Tue, 05 May 2009

Welcome [Subject: /Welcome] 

This is the VALD Development Weblog, an e-mail archive for
VALD developers up to Nov 2007.

This page is no longer updated and only exists for historical
reference.
For current VALD development information go to the VALD wiki.

Wed, 25 Jul 2007

Participation of VALD team members in relevant meetings [Subject: /Meetings] 

The 9th International Colloquium on
Atomic Spectra and Oscillator Strengths
for Astrophysical and Laboratory Plasmas
(ASOS9, 7-10 August 2007, Lund, Sweden)
* Ulrike, poster
* Tanya, poster

NLTE workshop
(30 July - 4 August 2007, Nice, France)
* Ulrike, Oleg, Luca, poster

VIth Serbian Conference on Spectral Line Shapes in Astrophysics
(11-15 June 2007, Sremski Karlovci, Serbia)
* Paul, talk

EURO-VO workshop
(21-23 March 2007, ESAC - Madrid, Spain)
* Nik, 15 min talk: "Practical tools for astronomical spectroscopy"
* Marlene, poster: "Abundance analysis of peculiar stars based on VALD"
* Bertrand, 30 min talk: "Stellar synthetic spectroscopy in the VO era"

Tue, 03 Jul 2007

To-do list [Subject: /Todo] 


This To-do list is roughly prioritized into two categories:
* Things to be done before the release of VALD-3
* Things which can also be done after the VALD-3 release

Some of the tasks are already assigned to a specific person
(indicated by ** Name **), as discussed before or during the Vienna workshop
(Feb 2007). The list may appear quite long at first sight, but some of the
tasks have been broken up in smaller steps already.

Please

1) Check the list for any corrections or additions.

2) Send suggestions for volunteers for the unassigned tasks
   (including yourself, of course) to Ulrike.

3) Send a note to Ulrike when you have completed a task.

Keep in mind our goal to release VALD-3 release before summer (June 2007)!


To be done before VALD-3 release:
=================================

* Extraction software
  - preselect3 and showline3: merge patches by Christian into Nik's latest
    version ** Friedrich **
  - add source code files to version management system ** Friedrich **
  - create portable Makefile to handle the byte-swapping option for all VALD
    tools ** Christian? **
  - update team documentation
  - implement extraction and output of references in all tools ** Nik **
  - implement extraction of isotopes (option to extract either only the lab
    mixture data or all individual isotopes) in preselect3
  - update user documentation

* Web interface
  - implement personal preferences (e.g. format) with e.g. cookies ** Eric **
  - update to include option for molecules ** Eric **
  - update to include option for isotopes ** Eric **

* VALD directory structure
  - remove all files not belonging to the default structure (put them in
    separate directory) -> same structure on all three servers ** Christian **

* Backup
  - Implement backup system for VALD

* Autoionization lines
  - define flag in format
  - implement reading flag in tools
  - update line lists

* Web documentation (for users)
  - update (including wavelengths - vacuum/air - and VALD-3) ** Christian **
  - convert to .tex and .pdf and make available for downloading
    ** Christian **

* "Inside VALD" document
  - correct existing sections ** All **
  - add text on Web interface (7. and 9.10) ** Eric **
  - add text on references (8. and 9.2) ** Friedrich **
  - add text on VALD-3 (9.)

* Cools star model atmospheres for EXTRACT STELLAR
  - copy MARCS (Plez or Uppsala?) models into ems/MODELS directory
  - implement switching between MARCS and ATLAS at solar temperatures in tools
  - update documentation

* Complete molecular line lists
  - provide data for all VALD fields for TiO list ** Bertrand **
  - insert data in VALD lists ** Nik **
  - provide partition functions ** Paul **
  - insert partition functions in VALD_list_of_species ** Nik **

* Fe II line list
  - Verification/check with Schnabel ** Tanya, Paul **

* Atomic line lists
  - convert new lists to VALD-2 format and make available immediately
    ** Tanya, Rimma **
  - split VALD-2 lists (one list per producer), except for Kurucz lists
    ** Tanya, Rimma, Christian, Ulrike **
  - check older lists for energy accuracy loss during conversion ** Ulrike **
  - test conversion from VALD-2 to VALD-3 format ** Nicole, Theresa **
  - extend conversion code to handle special cases for accuracy, source code
    and term designation ** Christian, Ulrike, Nicole **
  - convert all lists

* Type of lines
  - add accuracy flag 'P' for predicted lines in lists
  - update software to interpret this flag
  - update documentation
  - discuss flagging other types of lines (observed, astrophysical, ...)

* VALD-3 test
  - set up VALD-3 on test server and let a few users test for a few weeks

-----------------------------------------------------------------------------

To be done before and/or after VALD-3 release:
==============================================

* VALD meetings
  - finalize date for next VALD meeting in May 2007 ** Nik, Tanya **
  - organize next VALD meeting

* Contact Bob Kurucz ** Nik **

* Baltimore server
  - inquire if Baltimore server agrees with synchronization via rsync
    ** Nik **

* VALD-3 papers
  - REE data ** Tanya **
  - Fe-peak and other elements (Lund etc.) ** Tanya **
  - molecular data, including technical description of VALD-3 ** Nik **
  - all three papers should include astrophysical verification/applications
    ** All **

* New data
  - send references to Tanya (preferably pdf file of paper) ** All **
  - van der Waals data for Cr II - calculate and send to Christian
    when finished ** Paul **
  - data in latest Kurucz line lists - browse the Kurucz website and compile
    an overview of new data ** Oleg **

* Comparison of VALD lists
  - plot comparisons of all VALD lists (with script) and make them available
    on webpage ** Christian **

* VALD and VO
  - contact person ** Ulrike **
  - enquire about value for Austria joining ESO ** Werner **
  - final decision to join
  - build interface
  - test interface
  - register

* Funding
  - include VALD in grant applications when possible ** All **
  - enquire about national funding possibilities ** All **
  - revisit exchange programs between Sweden and Austria ** Werner **
  - consider programs on European level, e.g. eContentplus ** Werner **

* Meetings relevant to VALD
  - send information on meetings and participation to Ulrike ** All **

Mon, 02 Jul 2007

Paul Barklem [Subject: /Discussion/Molecules] 

Here is the list:


H2+ H2 H2- CH CH- C2 C2- CN CN- NH N2 OH OH- BO 
CO NO+ NO O2 HF NaH MgH MgO AlH AlO AlF Al2 SiH 
SiH- SiC SiN SiO SiF Si2 SH SH- CS NS SO MgS AlS 
SiS S2 HCl LiCl NaCl AlCl CaH CaF CaCl ScO TiO 
TiS VO CrO FeO YO ZrO LaO


It was based on everything I could find data for in a list Nik gave me.

/Paul


On Thu, Jun 28, 2007 at 04:18:50PM +0200, Ulrike Heiter wrote:
> Hi Paul,
> 
> Could you send me the list of molecules for which you calculated the
> partition functions?
> 
> Thanks,
> Ulrike
>

Tue, 12 Jun 2007

Past Meetings [Subject: /Meetings]

Tue, 06 Mar 2007

Tanya Ryabchikova [Subject: /Discussion/Conversion] 

Dear Ulli, 

  REE_exp list doesn't contain any error field, letters determine source.
s I remember (and I checked it for a few lines) for VALD2 I splitted this
file for separate elements and put corresponding abbreviations for the 
source, Therefore forget about ree_exp.file, it doesn't exist in 
default.cfg.

Just one more detail. Rimma and myself we have alredy converted fe_1 list 
in a proper VALD_3 format.
                                          Regards, Tanya

Mon, 05 Mar 2007

Christian Stütz [Subject: /Discussion/Conversion] 

Dear Ulli,

> * What exactly is the sense of the <Bcut> and <Ccut> in the output of
>   vald-filter --vald2
>   I looked at the source code but couldn't quite figure it out.

It's quite simple. Bcut cuts the term designation according to 
the longest consecutive row of blanks in between the characters
forming the term designations. Ccut looks for the first uppercase
character followed by a lowercase one (in between there might be
blanks or numbers). That works quite good in many cases.
If nothing is stated, both methods agree. Otherwise the two 
different results for the term designation are denoted by
<Bcut> and <Ccut>.

Try also /> vald-filter -h find_terms

> * Most of the problems that I identified with the previous version of
>   the conversion code remain.
>   - e.g. iron.nbs: "E" is put into the source field, when it is
>     actually an accuracy class ("E" would be the source for Cr III
>     lines)
>   - the NBS/NIST accuracy class is not always found.

See /> vald-filter -h find_source
and /> vald-filter -h find_quality
NBS gives problems because the Reference is not unique.
If no floating point number is given, the E flag is not recognized.
There is certainly some refining to do. That we mistake integer 
numbers as accuracies (eg. C.nlt) is a bug and has to be fixed.

>   I added the file that lists the problems to the version control system under
>   /home/vald/development/tools-source/trunk/EDIT/
>   file UH_conversion_test_accuracy_groups.txt

Thank you.

>   We have to split the work on this into parts (e.g. groups of lists)
>   and assign specific people to work on them (i.e. change/extend the
>   lconv code in the subversion system accordingly).

As you say, we have set up svn, everyone is invited 
to improve the filters.

Greetings 
          Christian

Ulrike Heiter [Subject: /Discussion/Software] 

Dear all,

1) Uppsala VALD server update

   I copied the pres3_2 and showline3_2 sources from Vienna to Uppsala,
   compiled them, made a few simple tests, and copied the binaries into
   the tools directory. So finally the Uppsala server is running the
   same versions of the extraction tools as Vienna and Moscow.
   (I checked that the Vienna and Moscow binaries are the same.)

...

Cheers,
Ulrike

Ulrike Heiter [Subject: /Discussion/Conversion] 

Dear all,

...

2) VALD EDIT and VALD2 -> VALD3

> From: Christian Stuetz 
> Date: Thu, 1 Mar 2007 19:26:10 +0100
> Date: Fri, 2 Mar 2007 09:13:21 +0100
> 
> the lconv() routine used for analysing the comment
> field from VALD2 seems to work quite fine now. 
> I put the binary into 
>   /home/vald/tools/vald-filter
> open for extensive testing.
> 
> have a look into /home/james/VALD2/Lists/FromVALD2/
> We are probably close to a good conversion.
>    

Thanks, Christian. I had a quick look at it.
Some comments:

* What exactly is the sense of the  and  in the output of
  vald-filter --vald2
  I looked at the source code but couldn't quite figure it out.
  
* Most of the problems that I identified with the previous version of
  the conversion code remain.
  - e.g. iron.nbs: "E" is put into the source field, when it is
    actually an accuracy class ("E" would be the source for Cr III
    lines)
  - the NBS/NIST accuracy class is not always found.

  I added the file that lists the problems to the version control system under
  /home/vald/development/tools-source/trunk/EDIT/
  file UH_conversion_test_accuracy_groups.txt
  
  We have to split the work on this into parts (e.g. groups of lists)
  and assign specific people to work on them (i.e. change/extend the
  lconv code in the subversion system accordingly).

  I will take over the lists in accuracy group
  B. quality class (NIST) only, that is:
  carbon.nist
  chromium
  chromium.nbs
  cobalt
  cobalt.nbs
  fe1_cor, fe1_new
  iron
  iron.nbs
  manganes, manganes.nbs
  nickel, nickel.nbs
  nitrogen.nist, oxygen.nist
  ree_exp
  scandium, scandium.nbs, scandium.nbs_cor, titanium, titanium.nbs,
  vanadium, vanadium.nbs, x, y

  I will however not be able to work on it this week (Gaia meeting)
  and only occasionally the rest of March (ESO deadline upcoming).

Cheers,
Ulrike

Thu, 01 Mar 2007

VALD Workshop Vienna [Subject: /Meetings]
Mon to Fri, Feb. 19 to 23, 2007
Thanks to all participants!
Here are the Notes by Ulrike taken during the workshop.
Here is a compilation of the answers to the evaluation form.
Here are some pictures taken during the workshop dinner.

Participants (see Team)

   Ulrike
   Werner
   Christian
   Thomas
   Nik (from Tue 20th evening, until Mon 25th)
   Oleg
   Paul
   Eric
   Tanya
   Rimma
   Friedrich
   Bertrand (from Mon 19th evening, until Wed 21st)
   Theresa
   Nicole
   Luca
   Marlene
   +
   Enrique Solano (Thu afternoon and Fri), EURO-VO

Schedule

Travel
A list of airlines that fly to Vienna can be obtained from the Website of the
Vienna Airport: http://www.viennaairport.com/ -> Flugplanung ->
Fluggesellschaften.

Information on how to get to the Institute as well as maps are available on
the Institute's homepage: http://www.astro.univie.ac.at/en.index.html
(For Visitors -> How to Find Us).
Accommodation
Please organize your accomodation yourself.

The guest rooms at the observatory are now fully booked.

On the Institute's homepage there is a list of hotels in the vicinity of the
observatory (For Visitors -> Hotels). Werner recommends Hotel Schottenpoint.
Lunch and dinner
On the Institute's homepage there is a list of restaurants and cafes in the
vicinity of the observatory (For Visitors -> Restaurants).
Coffee
Coffee breaks will be provided.
Workshop dinner
On Wednesday, Feb 21 we will have dinner together.

Past Meetings [Subject: /Meetings]

VALD team research - from Science Sessions at Feb 2007 Workshop (Notes by UH) [Subject: /Research] 

* Ulrike

  One of several projects: "Nearby stars"
  Spectroscopy in the Local Neighborhood:
  A High Resolution Survey and Abundance Analysis of Selected Stars
  With R. Earle Luck, Case Western Reserve University, Cleveland

  Four papers have been published up to now.
          
  Results for dwarf stars within 15pc and planet hosts:
  - no difference in the lithium contents of the hosts versus the non-hosts
  - mean abundances of other elements 0.1 to 0.2 dex higher in planet hosts
  - overall trends in abundances dominated by galactic chemical evolution

  A summary poster can be downloaded here:
The metallicity of nearby stars 

* Oleg
  
  One of several projects: "Advanced 1D model atmospheres"
  using "LL models" code, including magnetic effects, chemical stratification
  and individualized opacities using VALD linelists.

  Several papers have been published.

* Paul
  
  One of several projects: "Calculations of inelastic collisions"
  These are important for non-LTE calculations.
  In cool stars, electrons and neutral H are the dominant perturbers.
  Example: Li + H collision
  Movies showing the electronic wavefunction and potential energy surfaces
  during a collision can be downloaded here:
H+Li Movies 

* Luca

  PhD project: "Chemical evolution of atmospheres of B type stars
                in open clusters"
  Evolution of peculiar abundances and diffusion with time
  (clusters with log age (yr) = 6.8 to 7.9).

* Nik

  - Continue REDUCE development -> successful for transiting planets
    (measuring radial velocities from S/N=2 UVES spectra)
    Sahu, Casertano, Bond, Valenti, Smith, Minniti, Zoccali, Livio, Panagia,
    Piskunov, Brown, Brown, Renzini, Rich, Clarkson, Lubow (2006, Nature)
  - Unification of molecular equilibrium solver for SME, VALD and spectral
    synthesis codes is in progress, can handle very low temperatures
    (few 100 K to few 1000 K)
  - 3D radiative transfer calculations for Betelgeuse using hydrodynamic
    3D models by B. Freytag; will analyse obervations from D. Gray

* Tanya

  Abundance analysis and study of magnetic chemically peculiar stars
  
  One problem: bad knowledge of damping parameters
  Important for CP stars: Stark effect
  "The influence of Stark broadening on Cr II spectral line shapes in stellar
  atmospheres" - Dimitrijevic, Ryabchikova, Simic, Popovic, Dacic
  e.g. famous Cr II lines in near-UV 3402.26, 3402.40, 3403.25, 3403.31, ...
  Stark damping constants recalculated by Belgrade people using semiclassical
  approach -> better fit -> new data will be used for new release of Cr list
  in VALD
  
  Pulsating CP stars:
  "Rare-earth elements in the atmosphere of the magnetic chemically peculiar
  star HD 144897" - New classification of the Nd III spectrum
  (Ryabchikova et al. 2006, see list of VALD papers)

* Rimma

  Paper for Ni VI (Physica Scripta, Optical Spectroscopy)
  Obtain spectra in laboratory which allow classification of lines
  2 spectrographs (3m: 50 to 350Å; 6.65m: 250 to 2000Å)
  Hot/cold spectra (induction/capacity)
  Comparison between experiment and calculation
  Comparison of least square fitted and Hartree-Fock energy parameters
  
* Friedrich

  - Opacity binning code (derivative of ODF code)
    Applications: stellar atmosphere models, stellar structure models,
    3D and 2D hydrodynamical simulations (e.g. Muthsam et al.)
    
  - Coherent structures
    
    Improved modelling of excitation (energy injection) rates for solar
    p-modes (based on NeMo grid): Samadi, Kupka, Goupil, Lebreton,
    van't Veer-Menneret (2006, A&A)
    "Closure model with plumes": Belkacem, Samadi, Goupil, Kupka, Baudin
    (2006, A&A)
    
    2D simulation of solar photosphere with high resolution (3 km), 3D model
    with somewhat lower resolution
    Conclusions: current lower resolution models of the photosphere by
    other authors modelled correctly, but the higher resolution models show
    more structures in deeper layers (MNRAS paper in preparation)

* Eric

  - Calculating high resolution spectra to study T Tau stars
  - FIES spectrograph (La Palma) test verification run
  - High resolution spectra for transiting planet star
    -> high concentration of Li
  - Photometric and spectroscopic study of new pre-main-sequence eclipsing
    binary with Halpha emission and strong Li absorption
    A movie of spectrum variations and light curve was shown.

* Bertrand

  - 3D hydrodynamic simulations for Betelgeuse (B. Freytag's models)
    and radiative transfer calculations; ELODIE observations will be analysed
    and compared to D. Gray's observations
    Velocity and depth variation of spectral lines are studied.
    Other supergiants: SW Cep, mu Cep
    Correlation between delta v (velocity amplitude) and dust mass loss
    (linear relation in log-log diagram) -> probable link between convection
    and mass loss or dust formation?
    Paper in press
  - Movie of simulations (by PhD student Andrea Chiavassa):
    stellar surface intensity as function of wavelength
    (-> different sizes of granules)
    Observations at VLTI in H band planned (need 3 telescopes)
    + radio observations, CRIRES, NACO 
  - Most massive AGB stars in galaxy
    High resolution optical spectra of sample of stars displaying increasing
    strength of Li I line at 6708Å - fit synthetic spectra to observed spectra
    "Rubidium-rich AGB stars" García-Hernández, García-Lario, Plez,  D'Antona,
    Manchado, Trigo-Rodríguez (2006, Science)

* Christian

  - PhD project: expanding Friedrich's convection code for stellar envelopes
    into stellar atmospheres
    Problems: radiative transfer, computational grid (e.g. change of length
    scales), documentation, porting
  - Using LLmodels code for computation of small grids
    Starting models are generated using ODF approach, then continue with
    line-by-line (saves time)
  - Lambda Bootis stars (with E. Paunzen)
    Colors from models with individual peculiar abundances match observed
    colors
  - Automatizing abundance analysis procedure
    Semi-automatic approach involving helper routines that speed up routine
    work (line selection, fitting calculations to observations)
    Includes graphical interface (Tcl/Tk)
Astronomical Tools Collection 

* Theresa

  PhD project: Exploring Atmospheric structure of chemically peculiar stars
  via magnetic doppler imaging and the case of the roAp star HD 24712
  Using INVERS10 code and NOT observations (spectral polarimeter)
  Zeeman doppler imaging results were shown (distribution and magnetic field
  geometry of Fe, NdIII, NdII, ...).
  Fe and other Fe peak elements: depletion around maximum B
  Nd III and other REEs behave in the opposite way, Ce behaves "abnormal".
  Magnitude of abundance variations is different for different elements
  Analysis works nicely even for blends.
  RV measurements (Ryabchikova, Sabchov et al.) -> phase shifts

* Marlene

  - Did abundance analysis of candidate roAp star HD 3980 (Li spots)
  - Wrote manual for abundance analysis (bachelor's thesis)
  - Starting to do Doppler imaging

* Werner

  Coordination of several projects supported by Austrian funding agencies
    
  - FWF project "The Core of the HR diagram"
    (FWF = Fonds zur Förderung der wissenschaftlichen Forschung
    = Austrian Science Fund)
    Spectroscopic and photometric analyses of stars in the classical
    instability strip + hot and cool giants
    Cool giants: Thomas Kallinger (together with David Guenther)
    Most of money for VALD comes from this project
    
  - FWF project "Stellar magnetism at the main sequence"
    
  - FWF project "Dynamic Stellar Atmospheres: Convection & Pulsation"
    PI: H. Muthsam (Mathematics)
    
  - FFG-ALR project: MOST (3rd year of operation, Vienna ground station)
    (Forschungsförderungsgesellschaft and Agentur für Luft- und Raumfahrt
     = former ASA)

  - BMBWK (Ministry for Education, Science and Culture) project:
    COROT (launched Dec 27, after 8 years of preparation)
    Includes Vienna ground station (backup for French ground station)
    Receiving "zillions" of data
    Need for ground based collaboration is evident
    
  - BRITE-Constellation project
    Nano-satellites dedicated to high-precision photometry (millimag. accuracy)
    2 identical satellites (with blue/red filters,
    FFG-ALR: BRITE-Austria, U Vienna: UNIBRITE)
    Targets are brightest stars (-> massive stars)
    Strong link to stellar evolution and stellar atmospheres
    Second pair of satellites under consideration to be built in Canada
    Launcher to be found

Wed, 28 Feb 2007

Ulrike Heiter [Subject: /Discussion/Term] 

Here are the results of the group discussion on term designation format
at the VALD Workshop (Vienna, Feb. 2007, Paul, Oleg, Ulrike):

Recommendation: follow NIST format

- groups separated by dots = electron configurations 
- brackets around parent configurations
- the term in last group
- asterisk for odd level
- no spaces

- seniority: as in source lists (might be inconsistent -> warn users)

Flags: up to 4 characters (e.g. 'LS', 'J1J2', ...)

Examples:

5797.57  La II (from VALD/Wisconsin list):
LS
5d2.a3F
LS
4f5d.z3G*

5627.497 Fe II (from NIST):
LS
3d6.(3F2).4s.a2F
LS
3d6.(5D).4p.z4F*

5615.644 Fe I (from NIST)
LS
3p6.3d6.(5D).4s.4p.(3P*).z5F*
LS
3p6.3d6.(5D).4s.(6D).5s.e5D

Mon, 26 Feb 2007

Ulrike Heiter [Subject: /Discussion/Software] 


Thanks, Christian and Eric.

> I just added source files from ~/sources/SELECT, ~/sources/PRESELECT, 
> ~/ems/SAFE to the development/tools-source tree. Source files from ~/ems have
>  
> been added to the development/ems tree.
> 
> Many of these files are relatively old, and some of you might have updated 
> versions from last week. Please make the development tree as up-to-date as 
> possible.
> 

I strongly suggest to have only ONE source file for each tool in the
development tree.
For example, for PRESELECT, that would be "preselect3.f" (or whatever
the final name is), the version Friedrich is making by "merging" the
various small patches by Christian into Nik's latest version (the
people involved know what I mean). This will be the version developed
further. 

Cheers,
Ulrike

Eric Stempels [Subject: /Discussion/Software] 

Well done Christian, and great timing!

I just added source files from ~/sources/SELECT, ~/sources/PRESELECT, 
~/ems/SAFE to the development/tools-source tree. Source files from ~/ems have 
been added to the development/ems tree.

Many of these files are relatively old, and some of you might have updated 
versions from last week. Please make the development tree as up-to-date as 
possible.

Development files of the web server were already added on Friday. Everything 
that should be under version control now is. If not, add it using 'svn add 
[file]'.

(Nik, where is the ~/sources/EOS directory?)

Regards,
Eric.

Christian Stütz [Subject: /Discussion/Software] 

Dear all,

I added the Ascii definition and basic filter program to
the development branch:
  ~/development/tools-source/trunk

Details as vald@jan in this directory with 
  /> svn log EDIT/ 

Cheers
         Christian

Christian Stütz [Subject: /Discussion/Database] 

Hello Oleg,

I have updated the Models in the Vienna ems.
The old set of models is archived in ems/ARCHIVE/.
Since ems/MODELS/ is not syncronized right now,
someone should copy the new models to the Uppsala
and Moscow servers.

Cheers
        Christian

Oleg Kochukhov [Subject: /Discussion/Database] 

I have downloaded ODF NEW Atlas9 vmicro=2 km/s models from Kurucz's
website and produced files in krz format. The attached file contains all
available *.mod and *.krz files. It is suggested that these models are
used for "extract stellar" in place of the old Atlas9 models. There are
two main advantages:
   
1) models are computed with updated opacities
2) model headers contain correct solar abundances
   
Cheers, Oleg

Sat, 27 Jan 2007

Werner Weiss [Subject: /Discussion/VO] 

Dear Kolya, et al.,

some time ago we received in Vienna an invitation to consider VALD as 
part of the VO and as this basically is a good idea I was in principle 
in favour for it. The onbious time to discuss this matter is our 
upcoming workshop and Christian pointed out to me that Nik has proposed 
a topic on the same issue. Perfect!

Please find attached an email from Enrice to Ernst which he has 
forwarded to me. Enrice suggests to attend (part of) the workshop and to 
present the EuroVO scenario. I think this is a good idea and I am 
infavour of inviting him for this presentation.

What do you think?

Best regards,
Werner

Fri, 26 Jan 2007

Enrique Solano [Subject: /Discussion/VO] 


Dear Ernst, 

# VALD: OK. If you consider convenient my attendance to the Workshop to
# explain what EuroVO-DCA is and the type of support it can provide to
# projects like VALD or WEBDA, just let me know.

# WEBDA: We can start discussing WEBDA at any time (before, after or in
# parallel to VALD). Tell me what you prefer.

# EuroVO-DCA Workshop: As said in my previous mail,  it would be very
# interesting (for many reasons) to present a poster on the VO compliance of
# VALD during the workshop "Astronomical Spectroscopy and Virtual Observatory"
# that will take place at ESAC next March. The poster could have two parts;
# the first one describing what VALD is (you could take care of that) and the
# second one describing the work to be done for its VO compliance (we 'll do
# that).  Please tell me what you think ASAP as the deadline for registration
# and abstract submission is January 31st.

Best regards, 

Enrique. 



Ernst Paunzen
24/01/2007 10:28 
To: Enrique Solano
Cc:  "Weiss W.W.", Christian Stütz
Subject: EuroVO-DCA WP6 support for VALD 



Dear Enrique, 
  
There will be a Workshop for VALD from 
the 19th to the 23rd of Februray. EuroVO 
will be one of the major topics. I will sent you 
the schedule, aims, etc. after this Workshop. 
  
I hope thats fine with you. 
  
Should we discuss WEBDA after that? 
  
Best regards, Ernst

Thu, 18 Jan 2007

Hampus Nilsson [Subject: /Discussion/Atoms] 

Hi Ulrike,

As Nik told you Sveneric is on sick leave, he is however recovering 
nicely. I was in contact with Nik before Christmas and provided him with 
a bunch of atomic data. We will continue to produce data like: 
oscillator strengths, wavelengths, hyperfine structure, isotopic shifts 
and energy levels, but probably focus more on infrared lines since we 
have a new IR FTS. We are also involved in measurements of VUV lines 
using the MaxLab syncrotron.

Unfortunately nor I or Henrik Hartman have the time to attend the 
workshop. I you need more information don't hesitate to ask me.

Cheers,

//Hampus

Ulrike Heiter wrote:

>Dear Hampus,
>
>I am writing in the name of the team developing and maintaining the
>VALD database (http://ams.astro.univie.ac.at/vald/).
>I have tried to contact Sveneric Johansson before (see below), but Nik
>told me that he is probably still on sick-leave.
>
>Can you give me some information if and what your group could
>contribute to VALD (see e-mail below)?
>
>In the week 19-23 Feb 2007 we will have a workshop on VALD development
>in Vienna. We would be happy if someone of your group could
>participate.
>
>Best regards,
>Ulrike
>
>
>----- Forwarded Message
>
>Date:    Wed, 20 Dec 2006 17:33:20 +0100
>From:    Ulrike Heiter
>To:      "Sveneric Johansson"
>Subject: VALD development workshop Feb 2007
>
>Dear Sveneric,
>
>As I wrote a few weeks ago, we will organize a workshop on the
>continuing development of the VALD database from 19-23 Feb 2007
>in Vienna.
>Will you be able and willing to participate in this workshop? 
>
>I will send further information on the workshop by mid-January.
>Information is also posted on the VALD weblog:
>http://www.astro.uu.se/~ulrike/VALD/Meetings/
>
>In general, are you willing to participate in VALD development?
>As Nik has told me, you agreed to provide Fe II line data
>for inclusion in VALD. 
>Can you give me some information on which data can you provide?
>Also, I would appreciate any comments on any of the topics
>currently under discussion, see the VALD Development Weblog:
>http://www.astro.uu.se/~ulrike/VALD/
>
>Best regards and best wishes for Christmas and the New Year,
>
>Ulrike
>
>
>----- End of Forwarded Message
>
>  
>

Mon, 15 Jan 2007

Bertrand Plez [Subject: /Discussion/Molecules] 


The molecules I have compiled are often not checked at the 
high-resolution spectroscopic
level. Sometimes they are good, sometimes they are not. They were 
primarily compiled/computed
to include in opacities for cool stars. However some turned out, or 
were intently computed in order to,
be of high quality (e.g. CH, which I am currently further iomproving).

I have :
TiO, ZrO, CN, CH, VO, and a few more.
I have plans for C2 and NH.

Thu, 11 Jan 2007

Ulrike Heiter [Subject: /Discussion/Name] 

Incidentally, "vald" in estonian means borough/parish/township, and
"vald" in icelandic means power/authority/force.

Wed, 13 Dec 2006

Eric Stempels [Subject: /Discussion/Web/Interface] 

(Cc: Ulrike, Yury, Christian)

Hi Nik,

With some good help of Yury, VALD in Moscow is up and running now. I installed 
a new, updated version of the web interface that includes the 'via ftp' 
option. Before I announce and link the site from the other mirrors, I would 
like to know if the ems software is up to date and synchronization is active. 
We can then test drive the new software for a short while, before I upgrade 
Vienna and Uppsala too.

I'm asking about the synchronization, because I found some inconsistencies 
between the sites. For example, I found differences between ~/data/default.cfg 
in Moscow and Vienna, with no 'DREAM' lists in Moscow, but with 'GFIRON 
observed' lists. When accessing the personal configuration editor on different 
sites, this will result in different configuration files, and thus to 
different results from different mirror sites. Also, if the provided datasets 
between the sites are different, that is going to give problems.

I'm also not yet completely satisfied with how I'm handling newly added 
linelists. Currently, popups warn the user of removed or added lists. Maybe 
displaying a  page about the recent changes would be better. Also, new lists 
are disabled by default, which may not always be what we want. I'll think 
about that.

Christian and I are trying to work out a scheme to merge the web manual and 
the interface, creating a common layout.

Looking forward to your comments,

Regards,
Eric.

Wed, 22 Nov 2006

Christian Stütz [Subject: /Discussion/Web/Page] 

Hello Ulrike and Eric
 
Thank you for the information on the Servers.

[...]

The first version of the new homepage will be ready in
the next weeks. It is a cgi script written in C, so we
could call the vald script from within it and so end up
with a consistent layout of webpage and webinterface.


Best regards
              Christian

Tue, 14 Nov 2006

Tanya Ryabchikova [Subject: /Discussion/Term] 

Dear Ulli,
    As Kolya told us we may use 96 characters field for both levels in free 
format. The only requirement is to include a separator between levels, '=' 
as Kolya has proposed. Also we have decided to mark a connection type for
each level individually. I'll send you files with this new level 
designation by the end of the week or with Lyuda Mashonkina, who will come
to Uppsala next Tuesday.
                                      Regards, Tanya

Ulrike Heiter [Subject: /Discussion/Term]
Based on the discussion below, the term designation should be stored in the VALD-3 record in three fields: a 2-character flag, and two 40 character fields for lower and upper level. Nik confirmed that this should be managable in terms of storage. See /Format.

Ulrike Heiter [Subject: /Discussion/Web/Page] 

At the Uppsala meeting (Sep. 2006) the VALD webpage was discussed.

It was stressed that the webpage consists of two very distinct parts:

* A manual and documentation for VALD users - "web page"

* The VALD "web interface" as a way for extraction (alternatively to VALD-EMS)

Since the purpose and technical aspects of these two parts are different,
it seems appropriate to separate responsibilities for the parts.

Christian Stütz will be responsible for the web page
(see /Discussion/Documentation/Users), 
Eric Stempels will be responsible for the web interface
(see /Discussion/Web/Interface). 

The web page should only be edited (by Christian) on the Vienna server and
synchronized on the other servers. For this to work, it is important to use
only *relative links* throughout the web page.
It was also suggested to add a "History" section (where one could e.g. explain
the former meaning of VALD). The "News" section should be updated regularly.

Ulrike Heiter [Subject: /Discussion/Synchronization] 

Some notes on synchronization from the Prague (Aug. 2006) and Uppsala
(Sep. 2006) meetings:

The status now is that only user data and statistics are synchronized,
via e-mail every 12 hours.

In the future, several VALD components have to be synchronized.
A combination of rsync and subversion seems to be most appropriate.

The components to be synchronized using rsync would be:
* the data itself (linelists)
* the web page
* maybe the user configuration files.
One should always keep a copy of one previous version and backup copies.

Subversion could be used to synchronize
* the software 
* the user configuration files.
Software synchronization could be problematic because of servers running on
different systems, and the software has to be recompiled on changes.
An alternative would be to only compare the source files regularly and send a
warning by e-mail if changes between servers are found.

Christian and Eric are asked to discuss and implement this.
The following persons outside the VALD team with experience in synchronization
can be contacted:
Barry Smalley   http://www.astro.keele.ac.uk/staff.html
Thomas Marquart http://www.astro.uu.se/people.html

Ulrike Heiter [Subject: /Discussion/Web/Interface] 

At the Uppsala meeting (Sep. 2006) it was discussed that Eric will make the
follwing changes to the web interface (list probably not complete):

* Work in FTP mode instead of e-mail mode

* Include links to documentation

* Add an option to choose energy units

* Adaptation to VALD-3 when all extraction software is ready

Ulrike Heiter [Subject: /Discussion/Name] 

At the Prague meeting (Aug. 2006) it was decided that the name of the
database will from now on be just

VALD

i.e. it's not an acronym anymore, just a name, where the individual characters
do not stand for anything. This approach has been taken by other projects in
the past (e.g. the Gaia space mission).

This discussion topic is now closed.

The VALD webpage should be modified accordingly (e.g. remove the original logo
by Eric, and adapt the text).

Ulrike Heiter [Subject: /Discussion/Communication] 

At the Prague meeting (Aug. 2006), Nik said that communication by e-mail
should be sufficient.

Considering all comments I decided to set up this static Blog webpage,
where I will include all e-mails which seem to be relevant for future
reference, sorted by Subject and Date. Some general information is also
included.

Note that this page is also available as RSS feed (Live Bookmark).

Ulrike Heiter [Subject: /Discussion/Documentation/Users] 

Full references in VALD extraction output:

At the Prague meeting (Aug. 2006), it was decided that Nik will
implement this into the extraction software, and Friedrich will provide
necessary input data (lists of references).

Ulrike Heiter [Subject: /Discussion/Conversion] 


For the format of a VALD-3 record see /Format.

From the Prague meeting (Aug. 2006):

For testing purposes, an additional, independent server will be set up,
running VALD-3 for a few weeks. Testing will be done e.g. on the new
REE line lists (see /Papers).

Ulrike Heiter [Subject: /Discussion/Database] 

At the Prague meeting (Aug. 2006) the following points were mentioned:

* Including data from the new line lists of Kurucz (calculated, Fe peak and
  light elements) would increase the number of lines by about one order of
  magnitude.

* Database organization:
  Lists will be organized by provider (1 list/file per provider),
  and there will be replacement lists (if data from several sources exist
  for one and the same line).
  Replacement lists are described in the VALD-1 paper (Piskunov et al. 
  1995A&AS..112..525P, Section 3.3).

* A VO interface should be available by the end of the year.
  I don't remember who will be working on this.

Mon, 30 Oct 2006

Paul Barklem [Subject: /Discussion/Term] 

Hi Ulrike,

Ok, I promised to write something about thsi so here it is.  My main thinking
in this, is that for calculations based on simple valence electrons models,
such as those that I have been involved in developing for line broadening and
depolarisation, one needs to know:

1) the configuration (i.e. principal and angular momentum quantum numbers) at
least for the valence electron

2) the parent or grandparent state, this is so you can work out the binding
energy of the valence electron, and in the case of depolarisation of complex
atoms one needs the total spin and angular momentum of the core electrons.

3) it can be useful too to have the configurations of any open shells in the
core; for example if there is so called fractional parentage

Therefore my suggestion would be to enable all this information to be included
(I am not saying it should be included, or is easy to find, merely there
should be enough space so it can be preserved if available, which it often is
from NIST).  Whether all this info should be in the same place is a question
for someone else also.  Also, it may make sense to split the configuration and
term information for example into separate entries, for clarity.

So my minimum suggestion would be:

  w   w   G p    w  G p
nl  nl  (a L ) nl  a L
-------  ----- --- ----
open    parent val  LS 
shells         ence term
               shell
	       
where n =principle quantum number
      l =electronic angular momentum
      w =occupation number
      a =seniority
      G = 2S+1, spin multiplicity in LS coupling  
      L = total electron spin in LS coupling
      p = parity
eg.
  
3d5 4s  (a3G ) 4p  t2Fo    (which is a state in Fe II)

Naturally, this only covers LS coupling. A few more bytes would probably be
needed for other coupling schemes, e.g. a typical example in jj coupling:

3d9 (2D5/2)4p3/2 (5/2,3/2)o

Examples for other coupling schemes can be found at
http://physics.nist.gov/Pubs/AtSpec/node09.html

Thus, while it's not perhaps needed, I would advocate adding one more open
shell, since we probably need more bytes for the more complex coupling
schemes.  That's about as complicated as things can possibly get in LS
coupling, for example:

4f7 (8So)5d (9Do)6s (8Do)7s 9Do

(noting we could do away with the grandparent and great-grandparent here).  In
any case, my feeling is that we shouldn't throw away info that may be useful
to someone, unless there is a good argument such as storage space or read-in
speed, etc.  About 35-40 bytes per state (ie. x2 per line) would be enough
based on the above.  If there are strong arguments for smaller size, then the
minimum model presented above (~23 bytes) would be enough from my point of
view.

Cheers, Paul

Fri, 20 Oct 2006

Ulrike Heiter [Subject: /Discussion/Term] 

Hi Paul,

[...]

You can also get the term designation information that is currently in
VALD for individual lines using the showline code

~vald/tools/showline3_2

The standard "data base information file" is /u40/vald/data/pres3_2.dat

The characters at the end of the term designation that you get with
showline are not part of it, but give the source for the line data
(e.g. "KP", "BWL").

I hope this gives you some ideas for a format suggestion.

Cheers,
Ulrike

Wed, 18 Oct 2006

Werner Weiss [Subject: /Meetings] 

Dear All,

According to the resonance to my previous circular I blocked the 
apartment for the time from Monday, Feb.19 to Monday Feb.26, 2007, for 
the VALD workshop. All of you know the quality of the large and small 
****guest rooms at the observatory. Three of you can make use of them.
Presently, end of Feb. is not oversubscribed, hence I can slightly 
shift/extend the reservations if you tell me so within the next days.

I assume that Ulli will organize the structure of the workshop, draft a 
schedule and let us (in Vienna) know in time which computing resources, 
etc. will be needed. Is the mailing list complete?

I hope to meet you all in Vienna!
Cheers,
Werner

Sun, 15 Oct 2006

Tanya Ryabchikova [Subject: /Discussion/Term] 

Ulli,
  I attach a full list with the experimental measurements for REE lines,
where Rimma give a full LS description (according to NIST) of the levels + 
LS term designation 
needed for Lande calculations. If we save it (and who knows how longer it
might be for more heavy elements) than we need more space for term 
description. The alternative - to save only external electrons, like 6p, 
5s, 5d6s, etc. Please, discuss it with Kolya. The sooner we fix
formats, the sooner Rimma remakes all files. We cannot just cut a fixed 
number of positions. In this case some terms may still have a full 
description, while others - some strange cutted one. 

[...]

                                  Regards, Tanya

Wed, 11 Oct 2006

Oleg Kochukhov [Subject: /Discussion/Term] 

Dear Tanya,

If the length of a comment string is not a critical issue, I don't see why
should we stick to 42 characters. May be we can make it longer.

On the other hand, I can think of a number of ways to conserve spacing.

1) Use numerical flag instead "LS" string (e.g. 0=no information, 1=LS)
2) Remove spaces (if we deal with fixed format spaces are not needed)
3) Space between "E" and number that follows is unnecessary even for free
format strings.

Regards, Oleg