Progress report on analy lysis and calibration Progress report on - - PowerPoint PPT Presentation

progress report on analy lysis and calibration progress
SMART_READER_LITE
LIVE PREVIEW

Progress report on analy lysis and calibration Progress report on - - PowerPoint PPT Presentation

Progress report on analy lysis and calibration Progress report on analysis and calibration of f RAV AVE data and nd preparation n fo for DR3 of RAVE data and preparation for DR3 Toma a Zwitter Toma Zwitter University of


slide-1
SLIDE 1

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 1 1

Tomaž Zwitter Tomaž až Zwitter

University of Ljubljana University of f Ljubljana na Faculty of Mathematics and Physics Faculty of f Mathe hematics and nd Phy hysics

Progress report on analysis and calibration Progress report on analy lysis and calibration

  • f RAVE data and preparation for DR3
  • f

f RAV AVE data and nd preparation n fo for DR3

3rd field plate

slide-2
SLIDE 2

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 2 2

Plan: Plan an: FP3 FP3 RVs RVs Stellar parameters Ste Stellar ar param amete ters Distances Distan stances DR3 preparation DR3 prepar arati ation DR4+ DR4+

Progress report on analysis and calibration Progress report on analy lysis and calibration

  • f RAVE data and preparation for DR3
  • f

f RAV AVE data and nd preparation n fo for DR3

3rd field plate

slide-3
SLIDE 3

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 3 3

FP3 performance FP FP3 p performance

3rd field plate

FP3 results FP3 results (Mary): (Mary): summary: very good summary: very good FP3 performance FP3 performance (Fred): (Fred): Fiber lifetime: very good Fiber lifetime: very good Mechanics: ok, caution on cable & slit Mechanics: ok, caution on cable & slit

slide-4
SLIDE 4

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 4 4

Reduction pipeline Reduc uctio ion n pipe ipelin ine

Fred & observers Paul Warren Alessandro Mary Input Catalog Group Data Management Group checking... Actually everybody Actually everybody connected to connected to everybody else. everybody else.

slide-5
SLIDE 5

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 5 5

Number of pointings Num umber r of po point ntings gs

~350.000 spectra covering ~17.000 square degrees ~350. 0.000 00 spectra c covering ~ ~17.0 .000 s square degrees ~20 objects/degree ~20 o

  • bjects/degree2

2 on average, maximum is 107 spectra/degree

  • n a

average, , maxi ximum is 1 107 s spectra/degree2

2.

.

slide-6
SLIDE 6

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 6 6

Measured radial velocities Measure red ra radia ial l veloci locitie ies

slide-7
SLIDE 7

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 7 7

Radial Radia ial velocity ve veloc

  • city

errors erro rrors rs

slide-8
SLIDE 8

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 8 8

Radial velocity shifts Radial al ve velocit ity y shi hifts

Uncontrolled shifts in wavelength calibration. Uncontrolled shifts in wavelength calibration. Affects 5 whole fields and in addition the fibers 3-9 in 66 fields Affects 5 whole fields and in addition the fibers 3-9 in 66 fields in DR3. in DR3. Solution: Solution:

  • few fields & few fiber 3-9 ranges with really bad RVs are removed,
  • few fields & few fiber 3-9 ranges with really bad RVs are removed,
  • the user needs to be suspicious of and RV corrections >10 km/s.
  • the user needs to be suspicious of and RV corrections >10 km/s.
slide-9
SLIDE 9

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 9 9

Radial velocity definition: Radial vel veloc

  • city defi

efinition

  • n:

gravitational redshift not included gravi vitat ation

  • nal redshift

ft not included ed

v = GM / Rc = v = = G GM / R Rc = = 0.636 km/s ( M/M = 0 0.6 .636 k km/s ( M/Mסּ סּ ) ( R/R ) ( ( R R/Rסּ סּ ))-1-1 ∇∇p = - p = = -

  • ρ

ρ g , g = GM / R g g , , g g = GM / R2

2

M and R enter stellar atmosphere M and R enter stellar atmosphere calculation through: calculation through: Gravitational redshift of light Gravitational redshift of light leaving the stellar surface: leaving the stellar surface:

> From: rkurucz@cfa.harvard.edu > Date: June 5, 2009 12:12:30 AM GMT+02:00 > Dear Urtzi, > SYNTHE always computes in the laboratory rest frame because the line > data are in the laboratory frame... That way I can check wavelengths > against laboratory line lists without having to think about red shifts.

slide-10
SLIDE 10

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 10 10

Radial velocity definition: Radial vel veloc

  • city defi

efinition

  • n:

gravitational redshift not included gravi vitat ation

  • nal redshift

ft not included ed

Morales-Rueda et al. 2005

Example: A white dwarf + M dwarf binary. Ex Example: A A white d dwarf + + M M d dwarf bi binary. y.

slide-11
SLIDE 11

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 11 11

Check with repeated observations Check ck with re repeated obs

  • bserv

rvations

19 May 2009 internal data release: 19 May 2009 internal data release:

∆ RV = 1.8 km/s RV = 1.8 km/s

276024 spectra, 276024 spectra,

∆ Teff = 135 K Teff = 135 K

241376 objects observed once, 241376 objects observed once,

∆ log g = 0.2 dex, log g = 0.2 dex,

15730 more than once 15730 more than once

∆ [M/H] = 0.1 dex [M/H] = 0.1 dex These errors on stellar parameters are These errors on stellar parameters are less than half less than half of what is

  • f what is

derived from comparison with external datasets. So external datasets derived from comparison with external datasets. So external datasets values may not be error free and RAVE errors are indeed smaller. values may not be error free and RAVE errors are indeed smaller. If If ∆ ∆ RV > 5 km/s this is likely due to binarity (single lined spectroscopic RV > 5 km/s this is likely due to binarity (single lined spectroscopic binary). This happens in 6% of the cases. In binary). This happens in 6% of the cases. In ~0.6 ~0.6% binarity in a form % binarity in a form

  • f a double lined spectroscopic binary detected from a single spectrum.
  • f a double lined spectroscopic binary detected from a single spectrum.
slide-12
SLIDE 12

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 12 12

Errors for repeated observations Errors rrors for for re repeated ob

  • bserv

rvations

  • ns
slide-13
SLIDE 13

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 13 13

Are parameter values reliable? Are parameter values reliable?

Last report by Corrado & Mary (25-May-2009)

slide-14
SLIDE 14

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 14 14

Are parameter values reliable Are re p para rameter r values re reliable le

(within limits expected for a dataset of this kind) (withi hin li limits expected fo for a dataset of f this kind nd)?

?

Two cases: low S/N spectra, high S/N spectra. Two cases: low S/N spectra, high S/N spectra. Low S/N: everybody agrees, will be discussed later. Low S/N: everybody agrees, will be discussed later. High S/N: no (Corrado, Mary) High S/N: no (Corrado, Mary) yes (Ulisse, Tomaz, Jon) yes (Ulisse, Tomaz, Jon)

Arguments to be presented during this meeting: Arguments to be presented during this meeting: Comparison with synthetic spectra, with Soubiran&Girard sample Comparison with synthetic spectra, with Soubiran&Girard sample and observations of clusters using modified (how?) RAVE pipeline and observations of clusters using modified (how?) RAVE pipeline to reduce them (Mary, Corrado). to reduce them (Mary, Corrado). High-res tests (Jon), tests on repeats (Jon), High-res tests (Jon), tests on repeats (Jon), better metallicity calibration (Jon). better metallicity calibration (Jon). Tests on temperature and metallicity using reliably Tests on temperature and metallicity using reliably classified HD stars (Ulisse), customary observations (Ulisse). classified HD stars (Ulisse), customary observations (Ulisse). Comments on theoretical spectra and their culprits (Urtzi). Comments on theoretical spectra and their culprits (Urtzi).

slide-15
SLIDE 15

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 15 15

Are parameter values reliable Are re p para rameter r values re reliable le

(within limits expected for a dataset of this kind) (withi hin li limits expected fo for a dataset of f this kind nd)?

?

Just two comments from my side taken from a conference Just two comments from my side taken from a conference in Palm Cove last month: in Palm Cove last month: The library we use was extensively praised by Mike Bessell. The library we use was extensively praised by Mike Bessell. Martin Asplund's presentation had a title Martin Asplund's presentation had a title „ „Is the Sun chemically unusual?“ Is the Sun chemically unusual?“

slide-16
SLIDE 16

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 16 16

Distance from Dis istan ance fro rom stellar models stellar r models

Breddels et al. 2009, submitted Bredde ddels et al. 200 2009, 9, sub ubmitted

Distance error < 1/3 for ~1/3 of stars. Distance e error < < 1 1/3 f for ~ ~1/3 o

  • f stars.

.

slide-17
SLIDE 17

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 17 17

Age Ag Age e (not published)

(not publi lishe hed)

slide-18
SLIDE 18

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 18 18

Improving distances from Improv

  • ving distance

ces from stellar models stel ellar mod

  • del

els

Klement et al. 2008 Klement nt et al. . 200 008

Currently: Currently: distance error < 1/3 for ~1/3 of stars. distance e error < 1 1/3 f for

  • r ~1/3 o
  • f s

stars. .

slide-19
SLIDE 19

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 19 19

RAVE stars RAVE stars rs

Data from Breddels et al. 2009, animation by G. Matijevič Data f from B Bredde dels et al. . 200 009, 9, animation

  • n by G

y G. . Matijevi vič blue bl blue ue: : dwarfs (log g>3.5) dw dwarfs ( (log g g>3.5) >3.5) red red: : giants (log g<3.5) giant nts (log g<3

  • g g<3.5)
slide-20
SLIDE 20

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 20 20

RAVE stars (all distances we have) RA RAVE stars rs (all ll dis istance nces w we h have)

Data from Breddels (private communication), animation by G. Matijevič Data f from B Bredde dels (pr priva vate c com

  • mmuni

nication)

  • n), a

, animation by G n by G. M . Matijevi vič blue bl blue ue: : dwarfs (log g>3.5) dw dwarfs ( (log g>3.5) g g>3.5) red red: : giants (log g<3.5) giant nts (log g<3

  • g g<3.5)
slide-21
SLIDE 21

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 21 21

RAVE stars 220 million years ago RAV RAVE st stars s 220 million

  • n yea

ears ag ago

Data from Breddels et al. 2009, animation G. Matijevič & M. Vodopivec Data f from B Bredde dels et al. . 200 009, 9, animation

  • n G. M

Matijevič & & M

  • M. V

. Vodo

  • dopi

pive vec blue bl blue ue: : dwarfs (log g>3.5) dw dwarfs ( (log g g>3.5) >3.5) red red: : giants (log g<3.5) giant nts (log g<3

  • g g<3.5)
slide-22
SLIDE 22

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 22 22

RAVE & RAV AVE & Geneva Gen eneva va Copenhagen Cop

  • pen

enhagen en survey su survey vey

animation G. Matijevič ani nimation G

  • n G. M

. Matijevi vič blue bl blue ue: : dwarfs (log g>3.5) dwarfs ( (log g>3 g g>3.5) red red: : giants (log g<3.5) giant nts (log g<3

  • g g<3.5)
slide-23
SLIDE 23

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 23 23

RAVE & RAV AVE & SDSS SDSS survey su survey vey

SDSS data by P. Re Fiorentin, SD SDSS SS d data by by P.

  • P. Re Fi

Fior

  • rent

ntin, n, animation G. Matijevič ani nimation G

  • n G. M

. Matijevi vič blue bl blue ue: : dwarfs (log g>3.5) dw dwarfs ( (log g g>3.5) >3.5) red red: : giants (log g<3.5) giant nts (log g<3

  • g g<3.5)
slide-24
SLIDE 24

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 24 24

Distances in GCS, RAVE & SDSS Distances ces in GCS, RAV AVE & SDSS

slide-25
SLIDE 25

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 25 25

Presence of Presen esence e of giants & dwarfs giants s & dwarfs fs

slide-26
SLIDE 26

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 26 26

Preclassification - scheme by Gal Matijevič Preclassi ssificati ation - sc scheme by by Gal Gal Mati tijevič

slide-27
SLIDE 27

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 27 27

Eyeballing Eyebal balling

slide-28
SLIDE 28

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 28 28

Eyeballing Eyebal balling

slide-29
SLIDE 29

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 29 29

Eyeballing Eyebal balling

slide-30
SLIDE 30

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 30 30

Properties of Properties of f parameters parameters at low S/N (Corrado) at lo low S/N (Corrado)

slide-31
SLIDE 31

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 31 31

Magnitude and noise distributions Magnit gnitud ude and nd nois noise dis istri ribut utions ions

For a typical giant I = 12.5 corresponds to V = 13.5 and g = 14.0.

Systematic errors at low S/N Syste tematic tic e errors a at lo t low S S/N /N

slide-32
SLIDE 32

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 32 32

DR3 proposal: 3 separate catalogs DR3 R3 prop ropos

  • sal:

l: 3 3 separa rate ca catalogs logs

slide-33
SLIDE 33

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 33 33

DR4 preparations DR4 R4 pre repara rations ions

We maintain the list of 3 catalogs Catalog A will have to be eyeballed (sic!) - August First using the new input catalog Seeking new partners – so no spectra published?

From: Sofia Feltzing <sofia@astro.lu.se> To: Tomaz Zwitter <tomaz.zwitter@fmf.uni-lj.si> Subject: Date: Thu, 4 Jun 2009 10:28:32 +0200 X-Mailer: Apple Mail (2.753.1) GALACTIC ASTRONOMY, GAIA AND THE NEED FOR HIGH RESOLUTION SPECTROSCOPY A meeting at the IoA, Cambridge 8+9 July 2009

slide-34
SLIDE 34

Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 Tomaž Zwitter: RAVE, Padova workshop, 11/6/2009 34 34

Strange things happen... Stra range nge things ngs happen...