Skip to content

Introduction to Planetary Atmospheres at Leicester

Jump to bottom
Leigh Fletcher edited this page Nov 17, 2021 · 4 revisions

Introduction

Leicester’s planetary atmospheres team is part of the Planetary Science Group, and specialises in the acquisition and interpretation of planetary infrared spectroscopy from a range of ground-based and space-based facilities. Whether you’re here for undergraduate projects, PhDs or starting as new postdocs, this quick-start guide should help you get to grips with the infrastructure.

You’ll be working under Linux on the high performance computing facilities at Leicester, known as ALICE (for staff and PhD students) or SPECTRE (for undergraduates). Help and assistance on these systems can be sought from the research group to start with, or by emailing rcs.support@le.ac.uk. Your username will be used to provide you access to the research groups file system as soon as you arrive.

It is likely that you will be working with data from observatories and spacecraft using IDL (Interactive Data Language). IDL is used for both reduction of data and for interpretation of the results, and can be started from the command line by typing ‘idl’ (once licensing has been sorted out).

Alternatively, you may be working with data reduction via Python. We offer both JupyterHub via https://spectre15.rcs.le.ac.uk for Leicester locals, or you can install your own version of Python 3 (we recommend Anaconda).

More advanced users will also be using the FORTRAN-based NEMESIS code (Non-Linear Optimal Estimator for Multi-Variate Spectral Analysis). NEMESIS was developed at the University of Oxford by Professor Patrick Irwin and colleagues, and is now used across the world to model, fit and interpret planetary spectroscopy, from solar system planets to exoplanets and brown dwarfs. You will sometimes see this referred to as Radtrans, which is the underlying spectral simulation model, the ‘engine’ of NEMESIS.

A key rule: please do not make changes to any files on /data/nemesis unless given permission to do so. It is generally better to copy files to your homespace before working on them.

Getting Started

Access to ALICE/SPECTRE:

Using BASH:

  • I recommend you use BASH as the main shell. This means having a file in your home directory (/home/u/username/) called .bashrc. The ‘dot’ in front of the file name means it is hidden. The basic contents of .bashrc are found here:

  • I recommend you copy the following files to your homespace as follows:

  • cp /data/nemesis/startup/bash_profile_fletcher ~/.bash_profile

  • cp /data/nemesis/startup/bashrc_fletcher ~/.bashrc

  • cp /data/nemesis/startup/flexlmrc ~/.flexlmrc

  • You can then edit the .bashrc file as you see fit. This lets you load commonly-used software (e.g. module load idl), set up the PATH variable (which tells the system where to look for particular software files), and set up aliases (i.e., instead of typing out a command in full, you can alias the letter ‘m’ to the ‘more’ command to read the contents of a text file).

  • Once these files have all been copied across, try opening a NEW terminal window (which executes your new .bash scripts), and typing idl - it should open without complaining about licenses.

Text Editors:

Typing text into a window is the most common thing that you’ll be doing, and there are a wide variety of editors to help you. EMACS and VIM (or GVIM for a graphical user interface) are the oldest players, but have a steep learning curve associated with them. ATOM is an open source editor which looks great but can be rather slow as it’s big. NANO goes the opposite way by being small and only accessible via the terminal window. My favourites are in between: NEDIT and JEDIT (the latter of which requires JAVA via module load java/1.8), which open graphical windows but are very simple.

IDL:

You should be able to open IDL from the command line by typing ‘idl’. You design small scripts called *.pro files, and then execute them by typing: IDL> .r test.pro …where ‘test’ is the code name. Type IDL introduction into Google for hundreds of examples of beginners tutorials to get you started.

If you followed the instructions above, then the following entries should be in your .bashrc file to get started (i.e., no need to do them again):

module load idl

export LM_LICENCE_FILE=1705@license.rcs.le.ac.uk

export IDL_PATH=$IDL_DIR/lib

export IDL_PATH=$IDL_PATH:+/data/nemesis/idl

If these are in place but you still have problems starting IDL, talk to Dr. Fletcher about getting your license up and running.

A repository of useful IDL codes can be found here:

/data/nemesis/idl

Please do not modify the contents of these directories - if you need to change anything, then copy them to your system before making modifications.

Interactive Sessions

  • Do not run large software sessions on the login nodes (such as large IDL jobs or NEMESIS). For these, you should set up interactive sessions, which means you’re reserving space on one compute node for a set period of time, allocating a set amount of virtual memory. In the .bashrc file there are two types of sessions defined:

  • alias qint='qsub -I -X -l walltime=04:00:00 -l pvmem=1gb -l nodes=1:ppn=1'

  • alias bigqint='qsub -I -X -l walltime=24:00:00 -l pvmem=4gb -l nodes=1:ppn=1'

  • So just type qint or bigqint at the command line and wait for the session to open up, from which you can then use it as a normal Linux machine.

Data Storage and File Management

We each have our own homespace and a shared filespace for the project (/data/nemesis/).

  • Do not store anything important in your homespace as it will not be backed up.

  • Do not change any files in /data/nemesis/ unless they are within your own personal file storage. That’s particularly true of:

  • /data/nemesis/GBdata – all our ground-based data archives

  • /data/nemesis/idl – IDL codes for use by the whole group.

  • /data/nemesis/radtrancode – Source code for NEMESIS and Radtrans

  • /data/nemesis/specdata – Sources of spectral line data

The best course of action is to set up your own directory /data/nemesis/username/, and organise your files within it, copying over directories (e.g. from GBdata) as necessary. If you find that you don’t have permissions to access a particular directory, contact me.

Data Reduction Manager

The Data Reduction Manager (DRM) is a widget-based IDL code that allows us to reduce planetary images. Within this software package are tools to process raw data, assign latitudes/longitudes to the pixels in the image, create cylindrical and polar maps, and to absolutely calibrate the images (i.e., convert raw counts into planetary fluxes and brightness temperatures). To ensure DRM is working on your system, check that the following parameters are defined in your *.bashrc file (they should be if you followed the instructions above).

export DRMDIR='/data/nemesis/drm'

export IDL_PATH=$IDL_PATH:+$DRMDIR/current/

export IDL_PATH=$IDL_PATH:+$DRMDIR/programs/

export IDL_PATH=$IDL_PATH:+$DRMDIR/lib/

export PATH=$PATH:$DRMDIR/bin

export PATH=$PATH:$DRMDIR/current

export DRM_PATH=$DRMDIR/current/

export DRM_DATA_PATH=$DRMDIR/data/

Once defined, type: > idl drm_new ...and you should see a series of subroutines loaded and a file-browser window opened up. Be aware that this software is not particularly user friendly, but does allow you to view and manipulate the *.fits files (flexible image transport system = FITS) in the GBdata directories.

For background material on DRM please see Fletcher et al. (2009), Icarus

Particular instructions for DRM usage can be found in our Introduction to DRM. Note that not all data types will be supported by DRM, and that some (e.g., VISIR on the VLT) have their own codes associated with their reductions.

Clone this wiki locally