daily real-time MHD simulation of global solar corona and solar wind

[ Gallery contents ] [ Publication/Presentation ] [ Daily real-time coronal MHD simulation ]

[ version 1 of this page ] [ version 2 of this page ] [ Japanese version of this page ]

The real-time Sun, the global solar corona and solar wind
derived with time-dependent three-dimensional MHD simulation
using daily-updated SOHO/MDI and SDO/HMI magnetic field synoptic data

Please visit SOI/SOHO/MDI website and HMI website, for more information about instruments, data, and the status of the data processing pipeline.

Schedule : Around 14:00 UT (7 or 8AM PT) everyday, the latest daily-updated synoptic map will be generated in JSOC data pipeline.

synoptic frame map

FITS file (about 20MB)
WSO-like format (about 21kB)
an mpeg movie (35MB; *not regularly generated)

PFSS model plot

left : field lines and bases of open field
right: zero-Br contour surfaces.
an mpeg movie (17MB; *not regularly generated)

At 9 AM PT, the latest near-real-time data is loaded, processed and fed to the coronal MHD model and PFSS calculation code, then the right plot above is replaced with the latest one. Around 10 AM PT, a PFSS calculation of moderate spatial resolution is done and the left plot above is updated. We conduct thee MHD runs for a day, each of which uses the same magnetic field data but has a different sub-Alfvenic boundary treatment. Around 1 to 3 PM PT, each of three MHD runs (low-resolution, about 5-degrees in latitude and longitude) is completed, then the plots below for each run in this page are updated. The observational images from the SOHO and SDO are updated at around 12UT (around 4 or 5AM, PT).
*) The schedule times are subject to change, depending on status of the JSOC data process. Elapsed time of MHD runs depends on the CFL condition (i.e., the largest value of the magnetic field strength in an input synoptic map data).

Simulation with choice O : The density and temperature are fixed at the base of the open-field coronal hole.


plots made: 
Tue Jun 18 20:32:44 UTC 2013 (Tue Jun 18 13:32:44 PDT 2013)

lat-lon map at various distances click here to enlarge.	line-of-sight integration of coronal density

recent 27 plots (AniS java applet animation) year-to-date (mpeg. *not regularly generated)	recent 27 plots(AniS) recent 27 plots(mpeg)

Latest (with test HMI data):
	*Please contact us for details of the simulation and data: The latest CSV and FITS data files above are made from low-resolution simulation data.
	maps at 30 Rs: Vr (in km/s) and Br (in nT) in CSV (Comma Separated Values; plain text) format
	cubic data of 8 MHD variables : click here (FITS format, approx. 6.5MB). The size of this cubic data is 32 x 64 x 640. The first 80-layer block, (,,1:80), is for the density (in count/cc), from 1.01 Rs to 4.96 Rs with a fixed equi-radius interval of 0.05Rs, and the second to eighth 80-layer blocks are respectively for the temperature (in MK), V_r, V_theta, V_phi (in km/s), B_r, B_theta and B_phi (in Gauss).
Summary mpeg movies
	with SDO/HMI (test) data, since July 1. 2011, * without polar field correction, thus somewhat unstable.
	2nd half of 2012 :	lat.-long. map (30MB), Corona and CH-base (2MB)
	1st half of 2012 :	lat.-long. map (30MB), Corona and CH-base (2MB)
	2nd half of 2011 :	lat.-long. map (30MB), Corona and CH-base (2MB)
	transition phase, from MDI to HMI data as input. * MHD runs were very unstable, with less frequently-updated data
	1st half of 2011 :	lat.-long. map (30MB), Corona and CH-base (2MB)
	with SOHO/MDI data, from Feb. 08, 2008 Dec.31, 2010
	2nd half of 2010 :	lat.-long. map (30MB), Corona and CH-base (2MB)
	1st half of 2010 :	lat.-long. map (30MB), Corona and CH-base (2MB)
	2nd half of 2009 :	lat.-long. map (30MB), Corona and CH-base (2MB)
	1st half of 2009 :	lat.-long. map (30MB), Corona and CH-base (2MB)
	part of 2008 :	lat.-long. map (50MB), Corona and CH-base (2.5MB) (* without polar-field correction for the first 21 days, then with polar-field correction)

Observations / Measurements


image/data taken: 
Tue Jun 18 11:30:19 UTC 2013 (Tue Jun 18 04:30:19 PDT 2013)

Simulations with other choices of boundary treatment (c.f. Hayashi (2005) ApJS vol. 161 pp. 480)
The daily quasi-real-time simulation shown above assumes fixed density and temperature at the coronal hole base (labeled choice O in my paper (2005)). The plot(s) below show the simulation results with other choices.

With choice AB' : Density and bulk velocity at the base of the coronal hole can vary under the constraints; the fixed temperature and the limited mass flux.


plots made: 
Tue Jun 18 21:46:14 UTC 2013 (Tue Jun 18 14:46:14 PDT 2013)

*) The 10-Rs values of the corona simulated with choice AB' may not be of the time-relaxed one: This choice allows the boundary coronal-hole density to decrease, and it takes another extra 40 hours for the low-density wind to reach that height.


recent 27 plots(AniS) year-to-date plots(mpeg)	recent 27 plots(AniS) recent 27 plots(mpeg)

With choice A : Density, temperature and bulk velocity at the base of the coronal hole can vary under the constraints; the fixed specific enthalpy and the limited mass flux.


plots made: 
Tue Jun 18 20:39:07 UTC 2013 (Tue Jun 18 13:39:07 PDT 2013)

*) The 10-Rs values of the corona simulated with choice A may not be of the time-relaxed one: This choice allows the boundary coronal-hole temperature to decrease, and it takes another extra 40 hours or longer for the low-temperature, slow-speed wind to reach that height.


recent 27 plots(AniS) year-to-date plots(mpeg)	recent 27 plots(AniS) recent 27 plots(mpeg)

Modification / change / history

currently : Tweak scheduling.
Jan. 16, 2012 : Start using SDO/HMI daily-updated synoptic maps.
July 1, : Start using test version of SDO/HMI daily-updated synoptic maps.
Jan. 2011 : SOHO/MDI measurement ceased.
July 1 : Newkirk filter and other modifications are applied to synthetic LoS images for choice O.
June 15 : Newkirk filter and other modifications are applied to synthetic LoS images for choice A and AB'.
May 11 : This page was made publicly-accessible.
May 1 : Simulations with boundary treatment choice A and AB' are started.
Apr. 21 : Test plots of line-of-sight coronal density integration are added.
Mar. 1 : Polar field correction is introduced.
Feb. 8, 2008 : Daily simulation starts.

Description

general :
The simulation is scheduled to start 1 AM Pacific Time. It usually takes about 6 hours (in wall clock time, depending on the computational choices such as the boundary treatment and the computation resources available) to conduct this near real-time MHD solution. New plots may appear around 7 AM (P.T.) at earliest.
data input :
The simulation uses the EOF (Experiment Operations Facility) version of SOHO/MDI solar surface magnetic field synoptic data so that the recent state of the corona can be determined. The EOF magnetic field map is regularly updated using the measurement data made 24 hours prior, and the dashed rectangle in the plots above approximates the Earth-side heliographic region for which the data is updated.
The simulation script uses "wget" command to get the magnetic field data that is located at the web-accessible place above, and, for sake of maintaining the regular basis, uses the second newest synoptic data that contains the measurement data from 24 hours to 28 days prior to the start time of the simulation.
polar field correction :
Because of variety of factors such as the apparent inclination of the solar rotation axis, the measurement of the polar field is generally of low quality. The polar field correction is needed to to enhance studies on the solar corona and solar wind.
Our correction is based on the spline extrapolation with respect to both time and space, proposed and tested by X. Sun, Y. Liu, J.T. Hoeksema, and X.P. Zhao. See details in the paper.
initial and boundary values of magnetic field :
The boundary value and the three-dimensional initial value of the magnetic field are then calculated with the potential-field source-surface (PFSS) model. In this daily simulation, the spherical harmonics decomposition is used in the PFSS model, and the polynomial up to 5th order is used to give the initial and boundary values.
MHD simulation :
To obtain the steady state of the sub/trans-alfvénic solar corona, the time-dependent three-dimensional MHD simulation is conducted. The simulation scheme here is based on the Lax-Friedrichs method that is a simplified version of the TVD Roe/Godunov method; the Lax-Friedrichs method is much much faster than the Roe/Godunov method. The number of grids are 72, 32 and 64 for the radial, latitudinal and longitudinal direction, respectively. The inner boundary of the simulation is set at 1.01 solar radii, so that we will handle the coronal plasma already heated up to the order of 1 MK though the solar transition region. The daily MHD simulation will solve the time-relaxation process of the sub/trans-alfvénic solar corona over 40 hours.
   The boundary values of the eight MHD variables at this height are controlled by using the normal projected characteristics method, so that the obtained time-relaxed solution can be fully consistent with the given boundary map, the basic nonlinear hyperbolic MHD equations, and the concept of the characteristics. This is a quite capable mathematical method applicable to any computation scheme; this method re-distribute the calculated temporal derivatives in accordance with the given constraints and the direction of the characteristics curvature. In this method, all three components of plasma flow and the two tangential components of magnetic field can vary in accordance with the concept of the characteristics. The radial component of surface magnetic field is fixed all through the computation with any choices above.
   Some choices are selected and used in the daily MHD simulations to treat the open field region (coronal hole). The choices other than O limit the mass flux at the coronal hole base. For the closed-field region (base of coronal streamer), the plasma density and temperature can vary responding to the plasma precipitation from the stagnant closed field regions. Whether the plasma will be flowing outward or stagnant is determined by the simulated nonlinear interaction with the magnetic field, thus small or weak closed-field can be open up to the interplanetary space.
   The polytrope plasma with specific heat ratio of 1.05 is assumed to mimic the near-isothermal coronal situation due to the high thermal conduction. This assumption yields somehow small contrast in flow speed and density, and the absolute values of plasma (velocity and density) are less realistic. However, the magnetic field quantities and the trends of these plasma variables would be well retrieved. Because of the simplicity, we assumed the polytrope. Note that the specific enthalpy is a conservative quantities along the stream and that the values differ from place to place.
quick links to related sites : CISM, SDO / HMI, SOHO / MDI

K. Hayashi
last (minor) modification: May 6, 2012
The last major update/modification: Jan.15, 2012
The first simulation plots appeared on Feb. 8, 2008
The first version of this page was created on Feb. 1., 2008

The real-time Sun, the global solar corona and solar wind derived with time-dependent three-dimensional MHD simulation using daily-updated SOHO/MDI and SDO/HMI magnetic field synoptic data

The real-time Sun, the global solar corona and solar wind
derived with time-dependent three-dimensional MHD simulation
using daily-updated SOHO/MDI and SDO/HMI magnetic field synoptic data