Christer Sandin, PhD

Current address:

Astrophysikalisches Institut Potsdam; AIP
An der Sternwarte 16
D-144 82 Potsdam, GERMANY

1.9.2005-
(1.9.2003-30.4.2004)

Previous address:

Department of Astronomy and Space Physics; UAO
Uppsala University
Regementsvägen 1
S-751 20 Uppsala, SWEDEN

1.9.1998-21.5.2003

Quick links

You can find my list of publications here.
You will find everything about the data-reduction tool p3d here.

About my work and skills

I work with projects ranging from theory and modeling, observations, to writing tools for data reduction and data analysis. Here follows a brief summary of my work:

Theory: In my PhD project I worked with physical and numerical extensions of a radiation hydrodynamic model of evolved cool stars. The main goal of my work was to self-consistently include a description of how dust moves with respect to gas in these models. In order to be able to do this I also developed a unique – and automatic – analysis tool. The large amount of physics, which all must be simultaneously appropriately described, makes these models extremely complex to build, which is why no models exist (yet) with similar capabilities. For further details, see this page.
- Keywords: problem solving, thermodynamics, radiative transfer, hydrodynamics (gas dynamics, CFD, aerodynamics, radiation hydrodynamics), combustion processes (reactive flows), multiphase flows.

Observations: In the stellar physics group at AIP I have worked with studies of faint halos of planetary nebulae (PNe), both in the Milky Way and in other galaxies. The main result of this work has been to provide estimates of the final mass loss evolution of cool stars – this is unique work that for the first time allows the mass loss process of this decisive stage of stellar evolution to be explored. This project also provides a good basis for comparisons with my theoretical work.
- In this work we solely used the method of Integral Field Spectroscopy (IFS, that is also referred to as 3D spectroscopy, since this method works well with faint and extended objects, such as halos of PNe.
- In this project I have worked with the following instruments (telescope – instrument – integral field unit [IFU]):
  - Calar Alto 3.5m – PMAS – LARR & PPAK
  - VLT UT2 – FLAMES/GIRAFFE – ARGUS
  - VLT UT3 – VIMOS
  - McDonald, Harlan J. Smith, 2.7m – VIRUS-P. Although the telescope main mirror has a few bullet holes this is a very competent instrument with a very good light collecting power.
- In order to analyze the resulting data I have developed new instrument independent software – ifsfit. The plan is to make a public GPL release of this software as soon as I get some time to do this work. The software includes the following features:
  - Automatic line fitting of any number of spectra and lines (including continuum); with simultaneous subtraction of telluric lines (when present in the spectra). The code is based on Craig Markwardt’s routine MPFIT.
  - Arbitrary binning of spectra distributed in two-dimensions – allowing an increase of the signal-to-noise in the data.
  - Correction for differential atmospheric refraction (DAR) – allowing comparison of lines of different wavelength also at high airmass. (Note that correcting for DAR only makes sense with IFUs where the spatial elements are not separated from each other.)
- Keywords: observations, data reduction and analysis, modeling (with photoionization models), method development.

Data reduction: I have finished a general tool for reduction of data of fiber-fed integral-field spectrographs (IFS). The tool is called p3d. In comparison to other pipelines p3d is designed to be both general and easy to use, which should appeal also to non-experts. The paper about p3d is freely available from A&A, and the tool can be downloaded from the project web site that is hosted by sourceforge.net.
- p3d is a general tool with the following highlights:
  - Full logging of all performed tasks.
  - Graphical presentation of intermediate and final results.
  - Calculation of full error propagation through all steps.
  - Support for any kind of fiber-fed IFU (although it has been setup for four instruments so far: PMAS/LARR, PMAS/PPAK, VIRUS-P, and SPIRAL).
  - p3d is based upon the coding platform IDL. The code is thereby fast and works, without changes, on Linux, Solaris, MacOS, and Windows platforms.
  - p3d can be used with the IDL Virtual Machine (through automatic scripts that are provided). In this case no IDL-license is required to use p3d.
- Keywords: problem solving, numerical methods, programming, method development.

More information related to me

For more information about me, click here.
For various useful links, click here.

Useful computing-related information

If your remote machine is behind a thick firewall it is necessary to use an ssh tunnel when logging in or copying a file from that machine. Here is some information on how to do that. I use the following notation: RUSER is your login name on the remote machine, you do not have to use it if it is the same as on your local machine; MACHINE is the name of your remote machine (for example: star.aip.de); DMZ is the name of the remote login computer (for example: starport.aip.de), which is not the same computer as your remote machine; I use the port number 50809, you may use any other port, if necessary; <file> is the name of the file you want to transfer, with the full path if it is not available in the current directory; <path> is the path on the remote machine.

At first, from your local machine, open an ssh tunnel using the following command:

ssh -l RUSER -L 50809:MACHINE:22 DMZ

Now you can log on to the remote machine from your local computer using (add the -X or -Y flag if you need to be able to open X-applications from your remote machine – yes, in this case X11 forwarding needs to be switched on at the remote machine):

ssh -p 50809 -l RUSER localhost

If you instead need to copy one or several files to or from the remote machine you can use (yes, change the order of <file> and RUSER@localhost:<path> if you want to copy the file(s) the other way):

scp -P 50809 -p <file> RUSER@localhost:<path>

In fact, it is better to use rsync since it can continue an interrupted transfer without starting from scratch (it is also possible to continue an interrupted scp-transfer using this approach):

rsync --partial --progress --rsh="ssh -p 50809 -l RUSER" <file> localhost:<path>

The procedure is very similar if you want to use vnc to connect to computer behind a firewall using a remote login machine. At first, make sure that you have started the vnc server on your remote MACHINE using a command such as the following:

vncserver -geometry 1280x1024

Thereafter, from your local machine, open an ssh port forwarding tunnel using the following command (assuming that you will use port 5901):

ssh -l RUSER -N -f -L 5901:MACHINE:5901 DMZ

Finally, start the vnc session on your local machine:

vncviewer localhost:1

Good luck!

Astronomy Picture of the Day (follow link for all details):

Trees, Stars, Aurora!