The CISM Code Infrastructure

1
The CISM Code Infrastructure
Bob Weigel
The CISM Knowledge Transfer Short Course AFWA
Omaha, November 2-3, 2005
2
Outline

• Sun to Earth Numerical Modeling
• Model Overview
• Pairwise Couplings
• Coronal and Heliosphere Model Coupling
• Magnetosphere and Thermosphere-Ionosphere
  Coupling
• Magnetosphere and Inner Magnetosphere Model
  Coupling
• The CISM Forecast Model
• Visualization Products

3
CISM Heliospheric Models (CORHEL 1.0)
Solar Corona Model Magnetohydrodynamics Around a
Sphere (MAS) Linker et al., 1999 Mikic et al.,
1999 Riley et al., 2001
Solar Wind Model (ENLIL - Sumerian wind
deity) Odstrcil and Pizzo, 1999 Odstrcil et al.,
2002
4
CISM Geospace Models (LTR 1.0)
Magnetosphere Model Lyon-Fedder-Mobarry (LFM) Fedd
er and Lyon, 1995 Lyon et al., 1998 2004
Rice Convection Model (RCM) Wolf and Spiro, 1987
Toffoletto et al., 2004
Thermosphere/Ionosphere Model Thermosphere-Ionosph
ere Nested Grid (TING) Roble et al., 1988 Wang
et al., 1999
5
Numerical Model Chain (CISM 1.0)
Solar Corona Model (MAS)
B, V, ?, p
Solar Wind Model (ENLIL)
B, V, ?, T
Magnetosphere Model (LFM)
J, ?, T
?
B, ?, p
?, p
Magnetosphere- Ionosphere Coupler
Inner Magnetosphere Model (RCM)
?, ?P, ?H
?, ?, F
?P, ?H
Thermosphere/Ionosphere Model (TING)
6
CISM Numerical Code Coupling Framework

• Need to develop programming paradigm which allows
  for efficient coupling of models and is flexible
  enough to allow adding new physics and models
• efficient transmission of information among codes
• interpolation of data between grids
• translation of physical variables between codes
• control mechanisms to synchronize execution and
  interaction between codes
• minimal modifications to existing code base
• Intercomm University of Maryland A. Sussman
• Solution to the MxN problem in coupling parallel
  codes
• Will address the control issues
• Overture - LLNL B. Henshaw D. Quinlan
• C framework for solving differential equations
  on overset grids
• Used to handle interpolation between model grids

7
Framework Example

• Overture grid functions identify regions of
  overlap between models as well an interpolatant
• Intercomm creates data descriptor to determine
  how to move data in regions of overlap between
  coupler and models A B

• Applied Intercomm library to test case based upon
  splitting 2 fluid MHD simulation
• Overture application being developed to replace
  coupling infrastructure in CMIT

8
The Forecast Model Concept
(From Baker et al., 2004))
9
CISM_DX
a community-developed set of data, models, and
data and model explorers
10
CISM_DX

• Needed to create a visualization and data
  analysis tool which could be used by validation
  team as well as model developers
• Used Open DX and Octave as a basis
• Import modules currently implemented
• MAS, ENLIL, LFM, TING, and WSA
• RCM and TIEGCM are under development
• Also includes and extensive collection of
  observations and tools for calculating metrics
  and skill scores
• KT thrust is developing push button
  installation process
• includes examples and documentation
• Available for download from http//lasp.colorado.e
  du/cism/CISM_DX
• User group with mailing list and network
  archive

11
CISM_DX Lead Developers
Michael Wiltberger NCAR/HAO Michael Gehmeyr -
CU/LASP and NOAA/SEC Bob Weigel - CU/LASP
12
CISM_DX Contributors
Fabio Acero Scot Elkington Tim Guild Elly
Huang John Lyon Bob McPherron Asher
Pembroke Marius Schamschula Dimitris Vassiliadis
Logo by Susanna Lamey
13
What

• A collection of analysis and visualization tools
  by scientists, that use, or can be compiled with,
  Open Source Software on multiple operating
  systems. The tools are integrated.

• CISM-DX is a software package that contains
• Extensions to Octave and OpenDX for import and
  analysis of numerical and measurement data
• Extensive suite of networks and programs for both
  novice and expert users
• Extensive set of time series data in a common
  format

14
What
15
Integrated?

• It is not a web page with a list packages in
  source code format. These exist. It is not a
  web page for obtaining a restricted set of
  information.
• It is not a set of libraries that is called by a
  Java or other front-end (examples VisBARD, IDV,
  SSC 3D Orbit Viewer, AFGeospace). Although it
  could be.
• It is a set of source codes, pre-compiled for
  several operating systems, usage examples,
  educational tutorials, and programs that
  integrate the codes written in different
  languages. What calls the library (or the
  front-end) depends on the community - the
  library is separated from the GUI the libraries
  are integrated so that the GUI can easily change!

16
Code distributed with CISM_DX

• 50 .m files compatible with Octave/Matlab
• 30 C/Fortran codes wrapped with OpenDX or
  Octave
• 50 .net files for visualization in OpenDX
• 1000 years (1 GB compressed) of time series
  data _at_ 1-minute resolution
• 2 .pro files for IDL
• 10 .pl files for data file manipulation with
  Perl

17
Datasets distributed with CISM_DX

• Numerous data sets from across the community
  where collected moved on uniformly spaced time
  grids and stored in HDF files
• Provides the user with essential data quickly so
  they can move onto doing science with having to
  learn new data formats

• Dataset includes
• Satellite Observations
• ACE,WIND,GEOTAIL,GOES
• Ground Magnetometer
• WDC, FMI
• Indices
• AE, Dst, Ap, F10.7, aa, am

18
Why

• Space physics needs a tool for visualizing and
  analyzing numerical and empirical models as well
  as observational data
• Must support all levels of users from novice to
  expert
• Must be inexpensive
• Should draw upon the previous work and knowledge
  of the space physics community

19
Why continued

• It has been successful in CISM
• In allowing scientists to have access to
  Sun-Earth analysis programs without having to
  learn specialized data formats and data sets
• In getting students working with expert tools
• In speeding up model transition into operations
• The number of data formats and programming
  languages scale with time. Integration can often
  be more important than optimization (of data
  formats, download speed, etc.) Without a
  framework, this bottom-up approach is the best
  alternative (and some may say better).

20
The CISM_DX contribution framework
Our approach
Give researchers and students a useful tool and a
mechanism and framework for improving upon it.
21
The CISM_DX contribution framework

• User develops tool using software in the CISM_DX
  package.
• Submits a contribution package
• Package is review and suggestions for improvement
  are made
• Final package is posted to CISM_DX web site and
  announced on email list.
• Developers meet to decide what and how parts of
  contribution can be integrated in a future
  CISM_DX release.

• Ideally all contributions would be in Octave,
  OpenDX, C, Fortran 77, Perl, anything that has
  cross-platform support and can be executed,
  compiled, or wrapped with Open Source Software.
• However, there are many useful IDL and
  Matlab-specific codes that we would consider
  adding (especially graphics tools). They may not
  be as well-integrated as the other programs, but
  will still find use.

22
Web page http//lasp.colorado.edu/cism/CISM_DX
Source code and pre-compiled binaries
available for version 0.43. Version 0.5 expected
in the end of June.

• OSX and Windows XP versions in August.