices Hauck and Mermillio d and Crawford et al

are H b y m c while the Geneva photometry

Ruferner yields d B V m From the Stromgren

photometry one can see that Pictoris is unusual in that it has the b y c indices as

exp ected for a AV but a lower than average metallicity as indicated by its p osition in

the m vs c diagram Stromgren

The observed values of the Geneva indices can b e placed on the d vs B V graphs

for various M H values and iterated until the appropriate parameters converge The

resulting parameters are log g T K and M H dex This

gives Pictoris metallicity as appropriately of the Solar abundance which would

explain why Pictoris app ears underluminous for an AV star Paresce Some

other fundamental stellar parameters can now b e extracted such as R R and

Note that there are no metal lines from which a direct determination of metallicity can b e made

Chapter Pictoris

M M b oth slightly lower than the typical A dwarf but still within a quite

plausible range

The kinematics of Pictoris as measured by its UVW velocities which are all less

than kms Woolley et al suggest a normal galactic disk star while its

metallicity is characteristic of a more distant and older thick disk star But its high

rotational velocity of kms is typical for a AV star Uesugi and Fukuda

On closer insp ection however this seemingly inconsistent result with the standard picture

of galactic evolution app ears to b e a p eculiar and somewhat widespread phenomenon

among A stars Wol Lance

Using the Green Demarque and King revised Yale iso chrones for the Z

Y T K case which is most appropriate for the ab ove data one

obtains given the uncertainties in the data and the analysis an upp er age of yrs

Therefore Pictoris is most likely on the ZAMS and has an age of yrs and certainly

not exceeding yrs Paresce

The circumstellar disk

Although Pictoris had b een known as a shell star for decades the rst detection of the

circumstellar disk came when Pictoris was examined by the Infra Red Astronomical

Satellite IRAS in The observations of Pictoris revealed it was among the Vega

like stars which all exhibit infrared excesses in the IRAS bands of and m

This was interpreted to mean that these stars had co ol circumstellar material surrounding

them resulting in the observed excesses Aumann et al

In fact of all the Vegalike stars Pictoris has the largest overall IR excess seen by

IRAS and exhibits the only substantial excess at m Gillett Quite so on

after this detection the circumstellar material was imaged around Pictoris by Smith

and Terrille in whose coronagraphic images revealed the dust to b e distributed in

a highly attened circumstellar structure suggestive of a disk seen nearly edgeon The

00 00

disk was traced out to from the star in the NE and SW directions AU

at p c Smith and Terrille measured an r p ower law for the pro jected brightness

of the disk Smith and Terrille in later observations Smith and Terrille also

traced the disk out to more than AU and also noted that at ab out AU the disk

app ears to b e AU thick

Broadband B V R I multicolour photometry by Paresce and Burrows

and Smith and Terrille revealed no colour dierences from their coronagraphic imaging

Therefore the particle sizes are likely to b e m in radius When these multicolour

measurements are combined with the IRAS data the disk is found to have an inner region

nearly clear of dust or the extinction exp ected is greater than observed Artymowicz et

The circumstellar disk

m and m groundbased imaging with arcsec resolution by Telesco et al

was able to constrain the mo dels by revealing an inner region relatively clear

of dust p ossibly out to AU from the star

Further imaging using antiblooming CCDs of the Pictoris disk by Lecavelier des

Etangs et al found the disk ts an r p ower law Multicolour photometry

reveals a at alb edo for the dust particles demonstrating the probable large sizes of the

grains in the outer disk as predicted by Paresce and Burrows and Backman et al

However the decrease in brightness in B diers from that in longer wavelength

00 00

lters From where measurements b egin out to the B signal is a factor of less

than would b e exp ected from inward extrap olation of the brightness outside This

change in alb edo indicates a change in the nature of the grains This is exp ected b ecause

of ice sublimation in this region AU from the star A dierence in the grains

o ccurs due to their less icy nature which should reveal more ab out the dust comp osition

of the circumstellar disk The disk also shows some asymmetry in the images which could

b e the signature of planet formation

Observations from m to m by Telesco and Knacke detect the sp ectral

signature of silicates This is only seen within arcsec of Pictoris but this silicate

emission at m suggests small grains m present near the star Aitken et

al from m m sp ectroscopy conrms the presence of silicate emission

probably from small m grains in the disk The sizes are b elieved to b e similar

to cometary dust They also note that large particle sizes are required for the dust

particles for orbital stability against radiationpressure blow out or PoyntingRobertson

drag However thermal mo delling by Backman et al requires particles near

m in size to repro duce the IR measurements Therefore the small particles must b e

replenished from collisions by larger b o dies The Backman et al mo del can b e

summarised by Fig

This mo del quite simply describ es what is currently known ab out the Pictoris disk

Polarisation observations of the Pictoris disk indicate it exhibits p olarisation ori

entation characteristics similar to those of a simple reection nebula with p olarisation

of Gledhill et al These observations in conjunction with the

known colours of the disk suggest large grains m typical of interplanetary grains

are present Also Knacke et al shows that the dust around Pictoris cannot

b e interstellar like indeed a b etter match is found with cometary sp ectra like Halleys

indicating that the dust ab out Pictoris has b een pro cessed in some way This data

is supp orted by Savoldini and Galletta whose CO measurements of Pictoris

shows the circumstellar material is depleted in CO which indicates that the star is in

quite an evolved state This requires some active pro cess of gas transformation by means

of accretion of solid b o dies or conversely evaporation

Lagage and Pantin in their pap er note dust depletion in the inner disk and

Chapter Pictoris

Outer Disk

Ices 1000 AU (>60")

Inner disk Large grains refractory material (silicates)

100 AU

0−50 AU? (6")

small (large?) grains

Figure Backman et al mo del for the circumstellar disk of Pic

toris Not to scale

Variations in the sp ectroscopic lines an overview

also an asymmetry in the structure of the disk which hints at the existence of planets

These asymmetries in the circumstellar disk are conrmed in the pap er by Kalas and

Jewitt which demonstrate the existence of many asymmetries such as

A size asymmetry where the NE extension of the disk can b e measured out to

AU from the star while the SW extension is seen only out to AU

A brightness asymmetry is also observed in the outer disk b eyond with the NE

extension b eing observed to b e brighter than the SW extension

Other asymmetries are also mentioned in their pap er which could lend supp ort to the

idea of a Jovianlike planet causing p erturbation in the dust disk

One of the more interesting observations to come to light is the p ossible detection of

a planet orbiting Pictoris in Lecavelier des Etangs et al pap er who show that

Geneva Observatory photometry of Pictoris shows a gradual brightness and dimming

in the visual magnitude which could b e interpreted as the dustfree zone surrounding a

planet coming into view and antiocculting the star allow more of the stars light to b e

seen

Of great interest in the future will b e the result gleaned from observations of the

refurbished Hubble Space Telescope like Fig which shows that the disk around

Pictoris is a lot thinner than previously thought This lends further supp ort to the theory

that planets have formed from the circumstellar disk and also to the presence of large

numbers of planetesimals which are b elieved to b e observed sp ectroscopically see Section

In g the disk app ears slightly warped had this b een present when the star formed

it should have attened out by now Therefore this has b een taken to mean there is a

planet or planets pulling at the disk and distorting it

Future observations of the disk which can see closer into the star should help to prove

one theory ab out the presence of arclike structures in the disk in the planetforming

region It is exp ected that if there are planets in orbit ab out Pictoris they may reveal

their presence by leaving regions clear of dust in their wake

Variations in the sp ectroscopic lines an overview

Because Pictoris is seen edge on it is well suited for absorption line studies of its circum

stellar gas High resolution observations of the sp ectrum of Pictoris reveal the presence

of various anomalous features in some of its sp ectral lines These features are sharp ab

sorption features in the cores of some photospheric lines as well as broad absorptions in

the wings of those lines The observed gas consists generally of lowionization material

somewhat similar to that in interstellar HI regions Its temp erature is T K and a

typical density is nH cm Kondo and Bruhweiler Hobbs et al

Chapter Pictoris

Figure HST coronagraphic image of the circumstellar disk of

Pictoris which reveals a thinner disk than previous ground based images

reveal

VidalMadjar et al The circumstellar gas elements so far observed with these

features are Na I Fe I C I Fe I I Mn I I Ca I I Zn I I Mg I I and Al I I I The observed

features can b e divided into broad categories

A saturated circumstellar line centred at the radial velocity of Pictoris

Sharp absorption lines at low velocities almost always redshifted

Broad absorption features found mainly at high velocities

The circumstellar line

Is this line interstellar or circumstellar in origin Up on examination of the narrow ab

sorption lines of Ca II and Na I seen in the sp ectrum of Pictoris the interstellar origin

can b e ruled out on several grounds

Variations in the sp ectroscopic lines an overview

Figure HST coronagraphic image of the circumstellar disk of

Pictoris which shows an S shap ed warp which is interpreted as the

signature of planets closer to the star

For a star at p c the K line is much stronger than the interstellar K line seen

towards any other comparably near star

A ratio NCaI INNaI is found only in interstellar lines arising in high velocity

gas in contrast to the case for Pictoris

Narrow absorption lines arising from excited metastable lower levels of Fe I I from

which no interstellar absorption has b een detected have b een extensively observed

in the sp ectrum of Pictoris

Having determined that the line cannot b e of interstellar origin Hobbs et al

where do es the gas come from to form this circumstellar absorption line given that an

AV star would quickly dissipate any Calcium around it

There is considerable evidence the circumstellar Calcium K line comp onent is fully

saturated Simultaneous observations of K and H allow the measurement of the doublet

ratio The equivalent widths of K and H were found to b e and mA re

sp ectively giving a doublet ratio of The doublet ratio strongly implies that

there is negligible ux in the core of the circumstellar lines Ultrahighresolution obser

vations of the Ca K line by the Ultra High Resolution Facility at the AngloAustralian

Chapter Pictoris

Telescope show that the circumstellar Calcium K line core is at within the noise which

is characteristic of fully saturated absorption lines It should b e noted that if this line

is truly saturated it is unique in that no interstellar line has b een observed to b e fully

saturated Crawford et al

The source of the Ca II ions for the stable circumstellar line is probably from grains

in the circumstellar disk evaporating near the star at a distance of ab out AU As the

grains spiral towards the star due to PoyntingRobertson eect the intense starlight causes

the grains to evaporate releasing the Calcium back into the gas phase and pro duce the

bulk of the apparent stable part of the absorption Lagrange et al This pro cess

insures that there is a constant supply of material to maintain a saturated absorption line

This has b een termed the grainevaporation hypothesis by VidalMadjar et al

Hobbs et al concluded that most of the Ca II absorption o ccurs at a distance

of order AU This is consistent with the grainevaporation hypothesis

When other lines are examined it app ears that the source of the stable absorptions

are lo cated in dierent regions Na I and Ca II are lo cated in two dierent regions of

the circumstellar disk the Na I in an outer region which contains nearly all of the mass

of the disk Also in this region Calcium is underabundant by a factor of p erhaps

VidalMadjar et al Finally VidalMadjar et al have determined

that Ca II ions are blo cked from leaving the Pictoris systems by collisions within the

stable gaseous cloud that forms the torus around the star pro ducing the circumstellar

feature Hence why we observe the stationary ions

The variable features

Observations of the Calcium K line at dierent ep o chs shows drastic changes on

timescales of months weeks days and hours In the wings of the photospheric lines

these absorptions can app ear and disapp ear at will and tend to b e broad in app earance at

high velocities and at lower velocities narrow and deep These changes could b e related

either to a classical shell close to the star or to the much more extended circumstellar

dust disk around Pictoris These variations could b e due to a number of causes

A stellar photospheric phenomenon

This can b e excluded Such a mechanism o ccurring in the stellar atmosphere should

aect all the main photospheric lines The variable features however are only observed in

the strongest resonance lines and some metastable lines of Fe II LagrangeHenri et al

A shell phenomena close to the stellar surface

Having previously b een classied as a shell star by Slettebak in Pictoris shows

a Ca II K line prole similar to those of other Atype stars Since the velocity extent of

the Ca II p erturbations seen in Pictoris is kms for the easily detected absorption

cores signicantly b elow the stellar escap e velocity kms the corresp onding gas

was not able to escap e from the star It should thus fall back on the star and this could

Variations in the sp ectroscopic lines an overview

explain the observed redshift If at other ep o chs p ermanently blueshifted material is

detected it will strengthen this p ossibility

IUE sp ectra from late to early show

minimal shell column densities

sp ectra showing outowing plasma

broadened but stationary absorption

sp ectra showing infall

Ejection of a circumstellar envelope can b e exp ected to pro duce all phases that have

b een observed Bruhweiler et al These eects would b e observed as mass

gets blown o the star pro ducing the blueshifted features and falls back in to the

star having not achieved escap e velocity hence pro ducing the observed infall some of

this gas maybe for a time left in orbit ab out the star b efore falling back in hence the

broadened but stationary absorption

A phenomenon related to the dust shell present at larger distances from

the star

The circumstellar comp onent of the K line is due to the gas asso ciated with the in

ner disk at ab out AU and enriched in Calcium b ecause of dust grain evaporation

One therefore exp ects a relatively stable central core shap e as observed If the vari

able absorptions are due to inhomogeneities in the continuously infalling dusty material

and these inhomogeneities are stable their orbital p erio d should then pro duce p erio dic

p erturbations of the line

Discrete events of freely falling material

However one can interpret the observed variations as b eing due to discrete events of

freely falling material Large b o dies like cometary nuclei could plausibly account for such

observations Ferlet et al The clumpy structure of the infalling material is

conrmed by LagrangeHenri et al which lends further supp ort to some form of

discrete phenomenon

The observed variability in the infalling plasma would imply under a cometary inow

mo del star grazing comets p er year LagrangeHenri et al This

rate is currently higher than is observed in the Solar System but is compatible with

the rate inferred for the initial phase of the Solar System Beust et al Telesco

et al Lagrange et al This scenario which requires a large star

grazing comet b ombardment rate can account for the infalling plasma and p otentially

for intermittently enhanced stationary plasma in Keplerian orbits

The presence of Al III variable absorption features in the sp ectrum of Pictoris

presents something of a mystery Photoionization by an AV star like Pictoris is

unable to pro duce such a highly ionized sp ecies implying that an additional mechanism

Chapter Pictoris

is required to explain the observed ionization Deleuil et al The presence

of infalling b o dies may b e one of the few p ossibilities that can adequately explain the

presence of a highly ionized sp ecies such as Al I I I

The Falling Evaporating Bo dies scenario

The Falling Evaporating Bo dies FEB scenario was formed to explain the observed vari

ations in some of the metallic lines of the sp ectrum of Pictoris These variations were

observed to b e predominantly redshifted ie falling towards the star at relatively low

velocities

The suggestion that the observed variations could b e caused by cometlike b o dies

passing close to the star is quite app ealing given the age of the Pictoris system From our

own Solar System we know that the early Solar System under went intense b ombardment

in the rst million years evidence of this can b e clearly seen in our own Mo on So

one exp ects for the Pictoris system with its clearly detectable circumstellar disk that

cometary b o dies will have b een formed in vast numbers Given the edgeon orientation of

the disk the p ossibility of a cometary b o dy crossing the line of sight is greatly increased

if as we exp ect most of the material orbiting the star is conned to the plane of the disk

A mo del for these events has b een developed by Beust et al to explain the

observed variations The mo del assumes that the variations seen are due to evaporat

ing material from a cometary b o dy As a rst approximation they assumed a spherical

distribution of the cometary gas for reasons

the radiation pressure is not very ecient on hydrogen atoms at least not the EUV

ux from an A star

there is no appreciable stellar wind in normal A stars Therefore hydrogen atoms

are not eciently rep elled

One can then consider the density of the gas as a function of parameters

the distance r b etween the comet and the star

the distance x b etween the considered p oint and the nucleus

the outow velocity of hydrogen from the nucleus v

the hydrogen pro duction rate of the comet

The infalling b o dies treated as p oints are assumed to come from the circumstellar disk

itself ie quite far from the star so the orbits can b e taken to b e parab olic ones Beust

et al

The Falling Evaporating Bo dies scenario

Each redshifted absorption is then interpreted as the signature of an infalling solid

b o dy which evaporates in the vicinity of the star Ferlet et al and Beust et al

The grains evaporated from the comet are assumed to pro duce the evaporated

ion These ions are then sub jected to gravity radiation pressure and collisions with

the surrounding gas and grains Simulations are able to repro duce some of the observed

absorption features very well once appropriate values for the orbital parameters are added

LagrangeHenri et al

Figure Plots of the simulations of an FEB in an orbit close to

simple infall and passing close to the star The rst plot shows a

dimensional view of the Ca II cloud around the nucleus with each dot

representing a Ca II ion which is numerically followed The second plot

is the synthetic sp ectrum The third plot is a schematic representation

of the situation see Beust et al

From these simulations it app ears that the prep onderance of redshifted absorption in

the LVFs are due to b o dies falling from a precise direction in the circumstellar disk Beust

et al For these b o dies to b e p erturb ed from a particular direction in the disk

Beust et al prop osed a planet in an elliptical orbit e as the p erturb er to make the

stargrazing comets detectable by sp ectroscopic means Levison et al prop ose

Chapter Pictoris

Figure The plots here are similar to Fig but the b o dy ap

proaches the star much closer

that secular resonances from a planetary system of at least planets a prerequisite for the

existence of secular resonances can align the orbits of some b o dies in such a way one can

repro duce the general features seen in Pictoris Further simulations have repro duced

the features seen at kms as b eing due to b o dies arriving with an inclination of

LagrangeHenri et al

The time taken by a b o dy to cross the stellar disk is estimated to b e of the order of

hrs however many of the observed features are much longer lived hrs Since a

single infalling b o dy cannot explain this observation these events must b e due to several

individual b o dies crossing the line of sight at roughly the same moment but with spread

out p eriastrons This also explains the multiple structure seen in some features These

b o dies do not need to fall in as a simple queue the only constraint is in the inclination

to the line of sight LagrangeHenri et al A very interesting and plausible conse

quence coming from the observations of several b o dies falling in at once is that from time

to time a disk of gas might p ossibly form around the star which is fed by the evaporating

shower of b o dies Ferlet et al

The Falling Evaporating Bo dies scenario

Ca II ions will b e observed when they are AU from the star in to ab out stellar

radii This is the region where most of the observed features o ccur but should Ca II ions

b e closer than stellar radii then one should observe blueshifted features and or broad

shallow redshifted features Beust et al High redshifted features due to b o dies

getting much closer to the star should evolve quicker as the b o dy crosses the line of sight

These events are exp ected as they should also b e the b o dies which pro duce the observed

UV features Beust et al and LagrangeHenri et al

Finally in blueshifted features were observed in the sp ectra of Pictoris a narrow

feature at low velocity and also a broad feature at high velocity The narrow feature with

the small blueshift can b e repro duced with a larger inclination angle b etween the axis of

the orbit and the line of sight The broad feature is easier to interpret as b eing due to

the cometary tail b eing observed as the ions are blown away from the star via radiation

pressure LagrangeHenri et al

A new theoretical development is the ability to derive information on the size of the

cloud of ions from simultaneous observations of Ca II H K dep ending on the level of

saturation of the lines For saturated lines the fraction of the stellar disk o cculted by

the cloud can b e determined easily but for weak and unsaturated lines cannot b e so

well determined Unsaturated lines can only give a lower limit for the absorbing see

Fig

Figure Theoretical diagram of the depths of the H K lines The

curves are for dierent values of the lling factor of the absorbing cloud

for a nonrotating stellar disk The values of can infer the amount of

the stellar disk b eing o cculted PK and PH are the central depths of the

absorption features see LagrangeHenri et al

Chapter Pictoris

A test of the FEB scenario by Beust and Lissauer relies on the details of

the absorption line sp ectra and their temp oral variation and the eect on the absorption

lines of the rotation of Pictoris itself The ratio of the equivalent widths of certain

observed absorption doublets imply that the absorb er is often lo cally optically thick but

only covers a small p ortion of the stars disk Stellar rotation implies that the p ortion of

Pictoris photosphere b eing o cculted has a welldened velocity which may b e dierent

from the star as a whole The depth of the feature will dep end on the p ortion of the disk

b eing o cculted Comets crossing from the redshifted side to the blueshifted side or vice

versa will give dierent patterns for the evolution of the equivalent widths See Figs

and The eect is most striking in the K line

Figure The exp ected evolution of the

equivalent width of the H line for an FEB

equivalent width of the K line for an FEB

crossing the line of sight from dierent direc

tions when the rotation of Pic is taken into

account

The presence of Al III in the sp ectra of Pictoris came as a great surprise given that

the EUV ux of an AV star is not able to pro duce such a highly ionized sp ecies by

photoionization LagrangeHenri et al The FEB scenario suggest a natural

pro cess that can do this namely collisional ionization in the sho ck region of a coma

surrounding an evaporating b o dy A hydrodynamic mo del by Beust and Tagger

studied the comas surrounding evaporating b o dies in which sho ck surfaces form where

high temp eratures can o ccur allowing the ionization of exotic sp ecies such as Al I I Mg II

and Al III to o ccur VidalMadjar et al note in this situation temp eratures as

high as T K are easily achieved The conditions under which Al III can b e formed

is restricted to the immediate vicinity of the star dR Beust and Tagger and

Beust et al

Observations of Al III show that the redshifted line is saturated while the blue part of

it is the normal doublet ratio showing that the blue part is pro duced by a larger p ossibly

covering the whole stellar disk and more dilute plasma VidalMadjar et al

Protoplanetary disks around other stars

From the measurements of Al III it is p ossible to calculate the total mass of gas involved

in the variable features The Al III column densities multiplied by the crosssection of

the star provide a lower limit on the amount of variable aluminium The aluminium

can also b e in other states eg Al I I but once normalised to Solar abundances and to

meteoritic abundances one can determine the mass of the evaporated b o dy This gives

the ob jects having sizes in the kilometre range LagrangeHenri et al These

same calculations can b e carried out for the other ions and do indeed give similar sized

ob jects as a result

One can note that the terminal velocity for Al III is compatible to the free fall velocity

at a few stellar radii from the star while for Ca II the corresp onding terminal velocities

for the LVFs are typical of the free fall velocities at AU Beust et al The

velocities that these lines are observed at do agree with theory as the variations have yet

to b e observed at velocities that exceed the freefall velocities of ob jects near a star

Supp orting this FEB scenario is the completely dierent argument based on the theory

of cometary dynamics in the Solar System which lead Weissman to the conclusion

that the Pictoris disk of matter might b e very similar to the wide cometary reservoir

needed in the Solar System LagrangeHenri et al

Protoplanetary disks around other stars

With the rep orts from the IRAS satellite of certain nearby stars having infra red excesses

in some of the infra red bands b eing used it was b elieved that these could b e the rst

detection of dust disks around other stars and indeed the rst detection of circumstellar

material that could b e the precursor to planets The visual detection of the circumstellar

disk around Pictoris by Smith and Terrile conrmed that the IRAS detection was

indeed of circumstellar material from which planets could form Pictoris has since then

b een the b est candidate for a nearby planetary system that could b e studied from Earth

and indeed given its evolutionary status provides a go o d test of theories on planetary

formation

Of course the detections in the IRAS bands only indicate that there is some material

around the stars reradiating in the infra red light the radiation could b e from the stellar

wind or dust surrounding the star However any detection in the and m bands

in the ratio exp ected for co ol dust around the star would rule out the other p ossibilities

as the nature of the sources Given this the strong excesses around the stars Vega

PsA Eri and Pic were considered the b est indicators of the p ossible presence of

protoplanetary dust disks Aumann et al

With that rst detection of a disk of material around Pictoris other stars in the IRAS

p oint source catalogue were examined for further evidence of the existence of circumstellar

material Reexamination of the IRAS data and further observations of stars in the Point

Chapter Pictoris

Source Catalogue by Walker and Wolstencroft Cheng et al and Chini

et al have determined that a number of the other sources identied as likely

protoplanetary disks indeed have dust around them In particular several of the stars

such as Vega and PsA app ear to have disks seen face on or nearly so

Sp ectroscopic observations have b een made to detect any p ossible disks in a near edge

on orientation like Pic They to o might exhibit a circumstellar line in the core of some

of the photospheric features These searchs have b een made by Hobbs Lagrange

Henri et al and Grady et al who have b een lo oking in the Na I Ca II

lines and IUE sp ectra for evidence of a circumstellar disk or infalling plasma more

stars are now known to exhibit b ehaviour similar to Pic with the sp ectral signature of

infalling plasma they are HD Her and Oph

Smith and Terrile have continued their work in searching for further stars with cir

cumstellar disks and imaging them coronagraphically They rep ort in their pap er

that a survey of more than stars within p c lo oking for disks in an edgeon orien

tation or mo derate inclination found no other examples other than Pic Once again the

stars selected were from among the IRAS Point Source Catalogue with additional stars

added to give a b etter spread in sp ectral type

In Ophiuchus a sp ectroscopic binary was imaged with a disk around it by

VidalMadjar et al However the discovery of the disk has not b een conrmed

by other observers infact it is now b elieved to b e the result of imp erfect removal of the

scattered light from the telescop es secondary spider see Waters et al

In it was rep orted that millimetre wave mapping observations of Fomalhaut

revealed the presence of extended mm emission around the star see IAUC

in observations made in early Subsequent observations in failed to detect any

evidence of this emission and it is now b elieved to b e the result of some artifact in the

earlier data see Stern et al

The Hubble Space Telescope has imaged many protoplanetary disks around still form

ing stars in the Orion star forming region as well as other regions of star formation The

Orion images in question show the protoplanetary disks b eing silhouetted against the

bright background of the Orion nebula Other observations made of star forming regions

reveal all the exp ected phases of star formation that theory has predicted As a result of

these observations by the Hubble Space Telescope it would app ear that protoplanetary

disks ab out forming stars are quite common indeed many of the stages of star and cir

cumstellar disk formation are seen and that as a consequence planetary formation is also

fairly common

Protoplanetary disks around other stars

Figure HST images of protoplanetary systems in the Orion nebula

courtesy of NASA

Chapter

Aims of this Thesis

This thesis will lo ok into any variations of the Calcium II H K lines in the sp ectra of

stars which show promise as either protoplanetary systems or as stars surrounded by

the remains of a circumstellar disk from which planets may have formed The Ca II lines

of Pictoris will b e examined to study the variations and evolution of these variations

and compare them with the predictions of the Falling Evaporating Bo dies scenario

Selection of targets

Selection criteria for the stars to b e observed were that they exhibit some evidence of

circumstellar material and b e observable with the equipment available at MJUO To b e

considered for selection the stars had to have an IRAS excess which is considered as a

go o d indicator of p ossible circumstellar material Of these stars the following criteria

were applied to narrow the selection to just a few ob jects The stars had to b e brighter

than th magnitude in order for adequate sp ectra to b e obtained An additional but

optional criterion was that the star had b een coronagraphically imaged and showed some

indication of circumstellar material Two stars in addition to Pictoris were found

that satised the ab ove criteria namely Piscis Australis and Ophiuchus which were

rep orted to have had images taken that revealed the presence of circumstellar material

see the IAUC and Hop efully if these stars exhibit any of the

activity seen in Pictoris it would b e observed and recorded After analysis it was

hop ed something would b e able to b e said ab out the activity around all these stars

The data in the following tables were obtained from The Bright Star Catalogue

and the Hipparcos Input Catalogue See Table for the data on Pictoris Table

for the data on Ophiuchus and Table for the data on Piscis Australis

Evaluation of the Falling Evaporating Bo dies scenario

Previous sp ectroscopic observations of certain sp ectral lines at high resolution in the

sp ectrum of Pictoris reveal a changing structure in the cores of those photospheric

lines aected They have b een divided into types

a circumstellar feature at the radial velocity of the star

Evaluation of the Falling Evaporating Bo dies scenario

Name PIC

RA h m s

0 00

Dec

B V

U B

R I

Sp ectral class AV

PM yr

Parallax

Distance p c ly

RV kms V

v sin i kms

Table Stellar parameters of Pictoris

LVFs which are often quite deep and narrow and app ear to last for several days or

even weeks

HVFs which vary rapidly in timescales of only hours and which are generally broader

and less deep

To understand these varying absorption features in the Calcium II H K lines one

needs to derive the various parameters of interest

the velocities of the features

the widths of the features

the equivalent widths

To do this one must distinguish b etween the photospheric and circumstellar features and

prop erly identify the variable ones One must then either t the photospheric continuum

in the velocity range of interest or divide the sp ectra by another reference sp ectrum

assumed to b e free of variable absorption features Previously the parameters of the

LVFs were measured directly on the sp ectrum after tting the continuum lo cally The

HVFs have only b een observed in sp ectra that have b een divided by a reference one

These general observations of the variable absorption lines in the cores of the sp ectral

lines of Pictoris are all features naturally asso ciated with the predictions of the FEB

Chapter Aims of this Thesis

Name OPH

RA h m s

0 00

Dec

B V

U B

R I

Sp ectral class AVn

PM yr

Parallax

Distance p c ly

RV kms SB

v sin i kms

Separation

Table Stellar parameters of Ophuchius

scenario The FEB scenario has b een able to explain many of the observed features seen

in the Calcium lines and other sp ecies but still many observations need explaining To

test the FEB scenario further more information is required on

the o ccurrence of variable absorption features Particularly do the deep and shallow

absorptions dier signicantly in their frequency of o ccurrence

the velocities duration and the small time scale variations of the variable comp o

nents This will help in determining the orbital parameters necessary to repro duce

the observations in simulations from theoretical mo dels

the geometry and physical conditions in the absorbing cloud namely how large the

clouds of ions are with resp ect to the stellar disk

the prediction by Beust and Lissauer ab out the dierences in the absolute and

relative equivalent width of the Calcium H K features which varies dep ending on

what part of the stellar disk is b eing o cculted see Fig

single and group events Are we seeing single events of hrs duration or longer

lived events most likely due to groups of evaporating b o dies

Evaluation of the Falling Evaporating Bo dies scenario

Name PsA

RA h m s

0 00

Dec

B V

U B

R I

Sp ectral class AV

PM yr

Parallax

Distance p c ly

RV kms

v sin i kms

Table Stellar parameters of Pisces Australus

Another imp ortant aim of this thesis is to compare the results of the international

observing campaign with the more extensive data set contained in this thesis from MJUO

The results of the campaign which observed Pictoris over several days with near

complete coverage has revealed some interesting results The variable features app ear

to b e stronger and more numerous than ever b efore observed The equivalent widths of

the LVFs describ ed in the pap er by LagrangeHenri et al when present were

typically mA whereas the Dec LVFs are always present and with equivalent widths

larger than mA HVFs are also observed much more often For the rst time b oth

the app earance and disapp earance of these highly redshifted events were observed The

HVFs also vary b oth in central velocity and strength on very short timescales of a few

hours When they disapp ear it has b een noted that they get broader In fact the higher

the redshift the smaller the variability timescales the smaller the lling factors of the

clouds and the wider the lines

High resolution studies reveal a clumpy structure to the infall of the Ca II clouds

and the estimate on the rate of infalling events is events p er year under the FEB scenario

Chapter Aims of this Thesis

Figure The exp ected evolution of an FEB crossing the line of sight

Chapter

Observations

Observations were made of Beta Pictoris and the two other stars of interest Fomalhaut

and Ophuchius at the University of Canterburys Mt John University Observatory

The instruments used were the metre McLellan Telescope which had mounted on it

the Observatorys highresolution echelle sp ectrograph and the PM CCD system as

detector

Observing log

Observation runs allo cated on the telescop e were of approximately one week and two weeks

duration p er observing quarter Observing time was allo cated from April till November

to obtain the recent data contained in this thesis Data from earlier observational

runs as part of an international collab oration on Beta Pictoris were also available Ta

ble contains a list of the observations for this thesis

Table Table of observations made at MJUO in

The table ab ove shows the UT date star and number of sp ectra

taken in each of the Calcium II lines observed

Date of observation Star of sp ectra

December Pictoris CaI I K

January Pictoris CaI I K

April Pictoris CaI I K

Octob er Pictoris CaI I K

CaI I H

Octob er Pictoris CaI I K

Chapter Observations

Table contTable of observations made at MJUO in

Date of observation Star of sp ectra

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

December Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

April Pictoris CaI I K

CaI I H

May Pictoris CaI I K

CaI I H

Ophiuchus CaI I K

CaI I H

Pisces Australus CaI I K

May Pictoris CaI I K

CaI I H

Ophiuchus CaI I K

CaI I H

Pisces Australus CaI I K

CaI I H

June Pictoris CaI I K

Observing log

Table contTable of observations made at MJUO in

Date of observation Star of sp ectra

CaI I H

Ophiuchus CaI I K

June Pictoris CaI I K

CaI I H

June Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

June Pictoris CaI I K

June Pisces Australus CaI I K

CaI I H

Pictoris CaI I K

June Pictoris CaI I K

CaI I H

Ophiuchus CaI I K

CaI I H

Pisces Australus CaI I K

July Ophiuchus CaI I K

Pictoris CaI I K

CaI I H

July Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

July Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

July Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

July Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

Chapter Observations

Table contTable of observations made at MJUO in

Date of observation Star of sp ectra

August Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

September Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

September Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

September Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

September Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

September Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

September Ophiuchus CaI I K

Pisces Australus CaI I K

September Pictoris CaI I K

CaI I H

Octob er Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

Octob er Ophiuchus CaI I K

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

Octob er Ophiuchus CaI I K

The telescop e and echelle sp ectrograph

Table contTable of observations made at MJUO in

Date of observation Star of sp ectra

Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

November Pictoris CaI I K

November Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

November Pisces Australus CaI I K

Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

November Pictoris CaI I K

CaI I H

The telescop e and echelle sp ectrograph

Observations were made using the highresolution echelle sp ectrograph with a CCD de

tector to record the observations of the stars Calcium H and K lines at and

angstroms resp ectively The echelle sp ectrograph was mounted on the metre McLellan

telescop e at the cassegrain fo cus The telescop e is a DallKirkham type optical system

and has interchangeable secondary mirrors which allow the telescop e to b e used in

either a f or f cassegrain conguration In this case the f secondary was used

for the echelle sp ectrograph

The echelle sp ectrograph makes use of a gro ove mm echelle grating with a blaze

angle of tan The echelle has a resolving p ower in the visual part of

the sp ectrum of R x see Hearnshaw This gives disp ersion A mm at

A though in the ultraviolet at the Calcium II lines the disp ersion A mm

When using the echelle the blue crossgrating was installed which is used to maximise

Chapter Observations

eciency in the transmission of blue light through the sp ectrograph The slit size was set

to a micrometer setting of m irresp ective of seeing corresp onding to arcsec This

allowed most of the light to pass down the slit op ening in typical observing conditions at

Mt John The tilt settings of the echelle and crossgrating were set using the thorium

argon wavelength reference lamp to show the various lines that corresp ond to the sp ectral

regions of the Calcium H and K lines Later the camera was fo cussed to give the maximum

sharpness corresp onding to as high a signal as p ossible in the thorium lines When

the observing was b egun on the rst night of a run the telescop e was fo cussed on a

bright star usually Car or PsA at rst visually with the eyepiece mounted on the

sp ectrograph and then using the CCD to get the maximum signal p ossible in the orders

which corresp onds to the sharp est fo cussed image of the star on the slit When observing

the Calcium H K lines with this sp ectrograph it is not p ossible to observe b oth lines

simultaneously on the same CCD image due to their p osition on the echellogram and

size of the chip It was therefore necessary to change the echelle tilt which is altered

by a micrometer on the side of the sp ectrograph Every eort was made during the

alternating b etween the two settings to have the echelle tilt return to a p osition that kept

the same sp ectral wavelengths on the CCD to a precision of a few pixels This was done

by rotating the micrometers in one direction only when moving to a particular setting to

avoid backlash in the micrometer see Table for echelle parameters

Inside Filter BG

Crossgrating Blue

Collimator

Outside Slit width

Slit p osition

Slit Height

Crossgrating y

Echelle tilt yK

Camera fo cus y

Table Settings on the MJUO echelle sp ectrograph Ab ove

are given the various micrometer and other settings for the observations

of the CaI I lines

y These represent typical initial settings

For the smo othelding of the CCD also known as sp ectral at elds a blue lter was

used to blo ck out some of the signicant amount of scattered red light from the tungsten

The lter was only used for the smo othelds

MJUO CCD system

lamp used for smo othelding This is imp ortant b ecause it is known that the resp onse

of the pixels is wavelength dep endent see g for an example of the atness of the

CCD

MJUO CCD system

The MJUO CCD system is based around the Photometrics CCD camera system

purchased in December from Photometrics Ltd of Tucson The system is run by a

Photometrics PM computer which op erates the Thomson CSF TH CDA CCD

chip see Tobin for a fuller description All commands for op erating the camera

are in the FORTH command language which has b een tailored for the op erations of the

system at MJUO see Tobin and Photometrics Users Manual for Forth Pro

gramming System The CCD chip is overcoated with a uorescent dye Metachrome

I I for enhanced ultraviolet and blue resp onse The chip is x pixels in dimension

with a pixel size of m square and is contained within a co oled dewar with liquid ni

trogen The chip is maintained at a temp erature of ab out Celsius which is the

optimum op erating temp erature for the chip for observations in all parts of the sky

At nm the CCD chip has a quantum eciency of though the individual pixels

may vary by as much as part in from the average An example of this can b e seen

in the smo otheld see Fig

Observation pro cedure

At the b eginning of each run the echelle sp ectrograph was set up for observations in

the near ultraviolet This required the installation of the blue crossgrating which is

most ecient from the UV atmospheric limit to H The tilts of the gratings were then

adjusted to appropriate regions for observations of the wavelengths corresp onding to Ca

I I The slit size was set as used in the observation runs conducted in previous years

The slit size used was approximately arcsec The camera fo cus in the sp ectrograph

was adjusted as required on examination of several thorium reference sp ectra exp osed

at slightly dierent camera settings The camera fo cus p osition that gave the b est result

was then used Then nal adjustments were made to the gratings in the sp ectrograph to

match the thorium sp ectra used to calibrate the wavelength scales of the orders p ertaining

to the Calcium II lines to a copy of an image of a previous thorium sp ectra to consistently

observe the same wavelength region For ease of reduction a copy of a CCD image of

The tungsten lamp blew during the last observation run and was replaced by a new bulb which

app ears to b e stronger than the old one in the UV part of the sp ectrum improving also the ratio of the

UV light in the sp ectral orders compared to the background of scattered red and nearinfrared light

It should b e noted that the camera fo cus did not change to any signicant degree during individual

runs But a seasonal ie presumably temp erature dep endence was observed

Chapter Observations COLUMN OFFSET: 0 ; ROW No.COLS: 384 No.ROWS: 576 BINNING: 1 GAIN: 100 IMAGE TYPE: 3 University of Canterbury Mt John University Observatory Department of Physics & Astronomy

Orlon Petterson 1306 PHYSOKP GPL Ghostscript Thomson TH7882CDA Charge-Coupled Device S/W:1995 Feb 23 1342 SMOOTH_FIELD -DRK RA= : DEC= ( ) MJUO LST=23:12:47 X= HA= h m 1372 1403 1440

500 1485 1537 1660 1956 2346 400 2741 3111 3325 3441

300 3522 3589 3661 3880 4027

200 4117 4183 4242 4303 4412 100 4650 4741 4817 4880

0 4936 0300 100200 200100 300 0

FILTER X-GRATING RECORDED ACQUISITION DATE & TIME BEGUN ENDED CORRECTED MEAN [UT] 4986 blue 1995-Nov-14 08:14:58- 08:19:03 X-GTG MICR'R INSTRUMENT DECISECONDS EXPOSURE JULIAN DAY BG39 1.521 Ech 6000 ds 600.0 s 2450035.8466 5040 TELESCOPE OBSERVER ECHELLE TILT UT-CCD ID_TICKS OFFSET

1f13 okp 375 -55.0 INTPLN 816336898 433.17 6385 ACQUISITION SOFTWARE SL/W MICR'R INPUT FILE CONVERTED TO POSTSCRIPT FILE x=0.95

MtJohn 438 T864_0005.MT -->.PS at 17-JAN-96 11:55:23 NEGATIVE

Figure K line region smo otheld showing the atness of the CCD

Observations of Beta Pictoris

the thorium lines in question taken at an earlier time were used as the reference for the

setting of the sp ectrograph see Fig for the K line region and Fig for the H line

region

The observing pro cedure followed each night was to top up the dewar at least hour

b efore any use was made of the CCD so as to allow the CCD time to adjust to the addition

of liquid to the dewar Immediately after liquid nitrogen was added to the dewar it was

observed that the bias level changed quite signicantly for some minutes Some bias

readouts were made to see if there were any ma jor changes in the bias level of the CCD

that indicated the chip had not reached equilibrium Then dark frames of seconds

duration that were selected to b e free of cosmic ray strikes were acquired and used

to make a median dark frame for subtraction from all subsequent CCD images taken If

there was a change in the level of dark current during the night as indicated by a warning

from the acquisition software a new median dark was made Next the sp ectral atelds

or smo othelds were made by exp osing a tungsten light source through the slit of the

sp ectrograph A blue lter was placed b ehind the slit to reduce the amount of scattered

red light in the sp ectrograph and a minute exp osure was made with the CCD Typically

smo othelds were made at each of the two sp ectrograph settings which were then added

together in the reduction in order to reduce the magnitude of the systematic error in the

resp onse of the pixels hit by cosmicrays Median clipping of cosmicray strikes was not

p ossible b ecause sp ectral output of the tungsten lamp may change with time altering

the sp ectral shap e of the smo othelds which cannot b e scaled to identical mean signal

levels at all p oints as required for median clipping Shortly after sunset the dome shutter

was op ened to allow the dome and telescop e to reach the ambient outside temp erature at

this time also the covers were rolled back from the primary mirror and the baes Once

the smo othelding was completed the telescop e drives were activated and the telescop e

p ointed at the rst ob ject After every exp osure of a star a thoriumargon wavelength

calibration image was made to calibrate the wavelength scale for that image Once this

was done another stellar sp ectrum could b e taken or the telescop e moved on to the next

star After each image was taken it was then saved on to tap e Each H or K line sp ectrum

was also recorded on pap er for immediate examination

Observations of Beta Pictoris

Being a circump olar star at the MJUO latitude Beta Pictoris could b e followed all night

However in order to obtain go o d sp ectra with high SN in a reasonable time observations

were made whenever the star was b elow the preselected limit of airmasses of extinction

This would keep observations to the prop osed minute exp osure length required for

appropriate temp oral sampling It was found in practice that Beta Pictoris was b est

observed once it was ab ove airmasses to keep to the minute exp osure duration It

Chapter Observations COLUMN OFFSET: 0 ; ROW No.COLS: 384 No.ROWS: 576 BINNING: 1 GAIN: 100 IMAGE TYPE: 3 University of Canterbury Mt John University Observatory Department of Physics & Astronomy

Orlon Petterson 148 PHYSOKP GPL Ghostscript Thomson TH7882CDA Charge-Coupled Device S/W:1995 Feb 23 155 Th_arc for beta_Pic -DRK RA= : DEC= ( ) MJUO LST=07:04:41 X= HA= h m 160 164 168

500 173 177 182 187 192 400 197 202 208 214

300 221 231 242 257 274

200 295 320 355 400 461 100 542 656 822 1059

0 1450 0300 100200 200100 300 0

FILTER X-GRATING RECORDED ACQUISITION DATE & TIME BEGUN ENDED CORRECTED MEAN [UT] 2566 blue 1994-Nov-24 15:28:35- 15:29:23 X-GTG MICR'R INSTRUMENT DECISECONDS EXPOSURE JULIAN DAY None 1.520 Ech 1000 ds 100.0 s 2449681.1454 5747 TELESCOPE OBSERVER ECHELLE TILT UT-CCD ID_TICKS OFFSET

1f13 W< 376 -1.7 INTPLN 785690915 417.86 15979 ACQUISITION SOFTWARE SL/W MICR'R INPUT FILE CONVERTED TO POSTSCRIPT FILE x=0.95

MtJohn 438 T501_0035.MT -->.PS at 17-JAN-96 11:54:56 NEGATIVE

Figure Thoriumargon lines for CaI I K line region

Observations of Beta Pictoris COLUMN OFFSET: 0 ; ROW No.COLS: 384 No.ROWS: 576 BINNING: 1 GAIN: 100 IMAGE TYPE: 3 University of Canterbury Mt John University Observatory Department of Physics & Astronomy

Orlon Petterson 155 PHYSOKP GPL Ghostscript Thomson TH7882CDA Charge-Coupled Device S/W:1995 Feb 23 162 Th_arc for beta_Pic -DRK RA= : DEC= ( ) MJUO LST=04:16:32 X= HA= h m 168 173 178

500 183 189 195 201 208 400 216 225 234 244

300 256 268 282 297 315

200 337 364 397 444 508 100 612 756 968 1289

0 1915 0300 100200 200100 300 0

FILTER X-GRATING RECORDED ACQUISITION DATE & TIME BEGUN ENDED CORRECTED MEAN [UT] 3209 blue 1994-Nov-24 12:40:53- 12:41:41 X-GTG MICR'R INSTRUMENT DECISECONDS EXPOSURE JULIAN DAY None 1.520 Ech 1000 ds 100.0 s 2449681.0290 8131 TELESCOPE OBSERVER ECHELLE TILT UT-CCD ID_TICKS OFFSET

1f13 W< 475 -1.7 INTPLN 785680853 421.27 15977 ACQUISITION SOFTWARE SL/W MICR'R INPUT FILE CONVERTED TO POSTSCRIPT FILE x=0.95

MtJohn 438 T501_0024.MT -->.PS at 17-JAN-96 11:54:08 NEGATIVE

Figure Thoriumargon lines for CaI I H line region

Chapter Observations

was not always p ossible to keep to those constraints so when the seeing was bad or

observing was conducted through cloud go o d signals were deemed more imp ortant than

time resolution Sp ectra were taken of b oth the Calcium II H K lines usually alternating

b etween the two eg K H K H K In order to determine the b est times to observe the

programme stars a small pro cedure was written for MIDAS to calculate the airmass for a

star at Mt John Observatory for any time of the year and night see Fig The plots

were only used as a guide to observing times at the observatory The plots in themselves

are not meant to b e very accurate and have p oints plotted that represent at what times of

night the stars are at certain airmasses Airmasses were calculated for every minutes of

the year and the times that were within certain limits ie airmasses were plotted

as representing the airmasses which explains why there are sometimes more than one line

of symbols corresp onding to a particular airmass

Figure Plot of airmass vs time of year and time of night

for Beta Pictoris The ab ove plot shows the time of night for a given

time of the year and time of night when Beta Pictoris is at airmasses of

extinction or less Observations were planned from this plot in order to

make observations that record sucient signal for a reasonable exp osure

time This plot was used to plan observations that would record sucient

signal for a reasonable exp osure time

It should b e noted that the b est time for observing Beta Pictoris is during the southern

summer from ab out Octob er to April when it is at low airmass through most of the night

Observations of Ophiuchus

Ophiuchus was the only binary system observed and b ecause of its p osition near the

celestial equator was observed whenever it was b elow airmasses Because of its p o

sition in the northern sky at MJUO and though it is only half a magnitude fainter than

Beta Pictoris the typical exp osure times were of the order of an hour The absorption

features in the K line sp ectra are only very slight so it was necessary to get high sig

nal sp ectra so that the noise was very much smaller than the absorption features b eing

observed Observations were made mainly of the K line as it was decided early on that

unless something signicant was happ ening in the K line the H line would not b e ob

served It should b e noted that the H line sp ectrum of Ophiuchus shows none of the

circumstellar absorption features visible in the K line

Figure Plot of airmass vs time of year and time of night for

Ophiuchus The ab ove plot shows the time of night for a given time

of the year and time of night when Ophiuchus is at airmasses of

extinction or less Observations were planned from this plot in order to

make observations that record sucient signal for a reasonable exp osure

time This plot was used to plan observations that would record sucient

signal for a reasonable exp osure time

The b est time to observe Ophiuchus is during the months of May to September

when it is highest in the northern sky see Fig

Chapter Observations

Observations of Alpha Pisces Australus

Formulhaut or PsA is the brightest of the stars at V and is well placed

in the sky where it passes near the zenith at MJUO PsA was also observed when it

was b elow the airmass limit usually b etween observations of Oph and Pic see

Fig As a result of when it was observed most observations were at low airmasses

Observations were again almost all of the K line after b eing decided that the H

line could b e ignored unless something happ ened in the K line sp ectra that warranted

further investigation at H

Figure Plot of airmass vs time of year and time of night for

Alpha Pisces Australus The ab ove plot shows the time of night for

a given time of the year when Formulhaut is at airmasses of extinction

or less Observations were planned from this plot in order to make

observations that record sucient signal for a reasonable exp osure time

This plot was used to plan observations that would record sucient

signal for a reasonable exp osure time

Due to its brightness PsA is ideal for fo cusing the telescop e on the slit and max

imising the throughput to the CCD Also note that in Fig an error o ccurred in the

routine causing the triangular region of circles in the plot

Chapter

Data reduction and analysis pro cedures

The tap es containing the data taken at MJUO with the CCD were read onto disk in

Christchurch where the data could then b e reduced using the reduction program of choice

The Europ ean Southern Observatory Munich Image Data Analysis System ESOMIDAS

or simply MIDAS is the computing package used in the reduction and analysis of all the

data contained in this thesis

Munich Image Data Analysis System

MIDAS is essentially a set of general purp ose programs for the reduction and analysis

of astronomical data which has b een obtained from various detectors or sources The

MIDAS system can b e run b oth in interactive and batch mo des MIDAS is available in

b oth UNIX and VAXVMS versions It can b e adapted to suit the requirements for all

sorts of image pro cessing and data reduction It is also designed to allow the integration

of future application mo dules This is achieved by the use of pro cedure les which can

b e created by the user to p erform a simple or complex sequence of MIDAS commands

The pro cedure les are programmed in the MIDAS command language which consists of

a control language and set of commands In addition contextspecic commands can b e

added from the various MIDAS application mo dules MIDAS is run on the University of

Canterburys Computer Services Centre Sparc Centre known as cantua The

earlier versions of MIDAS were run on the Universitys VAX computer

Data reduction was undertaken using ESOMIDAS NOV and later NOV versions

of this package for the reductions of observations of Beta Pictoris acquired in earlier years

and only the NOV version for the rest of the observations made for this thesis

The MIDAS echelle reduction package

The reduction of the raw CCD data b egins with the starting up of a MIDAS session

The data on disk is converted from the FITS format to the idiosyncratic format used

by MIDAS Due to the orientation in which the CCD is mounted on the echelle sp ectro

graph it is necessary to rotate the images clo ckwise by into the orientation required

by MIDAS for sp ectral reductions that is orders increasing from top to b ottom and

wavelength increasing from left to right

Chapter Data reduction and analysis pro cedures

Next the echelle application package is enabled allowing the use of echelle sp ecic

commands The reduction pro cedure was as follows

The orders are lo cated from the smo otheld image and also their orientation as

recorded on the CCD The orders were found using either the DEFINEECHELLE com

mand for the Calcium K line region or the HOUGHECHELLE command for the Calcium

H line

The oset b etween the centres of the smo otheld orders and the stellar orders are

then found A gaussian is tted to one of the stellar orders and the dierence b etween

the centre of the stellar order and the corresp onding smo otheld order is recorded The

full width of the stellar order is also recorded This is imp ortant for determining where

on the image the dierent types of ltering are applied

Next the thoriumargon sp ectrum for the stellar image b eing reduced is tted by a

p olynomial that b ests ts the wavelengths of lines stored in a reference table to the lines

detected in the image The wavelengths reference table is a subset of one of the MIDAS

tables augmented with lines from the Atlas of the Thorium Sp ectrum The resulting

p olynomial t is applied to the extracted orders of the stellar sp ectrum to obtain sp ectra

plotted on a wavelength scale

Probably the most dicult and sensitive part of the reduction pro cess is the ltering

applied to the stellar sp ectral image In the removal of the cosmicray strikes on the

image two dierent types of ltering are necessary A median lter is applied to the

background b etween the orders to remove cosmicray strikes and smo oth the background

A cosmicray lter is applied in the orders which shouldnt aect the sp ectrum which

uses CCD information such as the readoutnoise in e inverse gain factor e ADU

and threshold value for evaluating spikes for cosmicray removal in the sp ectral orders

It works by tting a onedimensional p oint spread function PSF p erp endicular to the

disp ersion direction and replacing pixels that have large deviations from the PSF

Finally the ltered image is divided through by the smo otheld to remove pixelto

pixel sensitivity variations across the image and then recorded as dimensional sp ectra

Before this division is done the smo otheld image is scaled so that the signal levels in

the orders are close to unity The signal levels in the nal reduced sp ectra are then still

reasonably indicative of the total number of ADU in the original raw images The pro cess

followed is describ ed more fully in App endix A which describ es the general pro cedure for

reducing sp ectra recorded on the PM CCD system

The resulting sp ectra were then plotted out and overplotted with the background

subtracted order which should contain all the cosmic rays The plots are examined to

identify absorption features that are real and those that might b e due to overcorrection

for the removal of a cosmic ray strike in the order see Fig

The ab ove reduction pro cedure was then written in the MIDAS command language to semiautomate

the pro cess for sp eed and ease of use The program is named echorlon and is found in App endix B

Determination of resolving p ower

Figure A reduced K line sp ectrum with its corresp onding cosmic

ray sp ectrum which is used for conrming the reality of any variable

absorption features

Mo dication to MIDAS

Due to the unique combination of the echelle sp ectrograph and CCD found at MJUO it

is necessary to mo dify the MIDAS software The echelle orders are quite wide some

pixels due to the original design of the sp ectrograph b eing for photographic sp ectroscopy

Also the small size of the CCD chip means that less than orders are usually recorded

less than MIDAS was programmed for In the UV only a maximum of orders can b e

recorded on the chip

Because of the wide orders it is necessary to mo dify the cosmicray ltering routines

to remove cosmicrays more eciently The wavelength calibration routine needed to b e

mo died to work with as little as orders

Determination of resolving p ower

The determination of the resolving p ower of the sp ectra from the echelle sp ectrograph is

an imp ortant step in the reduction pro cess The resolving p ower at the two settings of

the echelle grating tilt is imp ortant to know as observations of the Ca II lines of H

K o ccur in dierent orders of the echellogram One would exp ect the dierence in the

FWHM to b e

or der

or der H

Chapter Data reduction and analysis pro cedures

When the thorium calibration frames were examined and then gaussians tted to the

emission lines it was found that the b est ts o ccurred for emission lines of intermediate

strength From these ts the average for those lines FWHM measured in the sp ectra

were taken and then an average of all the averages of the sp ectra to give the overall

average of the FWHM for the regions of the Calcium II H K lines The result of those

measurements are

The FWHM for the region of the K line A

which gives a resolving p ower of

or ab out kms for the clear separation of features in the K line

The FWHM for the region of the H line A

which gives a resolving p ower of

or ab out kms for the clear separation of features in the H line

The dierences in the resolving p owers at H K ab out are most likely due to

the small change in fo cus b etween the regions in the echellogram

Chapter

Presentation of Sp ectra

In this chapter the Calcium II H K sp ectra for all the stars are presented as a series of

graphical plots in which the sp ectra are stacked on top of one another in time order Any

changes in the sp ectra should b e readily apparent if the additional absorption is strong

enough to b e seen

The Beta Pictoris H & K lines

The sp ectra of Pictoris are presented here in order of time with b oth H K sp ectra

plotted next to each other where available The time of observation can b e estimated by

nding the time from the p oint of the circumstellar line in each sp ectrum The reference

sp ectrum for the H and K lines see Chapter are included for comparison and to aid

in identifying additional absorption features in the sp ectra

The observing campaign

Presented in this section are the stacked plots for the Pictoris International Observing

Campaign For the analysis of these sp ectra see Section

Chapter Presentation of Sp ectra

Figure Sp ectra taken on Dec

The Beta Pictoris H K lines

Figure Sp ectra taken on Dec

Chapter Presentation of Sp ectra

The observing campaign

Presented in this section are the stacked plots for the Pictoris International Observing

Campaign Also included are the follow up observation to the campaign on

January and also the sp ectra collected during the runs used for gaining exp erience

in observing Pictoris in b oth H and K For the analysis of these sp ectra see Section

Figure Sp ectra taken on Jan

The Beta Pictoris H K lines

Figure Sp ectra taken on Oct

Figure Sp ectra taken on Apr

Figure Sp ectra taken on Nov

Figure Sp ectra taken on Oct

Chapter Presentation of Sp ectra

Figure Sp ectra taken on Nov

Figure Sp ectra taken on Dec

The Beta Pictoris H K lines

The observing campaign

Presented in this section are the stacked plots for the Pictoris International Observing

Campaign For the analysis of these sp ectra see Section

Figure Sp ectra taken on Nov

Chapter Presentation of Sp ectra

Figure Sp ectra taken on Nov

Figure Sp ectra taken on Dec

Figure Sp ectra taken on Nov

The Beta Pictoris H K lines

Figure Sp ectra taken on Nov

The observations

Presented in this section are the stacked plots for Pictoris taken during for this

thesis by the author For the analysis of these sp ectra see Section

Chapter Presentation of Sp ectra

Figure Sp ectra taken on May

Figure Sp ectra taken on Apr

Figure Sp ectra taken on May Figure Sp ectra taken on Jun

The Beta Pictoris H K lines

Figure Sp ectra taken on Jun

Chapter Presentation of Sp ectra

Figure Sp ectra taken on Jul

The Beta Pictoris H K lines

Figure Sp ectra taken on Aug

Figure Sp ectra taken on Jul

Figure Sp ectra taken on Sep

Chapter Presentation of Sp ectra

Figure Sp ectra taken on Sep

The Beta Pictoris H K lines

Figure Sp ectra taken on Oct

Figure Sp ectra taken on Nov

Figure Sp ectra taken on Oct

Chapter Presentation of Sp ectra

Figure Sp ectra taken on Nov

The Beta Pictoris H K lines

Figure Sp ectra taken on Nov

Chapter Presentation of Sp ectra

The Ophiuchus K line

In this section the sp ectra taken of Ophiuchus are presented for examination Note

that the sp ectra without the small absorption features are H line sp ectra

The K line sp ectra of Ophiuchus is dominated by the smo oth photospheric prole

onto which are sup erimp osed narrow absorption lines due to interstellar clouds contain

ing Calcium II at and kms Helio centric velocity The H line sp ectra show no sign

of these absorption features so any additional features strong enough to b e detected in H

should corresp ond to infalling matter in the star system

At the available noise levels there are no clearlyapparent circumstellar absorption

features The two additional absorptions seen in the K line sp ectra obtained on Octob er

and are due to dust on the window in front of the CCD detector and are not evidence

of variable absorption features see Fig

Figure H K line sp ectra of Ophiuchus taken from MJUO in

May and June

The Ophiuchus K line

Figure H K line sp ectra of Ophiuchus taken from MJUO in

June and July

Figure K line sp ectra of Ophiuchus taken from MJUO in July

August and September

Chapter Presentation of Sp ectra

Figure K line sp ectra of Ophiuchus taken from MJUO in

September

Figure K line sp ectra of Ophiuchus taken from MJUO in

September and Octob er

The Alpha Piscis Australus H and K lines

The sp ectra of PsA shows only the photospheric prole but since the susp ected dust

disk is b elieved to b e seen in the plane of the sky it is unlikely any FEB in orbit around the

star would cross the line of sight given this orientation No variable absorption features

are apparent in any of the sp ectra

Figure H K line sp ectra of Piscis Australus taken in May

and June

Chapter Presentation of Sp ectra

Figure K line sp ectra of Piscis Australus taken in June and

July

Figure K line sp ectra of Piscis Australus taken in July August

and September

The Alpha Piscis Australus H and K lines

Figure K line sp ectra of Piscis Australus taken in September

Figure K line sp ectra of Piscis Australus in Octob er

Chapter

Characterisation of Pictoris sp ectra

Normalising the sp ectra

Normalising the data to a common intensity scale is an imp ortant step along the path to

extracting useful information

Each of the Ca II H K lines was plotted at the same wavelength scale and the ux

level was scaled so that the plots could b e overlayed for comparison Then the CaI I H

K line sp ectra were divided through by their resp ective reference sp ectra

Each reference sp ectrum was made from sp ectra of each star taken at MJUO using the

same equipment The sp ectra selected had to b e as free as p ossible of additional absorption

features if present then the sp ectra were smo othed using the MIDAS FILTERGAUSS

command The adopted H and K reference sp ectra are shown in Figs It was

decided that the b est way to normalise the sp ectra was to align the shortwavelength side

of the circumstellar line of each observed sp ectrum with the corresp onding feature in the

reference sp ectrum and then divide one by the other The resulting division should then

reveal the presence of absorption features by any departure from a horizontal straight line

Of course the circumstellar line is b elieved to b e saturated therefore the quality of the

background subtraction during the reductions aects the line depth Measurements near

this line are exp ected to b e more uncertain The example of Figs and illustrate

the pro cedure In Fig the observed sp ectrum of Fig has b een aligned with and

divided by the reference sp ectrum

In the particular case of the Beta Pictoris reference sp ectra the smo othed prole was

made from sp ectra taken in previous years and were nearly completely free of additional

absorption features The reference sp ectra should only have the photospheric lines and the

circumstellar feature present Since observations of Beta Pictoris b egan at MJUO there

has never b een a case where no additional features are present in go o d signal sp ectra It

was therefore necessary to adjust the smo othed sp ectrum after a few trials at normalising

other sp ectra from dierent observing runs turned up some common absorption features

which were then removed The normalised sp ectra are converted to a velocity scale

centred on the circumstellar line see Fig The MIDAS pro cedure which do es the

normalisation is found in App endix B

Fitting the absorption features

Figure K line sp ectra taken on June showing the broadened

circumstellar line and also a clear HVF

Fitting the absorption features

With the normalised sp ectra the existence of any absorption features can then b e tted

with gaussians to b est t the absorption features attributed to the cometary b o dies The

continuum is then tted by either of a constant straight line a line with a slop e in

wavelength or a quadratic in wavelength whichever app ears to t the continuum b est

The absorption features are tted using the MIDAS nonlinear tting routines The cursor

was used to mark on the plot the approximate centre and depth of each feature and also

the FWHM These initial values were then fed to the tting routine which was used the

following equation to compute each gaussian shap ed feature

x b

GAU S S x a b c a exp ln

This pro cess was rep eated for each of the features in the plot The Mo diedGauss

Newton metho d was used in the tting From the initial input from the cursor the values

of ab and c were iterated many times to pro duce an optimum synthetic t from which

one can then extract velocity information ab out the cometary absorptions as well as their

strengths and FWHMs see Fig The MIDAS pro cedure which do es the tting is

listed in App endix B

Chapter Characterisation of Pictoris sp ectra

Figure The normalised sp ectrum after b eing divided by the refer

ence sp ectrum

Figure Plot showing the normalised sp ectrum and tted gaus

sians to the variable absorption features Note the well dened gaussian

shap ed absorption at kms

Chapter

Analysis

Presented here are the results of the sp ectroscopic observations taken of Beta Pictoris

from MJUO during the international observing campaigns and the ob

servations made in The data is presented as a series of plots Figs

showing the velocity FWHM depth and equivalent width These data should b e con

sidered conjointly with the plots in chapter gures The symbols used for the

absorption features identify what range of velocity they are in There are a maximum of

velocity regions that can b e plotted usually the redshifted LVF region and then either the

blueshifted and redshifted HVFs or ranges of HVFs in the red Op en symbols represent

the data derived from the sp ectra taken of the Calcium K line and the lled symbols rep

resent the data derived from the sp ectra taken of the Calcium H line The same symbols

are used for the same velocity range from day to day during the same observing run The

data on the variable absorption features seen in the sp ectra of Pictoris is also displayed

in a table in App endix C

The observing campaign

Only observations of the Calcium K line were made in December The most

remarkable feature of the observations Figs is the presence of the long lived

absorption feature at kms In fact it is observed from December to December

and even app ears in a follow up observation on January It should b e noted that

the detection on the January has b een badly eected by a cosmic ray as can b e seen in

Fig App earing from time to time are slightly higher redshifted features at to

kms which app ear out of the red wing of the kms feature see Figs

This indicates that the infalling material is clumpy in structure which supp orts

the conclusions of LagrangeHenri et al Even higher HVFs are observed which

are generally broader with the centroid evolving in velocity see Figs

Also detected is the presence of blueshifted absorption features Figs

not detected by LagrangeHenri et al This is not surprising given the

For some unknown reason the lower error bars in the FWHM and Equivalent Width data plotted

b efore each Helio centric Julian Day is misplotted at a very low value This is clearly an error in the

MIDAS NOV plot command overploterror when a logarithmic scale is in use

Chapter Analysis

metho d of division by an unabsorb ed stellar prole followed here which allows for the

detection of broad shallow and some blended absorption features All the blueshifted

features were short lived typically hrs in duration they app ear and disapp eared

suddenly hr and once present show little evolution in strength of the features

app eared at low velocity while the other was a broad shallow absorption feature at high

velocity Fig much as exp ected for the observation of the ion tail of an evaporating

b o dy close to the star

The redshifted features at high velocity and the ones emerging from the kms

absorption feature app ear and then disapp ear over timescales of ab out hours which

corresp ond to the crossing times of orbiting b o dies across the stellar disk On the night

of December the feature seen at kms shows evidence suggestive of the eect

of stellar rotation on the evolution of an FEB features equivalent width The decrease

of equivalent width is slow suggestive of an FEB crossing from the blueshifted side to

the redshifted side of the stellar disk However the rate of decline may b e considerably

dierent if errors are underestimated and the apparent deceleration of the feature is

unexp ected Could it b e due to the FEB encountering some dense medium

In calculating the number of events observed during an observation run and then giving

a yearly estimate of infall a number of problems need to b e addressed The timescale from

which the estimate will b e made from can have a great deal of impact on the result Do es

one consider only the time observed and hence only count those features seen to app ear

or the days on which observations are made and then count what app ear to b e individual

features regardless of them b eing already present and those new ones that subsequently

app ear Of course if one lo oks at just the velocity a feature is seen to app ear at one could

miss the signs of or more features b eing present from examination of the equivalent

width of the feature With all these problems in mind it was decided that the estimate

of the number of infalling b o dies would b e calculated from the b o dies that app ear in the

velocity plots Also the equivalent width of the features would b e examined for strong

evidence of additional b o dies b eing present The long lived features would only count as

b o dy despite the b elief that they are due to more than as there is no measurable

evidence of additional features in this data The time over which these features would

b e counted is the days of observation since some of the features observed would app ear

during the day b efore observing b egan and disapp eared after observing ended for the

night Hence the resulting estimate should give a lower b ound to the number of infalling

b o dies that could b e observed crossing the line of sight

A count of the features observed gives events detected over days b oth red and

blueshifted plus the LVF at kms An estimate of the infalling rate gives events

observed p er year

The observing campaign

Figure December

Chapter Analysis

Figure December

The observing campaign

Figure January

Figure December

The observing campaign

Before the commencement of the campaign a few nights were used in April and

Octob er to optimise observing pro cedures and gain exp erience with altering the sp ectro

graph set up so that observations of b oth the H and K lines could b e made On April

an absorption feature is seen in the wing of the circumstellar feature also a broad feature

is present blended with the other feature

On Octob er features are seen at low redshift one deep and distinct and a shallower

feature b etween it and the circumstellar line Both however are gone in the following

nights observations see Figs and which show no discernible absorption fea

tures at all except for the app earance of a broad HVF in the last H sp ectrum

During the international observing campaign in November and December Figs

a long lived shallow blueshifted feature is observed in the K line all week long

but only conrmed in the H on November It is however clearly seen in some of the

original sp ectra most prominently on November and see Figs and The

lack of detection in H may b e entirely due to the dierence in the oscillator strengths of

H K At b oth ends of the run redshifted features app ear in the sp ectra at low velocity

predominately also many broad features are seen some even show suggestive evidence of

evolution in equivalent width On December and the evolution of equivalent width

Chapter Analysis

in features at kms is again suggestive of FEB passage from the blueshifted to

redshifted sides of the stellar disk

A count of the observed events indicates seen over days giving the infalling rate

at p er year assuming of course that the sample is typical for the whole year

Figure Octob er

Figure April

The observing campaign

Figure November

Figure Octob er

Figure November

Chapter Analysis

Figure December

The observing campaign

The campaign reveals lots of varying features which come and go Figs

On November a broad LVF is seen at kms in b oth H and K but has disapp eared

after only hours of b eing rst observed From the November a feature develops

and is seen to start evolving initially starting as a broad absorption in the red wing of the

photospheric line at kms It then develops into a clear and unmistakable feature in

the Ca II lines where it is seen in b oth H and K as a broad and deep feature on the

November It is p ossible there is more than one FEB present and this maybe part of a

shower of b o dies It then evolves into a broader feature on the November and on the

November near the end of the night it is seen to separate into features From this

and from the plots of equivalent width it is clear that this feature is evolving in a quite

complicated fashion which suggests that more than one FEB is involved

The circumstellar line is further broadened in its red wing indicating another feature

sup erimp osed typical of the features seen at low velocity from b o dies orbiting at quite

large distances from the star AU In the few nights it is observed it is seen to come

and go but is clearly seen from the November It app ears to b e remarkably stable and

is not saturated indicated by the dierence in the equivalent width of the H and K lines

The longterm stability is reminiscent of the similar phenomenon observed in

The observing campaign

For a brief p erio d of time on November a broad blueshifted feature is seen for ab out

hour which app ears suddenly and disapp ears just as suddenly

On the November a LVF at kms b ecomes broader as the night progresses and

is still present the next night It has clearly evolved over this time Also on November

broad HVFs are seen to come and go quickly seen in one sp ectra and gone the next in

sp ectra taken hour apart

Finally a count of events yields an infall rate of events p er year from the

events seen over the days observed This will b e greater if the longlived features are

due to more than one b o dy

Figure November

Chapter Analysis

Figure November

observations Beta Pictoris

Figure November

observations Beta Pictoris

During this year the bulk of the data was collected on Pictoris in week and week

runs Initially there were short observation runs to check instrumental settings gain

exp erience and check for p otential problems The sp ectra are summarised in gures

On April Pictoris showed a number of features a broad blueshifted feature at

kms and also some additional absorption in b oth wings of the circumstellar line

In the second initial observation run in May a LVF is seen at kms on May and is

seen to grow in equivalent width as it gets broader On the following night a feature is seen

initially at kms which then splits into comp onents in b oth H and K Examination

of the equivalent widths show it is entirely p ossible that this feature is a continuation

of the previous nights feature Also seen on May is a clear blueshifted feature at

kms which clearly changes in equivalent width

From June a deep absorption feature is present at kms which exhibits little

change throughout the run to June Of particular note is a feature observed on June

at kms as a deep and given its high velocity unexp ectedly narrow absorption

feature It app ears in only one sp ectra K giving it a timescale of less than hour as it is

not observed in either of the sp ectra taken b efore or after in H There can b e no doubt

Chapter Analysis

as to its reality Examination of the CCD image shows an absorption line right across

the width of the sp ectrum Such a feature is dicult to explain as its velocity indicates

it is very close to the star rough calculations estimate it to b e less than stellar radius

from the surface of the star but it may b e explained by the FEB scenario explanation

for LVFs with the axis of the parab olic orbit at low inclination to the line of sight see

Beust et al At the same time a broad HVF typical of a FEB close to the star

is present during the time of observation at kms which clearly changes in equivalent

width

The June sp ectra reveal further blueshifted activity in the one high signal sp ectrum

In the last and particularly noisy sp ectrum a p ossible broad feature at kms app ears

On June a broad and distinctive feature has app eared at kms which is still

visible the next night It has weakened somewhat by the b eginning of the night of

June during which it changes only slightly in equivalent width This is the inverse of the

b ehaviour of the other relatively stable feature present at kms A broad HVF is seen

at kms in the rst half of the night of June but is gone in the later half when

observations of Pictoris resume The plot of equivalent width reveals it clearly evolves

while it was observed with a slower fall than rise suggestive of an FEB crossing the stellar

disk from blueshifted to redshifted side

In late July a feature at kms is still observed from the July and though not

as strong it may b e a continuation of the previous runs feature at that velocity Also

the feature observed at kms is reminiscent of the previous run but clearly varies in

equivalent width July is lled with activity in addition to the previously mentioned

features a broad HVF at kms is seen which seems to also change velocity A HVF at

kms is briey seen in one sp ectrum but unlike that on June is broad A blueshifted

feature at kms is seen to clearly change in equivalent width Interestingly enough

it is observed in H twice b efore it is seen in K During the night of July the feature at

kms p ersists A blue feature app ears and rapidly strengthens towards dawn

On July the kms feature is unchanged the kms feature shows some inter

esting dierences in H and K in that the H absorption starts considerably weaker than

the K ones but strengthens towards dawn see Fig A HVF is seen at kms

which is broad and changes in equivalent width only slightly At kms a broad

and dicult to detect feature is seen intermittently in some sp ectra but do es double in

equivalent width The feature at kms b ecomes more distinctive on July A fur

ther broad HVF at kms app ears Another blueshifted feature at kms is seen

changing in strength

The nal night of the run Fig reveals the feature at kms as shallower than

the previous night Further blueshifted activity is also present The blue features are

clearly changing and the broad HVF at kms is only visible for half an hour

Again on September a feature is seen at kms which is only a little weaker than the

observations Beta Pictoris

previous run Is this the same feature seen in June and July If so the FEB scenario will

have great diculty explaining such a very longlived phenomenon but the existence of

such a longlived low velocity absorption might hold clues to the origin of the circumstellar

feature the p ersistence of which is also problematic Further blueshifted activity is seen

in a changing feature at kms The feature at kms is still present until

September but has disapp eared two nights later There is continuing activity on the blue

side of the circumstellar line which varies signicantly in equivalent width but is constant

in velocity It remains a clearly seen feature throughout the observation run in September

till the th when it quite suddenly changes in equivalent width and then disapp ears On

September the features at kms and in the blue p ersists from the previous night

but several HVFs are seen in the red They are seen to app ear and strengthen and then

the feature at kms disapp ears The HVFs app ear in the H sp ectra rst and vary

rapidly On September a feature is easily seen at kms which is broad and slowly

changes in equivalent width during the night Another one at kms app ears briey

changes rapidly in equivalent width and then disapp ears all in just over hours Both

HVFs indicate in their equivalent width that they are crossing from the blueshifted to the

redshifted sides of the stellar disk with dierent crossing times On September some

new HVFs are observed One feature is seen changing velocity from to kms and

staying there for the rest of the night Up on examination of the equivalent width it is

clear b o dies make up this feature the original seen to change velocity and the second

one that remains at a steady velocity A further HVF app ears at kms in the last

sp ectrum of the night The feature at kms remains unchanged in equivalent width

during the night while the blueshifted features show some variation On September

there are the blueshifted and kms features and a broad LVF at kms which

app ears and a single HVF at kms in one H sp ectrum On September the kms

feature is gone and the blueshifted feature is seen to disapp ear The only other activity

is the broad feature at kms

In Octob er another feature is seen at kms which is quite strong and changes little

in equivalent width on Octob er Further blue activity is seen in the form of a broad

absorption feature which is stronger in H than in K Another broad HVF at kms

is seen to slowly change in equivalent width during the night hours Further broad

absorption is seen in K at lower velocity late in the night The second night of observations

Octob er reveals little change with the kms still present and a broad feature at

kms which can only b e seen in the high signal sp ectra The nal night of observations

Octob er shows the kms feature to b e getting weaker as the night progresses while

in contrast the blueshifted feature strongly evolves and is even stronger in H than it is in

K For a brief p erio d of time broad HVFs are seen hours at and kms

November was last time data was collected and as b efore it shows a wide range of

activity During the entire run an additional absorption is found in the wing of the

Chapter Analysis

circumstellar line at kms and slowly evolves over the run from a dicult to detect

feature to a prominent absorption On November it is seen to split into features one

broader than the other The broader feature is missing when observations b egin on the

night of the November but the remaining line is still present On the November it has

b ecome a clear feature in b oth Calcium sp ectra and slowly changes in depth though the

rest of the run During the run features app ear and disapp ear mostly they are broad and

can b e easily seen in the sp ectra Some of the HVFs are dicult to detect due to b eing

so broad and their typically short timescales of ab out a few hours at most A number of

blueshifted features are seen all of which are broad and app ear in H more than K

In general the features all follow general trends which can b e in most cases quite well

explained by the FEB scenario The dierences in distance to the star p osition over the

stellar disk and the inclination of the line of sight all t well b etween observation and

theory

Some separate features were identied over the nights of observation which

yields an estimated infall rate of events p er year If the longlived features are made

up of many b o dies falling at the correct angles the actual infall numbers could of course

b e much higher

Figure May

Figure April

observations Beta Pictoris

Figure June

Figure May

Figure June

Chapter Analysis

Figure June

Figure July

observations Beta Pictoris

Figure July

Figure August

Figure July

Chapter Analysis

Figure September

observations Beta Pictoris

Figure September

Figure Octob er

Chapter Analysis

Figure November

Figure Octob er

Figure November

observations Beta Pictoris

Figure November

Chapter Analysis

Figure November

Summary of observational results

Of the general types of absorption features found in the Calcium II H K lines see

Sec the results of the observations from MJUO agree with there b eing general

types of features

a circumstellar feature at the radial velocity of the star

LVFs which tend to feature deep and narrow absorption lines which can last from

hours to months Also broad and shallow features app ear which last for shorter

timescales

HVFs which vary rapidly in timescales of hours and are almost invariably broad

and shallow

These descriptions certainly describ e the redshifted features adequately however the

blueshifted features tend to b e broad and shallow with short timescales Occasionally

however the blueshifted features while broad and shallow app ear unchanged over many

days b efore disapp earing as suddenly as they app eared It also should b e noted that

the circumstellar feature has exhibited variations in its FWHM either due to LVFs at

kms or changes in the circumstellar feature itself However measurements of this

variation fall outside the scop e of this thesis

Summary of observational results

Now the requirements for testing the FEB scenario will b e addressed from the p oints

raised in section

The o ccurrence of the variable absorption features dep ends on the distance they can b e

found in around the star The deep est and narrower lines are found at low velocity to

kms see Fig and and have FWHM from mA In their pap er

analysing the data only LagrangeHenri et al found a strong correlation b etween

velocity and FWHM see Fig and this is conrmed by our metho d of parameterising

absorptions see Fig The larger data set presented here shows the correlation is

weaker than that found by LagrangeHenri et al In the blueshifted absorptions a strong

correlation b etween velocity and FWHM is found However this result should b e treated

with caution in case the exp erience of the redshifted correlation turning out to b e weaker

as more data is added proves also to b e the case for the blueshifted absorptions

Figure Plot of the MJUO data for showing the strong correlation

in FWHM vs velocity as found in LagrangeHenri et al More data is

present in this plot than LagrangeHenri et al found

The deep absorption features in general are very long lived and the shallower absorp

tion features tend in general to last for approximately the crossing time of an FEB across

the stellar disk A quick count of the features nds ab out of all features seen are the

deep features which typically last many days at least However of all the features seen

ab out last long enough that more than FEB must b e resp onsible for its presence

From this data some conclusions can b e made

The deep features which are b elieved to b e due to groups account for only half the

group events observed Since these LVFs are supp osed to o ccur at large distances

Chapter Analysis

Figure Plot of the MJUO data for all years showing the much weaker

correlation in FWHM vs velocity

from the star one exp ects to see group events for FEBs closer to the star less often

at higher velocities or blueshifted

The single events will tend to b e the shallower features which app ear most often

and account for all the HVFs over kms This is not entirely surprising since

as FEBs pass closer to the star their crossing times get shorter and any other b o dy

in an orbit with a similar inclination to the line of sight has a smaller probability of

o ccurring during the crossing time of another b o dy

Examining the plot of the depths of the absorption features against their velocity it

b ecomes clearly apparent that the deep est lines are found only at low velocities b etween

redshifts of kms This corresp onds to clouds of ions orbiting the star at large

distances which can also cover large fractions of the stellar disk A second set of features

have depths ab out half as deep as the lower velocities lines which feature a denite cuto

at kms This may imply a limit in the inclinations of the parab olic orbits of FEBs

app earing in Ca II sp ectra Shallow and broad feature are found at most velocities and

tend to represent individual FEBs The blueshifted features show a similar cuto in the

depths of the features at kms see Fig

The equivalent widths of the lines for a given velocity are all b elow mAwith the

H line features generally smaller than K except at blueshifted velocities see Fig

This result implies that the blueshifted features o ccur under very sp ecial circumstances

since the oscillator strength of the Calcium K line is twice the oscillator strength of the

Summary of observational results

Figure Plot of the depths of the absorption features observed vs their

velocities over all observations

H line

In their pap er LagrangeHenri et al showed that the depths of the Calcium lines

taken simultaneously or nearly so can provide information on the size of the absorbing

cloud of ions as a fraction of the stellar disk see Fig The plot of the line ratios for

normal Ca II should have a value of pKpH and for saturated lines pKpH where

p is the depth of the line The line ratios in the FEB scenario sometimes have pKpH is

less than when the Ca II cloud masks the part of the stellar disk rotating away from

the observer As can b e seen in Fig cf Fig the deep lines can cover signicant

fractions of the stellar disk in some cases The HVFs have much smaller depths and

the amount of disk covered is also small The blueshifted features only give

lower limits on how much of the stellar disk they cover since none of those features have

app eared to b e saturated hence they at least cover of the stellar disk

The timescales over which the observed variations in the sp ectroscopic lines vary over

relate quite strongly to the velocity with resp ect to the star of the absorption feature

The lower the velocity the longer the timescale esp ecially if the absorption features are

deep and the higher the velocity the shorter the timescale of variation The redshifted

deep LVFs have b een observed to last for weeks at a time and indeed one even app ears to

last several months during which time they show little if any variation in their equivalent

widths As the velocity increases the timescales come down slowly till they reach the

crossing time for an FEB across the stars disk of hours at kms Ab ove that

velocity the timescales for the FEBs are less than the maximum crossing time and typically

Chapter Analysis

Figure Plot of the equivalent widths in mAof the absorption features

observed vs their velocities over all observations

Figure Plot of lling factors for quasisimultaneous observations in H

K Squares represent blueshifted features crosses LVFs from kms

and triangles for HVFs ab ove kms

Conclusion

only app ear for a few hours at most Many absorptions have b een seen in single sp ectra

only indicating they cross the line of sight in the exp osure time of the sp ectra hour

or slightly longer

The blueshifted absorption features at low velocity also last quite long times a few

days at least but most of the observed features lasted only short times typical of the

crossing times of b o dies in orbit around the star through the line of sight

Despite the retraction of the observation of a dust disk around Piscis Australus as

an IRAS p oint source it may have dust distributed at larger distances from the star and

also may b e viewed face on rather than edge on The sp ectra taken during only

show the photospheric lines with no additional absorption features

Observations of Ophiuchus in H and K failed to detect any absorption features

except for the lines b elieved to b e due to interstellar clouds While additional features

were observed in sp ectra they turned out to b e due to dust on the window in front of

the detector

Conclusion

This thesis contains sp ectroscopic observations of the stars Pictoris Piscis Australus

and Ophiuchus taken from December till November at the conclusion of

observations for this thesis From these sp ectra useful data has b een obtained from the

equipment at MJUO which have yielded new results from the prolonged time coverage

these observations cover

The metho d used here for the detection of absorption features pro duces the same

results as LagrangeHenri et al in their pap er on the observing campaign on

Pictoris as well as detecting many more features that their metho d is insensitive to More

shallow absorption features are seen in this data as well as the detection of a number of

blueshifted absorption features

The predictions of the FEB scenario seem to explain many of the observations of the

variable features seen in the resonance lines of Calcium II at H and K The timescale of

variations of the HVFs do agree well since all the HVFs vary in less than the crossing time

of hours The HVFs are generally broader and shallower than lower velocity features

and app ear more often in K than in H There is evidence of the variation predicted by

Beust and Lissauer of the equivalent widths of the H and K lines as an FEB crosses from

one side of the stellar disk to the other due to doppler shift in the stars light from each

side The LVFs are found in greater numbers and are easier to detect than the HVFs

also the LVFs are the domain of the deep absorption lines The features observed are

b oth made up of single events and groups of b o dies causing the observed variations The

numbers of each type of features observed were as exp ected however some interesting

features seem to defy explanation

Chapter Analysis

The longlived events which almost always app ear as deep absorption features are still

problematical in that there is no adequate explanation as to how these lines can app ear

unchanging for at least one case months at a time However a solution to this problem

could help explain the circumstellar feature p ersistence

There are many other puzzles yet to b e solved from these observations on the timescales

of features and the constant nature of the equivalent widths

The observations of the Calcium lines found no variable absorption features in the

sp ectra of either Ophiuchus or Piscis Australus While they remain go o d candidates

for further observations of the gas disk and infra red excess resp ectively the presence of

variable absorption features seems unlikely to b e observed

Chapter

Future work

Future work on studying the Pictoris system and other similar systems should include

improvements in three domains observations reductions and analysis

Improvements to the sp ectroscopic observation technique esp ecially for long term

programmes should include observations of Canopus Carina Canopus app ears to

have a very stable line prole and so it can b e used correct for any changes in the setup

of the echelle sp ectrograph causing changes in the shap e of the prole of the photospheric

lines in the programme stars Canopus is conveniently situated only from Pictoris

A new pixel CCD is on order for use at MJUO which has in addition to

its larger size has a far higher quantum eciency over all wavelengths than the

current CCD system Combined with the fo cal reducer which is b eing designed truly

simultaneous observations of b oth the calcium H and K lines will b ecome p ossible In

addition the simultaneous observations and greater quantum eciency will allow for

b etter temp oral characterisation of the variable absorption features and higher signalto

noise ratios in the observed sp ectra It will then b e p ossible observationally to search for

further and more convincing conrmation of the results of the result of Beust and Lissauer

that the H line can sometimes b e stronger than the K line as is predicted by the FEB

scenario

Simultaneous observations will mean that it is no longer necessary to change the

sp ectrograph settings during the night This will improve smo othelding and also make

it easier to establish an indep endent wavelength scale This observational improvement

should result in improved reductions The stability of the observing set up should insure

that the background subtraction imp ortant for accurate determination of the depth of

the circumstellar feature is done consistently

The indep endent established wavelength scale will eliminate the need to align observed

and reference sp ectra on the shortwavelength side of the circumstellar feature which is

a pro cedure which hides any lowvelocity blueshifted features The uncertain matching

of the circumstellar depths splits any broad zerovelocity features into narrow features

one blueshifted and one redshifted This uncertainty should b e eliminated by b etter

background subtraction and the improved wavelength scale

In the interim some improvements can b e made to the current reference sp ectra used for

division of the observed sp ectra for extraction of absorption feature parameters Primarily

the region around the circumstellar feature needs some attention given that the often low

Chapter Future work

signal levels can cause tiny dierences to b ecome problematic in the normalised sp ectra

when tting the gaussians to the features Also it may b e p ossible to make measurements

of the circumstellar line with improvements to the division pro cedure by tting a parab ola

to the currently used K line reference sp ectrum and another quadratic to the region of the

circumstellar line in the H line reference sp ectrum Since in the division by the reference

sp ectrum of the observed sp ectra the circumstellar line do es exhibits some variation it may

b e p ossible to measure this quantitatively and p erhaps even characterise these changes

The tting of the gaussians to the absorption features using the Mo diedGauss

Newton metho d can b e improved with b etter parameters for the tting routine to use in

the matching of the synthetic t with the data The mo dications needed at this stage

are merely renements on a pro cess that works quite well

Obviously detailed mo delling outside the scop e of this thesis would lead to further

renements of the FEB scenario and its predictions

Chapter

Acknowledgements

Id like to thank my sup ervisor William Tobin for oering this pro ject and for all the

help and advice he gave during this research I wish to thank all the sta at Mt John

Observatory Mike Clark who would come down whenever I had a problem with the

equipment and help set up the sp ectrograph John Baker for setting up the telescop e

Alan Gilmore and Pam Kilmartin for breaking the monotony of those long lonely runs in

winter with those chats and dinner and June Clark

I would also like to thank the rest of the astronomy group for the company and

discussions esp ecially Lyndon Karen John Pritchard for all their help when I started

using MIDAS and again to John for all his help with LaTex and BibTex when it seemed

to b e getting the b etter of me Irene Michael Jovan Ljiljana LauraBeth and Peter for

b eing such go o d company and friends Thanks also to John Hearnshaw and Peter Cottrell

for their supp ort and encouragement

I like to thank those who were up on the mountain at the same time working on the

MOA pro ject who help ed while away the hours on those cloudy nights esp ecially Tom

Love Ian Jordan for all those interesting discussions Tim Banks for letting me in on some

of the secrets of writing a thesis Dr M Honda and Toshi Yanigasaw who were overseeing

the MOA pro ject during my last run up at Mt John

This research made use of the Simbad database op erated at CDS Strasb ourg France

the valuable assist of which is greatly appreciated The online resources made available

across the Internet from the STScI are also appreciated in freely providing some of the

images used in this thesis and the information on the latest observations of Pictoris by

the Hubble Space Telescope

References

D Aitkin T Mo ore P Ro che C Smith and C Wright MNRAS LL

C Allen Astrophysical Quantities Athlone Press U of London

P Artymowicz C Burrows and F Paresce ApJ

H Aumann F Gillett C Beichman T de Jong J Houck F Low G Neugebauer R Walker

and P Wesselius ApJ Lett L

D Backman F Gillett and F Witteb orn ApJ

H Beust AM LagrangeHenri A VidalMadjar and R Ferlet AA

H Beust and M Tagger Icarus

H Beust A VidalMadjar and R Ferlet AA

H Beust A VidalMadjar R Ferlet and AM LagrangeHenri Ap Space Sci

H Beust A VidalMadjar AM LagrangeHenri and R Ferlet AA

A Boss Lowmass star and planet formation PASP

F Bruhweiler Y Kondo and C Grady ApJ LL

K Cheng F Bruhweiler Y Kondo and C Grady ApJ LL

R Chini E Krugel B Shustov A Tutukov and E Kreysa AA

E Churchwell In CJ Lada and ND Kykas editors The physics of star formation and early

stel lar evolution volume of NATO ASI series CMathematical and physical sciences page

NATO advanced study institute

A Crawford J Spyromilio M Barlow F Diego and A Lagrange MNRAS LL

D Crawford J Barnes and J Golson AJ

M Deleuil C Gry A LagrangeHenri A VidalMadjar H Beust R Ferlet H Mo os T Liven

go o d D Ziskin P Feldman and M McGrath AA

A Lecavelier des Etangs M Deleuil A VidalMadjar R Ferlet C Nitschelm B Nicolet and

A LagrangeHenri AA

A Lecavelier des Etangs G Perrin R Ferlet A VidalMadjar F Colas C Builand F Sevre

J Arlot A LagrangeHenri J Lecacheux M Deleuil and C Gry AA

References

T Encrenaz J Bibring and M Blanc chapter The Formation of the Solar System AA

Library SpringerVerlag Berlin Heidelb erg

C Turon et al Hipparcos Input Catalogue ESA Publications Division

R Ferlet L Hobbs and A VidalMadjar AA

R Ferlet A LagrangeHenri H Beust R Vitry J Zimmerman M Martin S Char M Belmahdi

J Clavier P Coupiac B Foing F Sevre and A VidalMadjar AA

F Gillett In FP Israel editor Light on Dark Matter volume page Reidel Dordrecht

I Glass MNASSA

T Gledhill S Scarrott and R Wolstencroft MNRAS

E Green P Demarque and C King The Revised Yale Isochrones and Luminosity Functions Yale

U Obs New Haven

Beust H and Lissauer J AA

B Hauck and M Mermillio d Astr Ap Suppl

J Hearnshaw Proc ASA

G Herbig In L Danly H Weaver editor The Formation and Evolution of Planetary Systems

page Cambridge U Press Cambridge

L Hobbs ApJ

L Hobbs A LagrangeHenri R Ferlet A VidalMadjar and D Welty ApJ LL

L Hobbs A VidalMadjar R Ferletand C Alb ert and C Gry ApJ LL June

D Hoeit and C Jaschek The Bright Star Catalogue Yale U Obs th edition

P Kalas and D Jewitt AJ

R Knacke S Fajardoacosta C Telesco J Hackwell D Lynch and R Russell ApJ

November

Y Kondo and F Bruhweiler ApJ LL April

P Lagage and E Pantin Nature June

A Lagrange R Ferlet and A VidalMadjar AA

A LagrangeHenri H Beust R Ferlet and A VidalMadjar AA LL

A LagrangeHenri R Ferlet A VidalMadjar H Beust C Gry and R Lallement Astron

Astrophys suppl ser

A LagrangeHenri E Gosset H Beust R Ferlet and A VidalMadjar AA

References

A LagrangeHenri F Plazy H Beust D Mouillet M Deleuil R Ferlet J Spyromilio A Vidal

Madjar W Tobin J Hearnshaw M Clark and K Thomas

A LagrangeHenri A VidalMadjar and R Ferlet AA

C Lance Ap J Suppl

H Levison M Duncan and G Wetherill Nature December

T Mouschovias In Lada CJ and Kykas ND editors The physics of star formation and early

stel lar evolution volume of NATO ASI series CMathematical and physical sciences page

NATO advanced study institute

G Nankivell and N Rumsey In JB Hearnshaw and PL Cottrell editors Instrumentation and

research programs for smal l telescopes number in Symp osium page p IAU D Reidel

publishing company

B Palmer and R Engleman Atlas of the Thorium Spectrum Los Alamos National Lab oratory

F Paresce AA LL

F Paresce and CBurrows ApJ Lett L

Photometrics Limited Photometrics users manuel for Forth programming system

K Pollard The nature of low mass supergiants RV Tauri and R Coronae Borealis variables PhD

thesis Dept Physics Astronomy University of Canterbury

F Ruferner Catalogue of Stars Measured in the Geneva Observatory Photometric System Obs de

Geneve

M Savoldini and G Galletta AA

T SchmidtKaler LandoltBornstein b

A Slettebak ApJ

A Slettebak and K Carp enter ApJ Suppl

B Smith J Fountain and R Terrille AA Research Note

B Smith and R Terrile Science

B Smith and R Terrile BAAS

A Stern M Festou and D Weintraub A map of a collionally evolving dust disk around fomalhaut

Nature June

S Stern Iauc IAU circular

B Stromgren Ann Rev Astr Ap

C Telesco E Becklin R Wolstencroft and R Desher Nature September

C Telesco and R Knacke ApJ LL

W Tobin Description of the system Performance note

W Tobin Forth acquisition software Performance note

A Uesugi and I Fukuda Revised Catalogue of Stel lar Rotational Velocities Dept of Astronomy

Kyoto U

A VidalMadjar and et al Iauc IAU circular

A VidalMadjar L Hobbs R Ferlet C Gry and C Alb ert AA

A VidalMadjar A LagrangeHenri P Feldman H Beust J Lissauer M Deleuil R Ferlet

C Gry L Hobbs M McGrath J McPhate and H Mo os AA

H Walker and R Wolstencroft PASP pages December

L Waters and et al Circumstellar gas and dust in ophiuchi aive AA

P Weissman Space Science Review

S Wol The astars Problems and p ersp ectives Technical rep ort NASA

R Wooley E Epps J Penston and S Poco ck R Obs Ann

App endix A

Reducing MJUO CCD Echelle Sp ectra Using

ESOMIDAS

This manual has b een up dated to b e applicable to any and all sp ectra taken with the

MJUO CCD and b eing reduced with ESOMIDAS NOV on cantua

To access Midas type inmidas unit p

This will put you into Midas using a unit number that noone else should b e using If

you are using a WYSE terminal the unit number is replaced with letters from XX to

ZZ Once this is done you can then load the sp ectra from your disk space and convert

them from FITS format to the MIDAS binary data format

The pro cess starts with the sp ectra in FITS format on the computer disk

A Creating Images of the Correct Format for MIDAS

MIDASi intapets numbers of images to b e pro cessed MIDAS les prex

FITS les prex

Convert the sp ectral images in FITS format into the idiosyncratic MIDAS format

will convert les to originally from Example MIDASi intapets T T

Tape T and save them as TBDF TBDF TBDF

Note all spacing in commands are imp ortant to MIDAS

MIDASi createicat name for catalogue MIDAS les prexBDF

Create an ASCI I le listing the image les for subsequent op erations

Example MIDASi createicat T TBDF

will create a catalogue named TCAT listing all les named in the form TBDF

MIDASi setcontext echelle

Make echellespecic commands available for the following op erations

MIDASi rotateechelle catalogue name prex for rotated les ip XY

Transform all of the images listed in the catalogue into the layout required by MIDAS

orders increasing from top to b ottom and wavelength increasing from left to right as the

images are displayed on the screen

Example MIDASi rotateechelle T T ip XY will transform all of the les listed

in the catalogue T retaining the same le names

A Initializing for the MIDAS Session

The ip option is used b ecause MIDAS do es not seem to b e able to rotate in one

op eration

Note that the capital XY must b e used to rotate all the les a small xy causes funny

things to happ en with MIDAS Also note that this op eration creates new les named MI

DAS le prexBDF with sequential numbering indep endent of the numbers of the

original les If the les listed in the catalogue are not sequentially numbered b eginning

with then it will b e necessary either to keep track manual ly of the corresp ondences

b etween the original and the transformed les or to rename the transformed les to corre

sp ond with the original les If the names are the same then UNIX is unable to transform

the images so a change in le name is required

A Initializing for the MIDAS Session

MIDASi creategraphics

Make a window for displaying plots

MIDASi createdisplay

Make a window numbered the right size for displaying the rotated images

MIDASi loadlut pastel

This will put any image into false colour and will show up all orders and cosmic rays more

clearly for identication Other palettes ie lo ok up tables include rainbow backgr

color heat light random random staircase and stairs

MIDASi initechelle previous session name

Set the echelle reduction parameters to those saved from a previous session if available

A Lo cating the Echelle Orders in the Images

This entire section is used only for the rst reduction of a set of images made with a given

setup of the telescop e and instruments Once the orders have b een found usually from

a smo otheld image the same order table can b e used for all stellar images obtained

with that setup But if the tilt of the orders changes then the ordernding must b e

rep eated Normally the tilt only changes if the CCD cryostat has b een removed from the

sp ectrograph and then replaced and in that case the observer should have obtained a

new smo otheld image to dene the new p ositions of the orders

MIDASi setechelle DEFMTDstd DEFPOL

The DEFMTDs available are std com when the width of the orders are larger than the

interorder spacing There is also the metho d hough when these ab ove metho ds fail to

work in the searchorder command See Houghechelle in Other Useful Commands at

end The DEFPOL ts a p olynomial of degree XY where X the p olynomial for tting

App endix A Reducing MJUO CCD Echelle Sp ectra Using ESOMIDAS

along the orders and Y the p olynomial for tting across the orders A dimensional

p olynomial of degree will normally b e used to approximate the p ositions of the orders

This may need to b e changed to degree if there is diculty in nding any orders eg

only orders detected or present

MIDASi showechelle

This shows what the echelle reduction parameters are It is useful to check that the pa

rameters have b een changed

Echelle Parameters

Instrument INSTR

Dispersers GRATING CROSS

Detector CCD CCDBIN

Image size IMSIZE SCAN

Orders orders from to

Wavel range wavel units

Order definition

DEFMTD hough DEFPOL

ORDREF ksmooth

NBORDI WIDTHI THRESI