Solar Wind and CMEs with the Space Weather Modeling Framework Bart van der Holst C. Downs, F. Fang, R. Oran W. Manchester, I. Sokolov, G. Toth, T. Gombosi Center for Space Environment Modeling University of Michigan http://csem.engin.umich.edu

Outline Space Weather Modeling Framework New: Lower Corona model New: Corona model with Alfvén wave turbulence New: Flux Emergence in Convection Zone model Outlook: Couple it all together B. van der Holst http://csem.engin.umich.edu

Space Weather Modeling Framework Synoptic Magnetograms Flare/CME Observations Upstream Monitors Radars Magnetometers In-situ F10.7 Flux Gravity Waves SWMF Control & Infrastructure Eruption Generator 3D Corona 3D Inner Heliosphere Global Magnetosphere Polar Wind Inner Ionospheric Electrodynamics Thermosphere & Ionosphere Energetic Particles Plasmasphere Radiation Belts Lower Atmosphere 3D Outer couplers

Space Weather Modeling Framework

Quantitative Comparison with SOHO for the October 2003 Halloween Storm Out of equilibrium flux rope superposed on a steady state MHD corona Total Mass, Observed =1.5x1016 gm Model = 2.2x1016 gm

Observed vs. Simulated Solar Wind and IMF near Earth (10o off) vX [km/s] vX [km/s] n [cm-3] n [cm-3] Simulated CME arrives 4.8 hours early. Solar wind speed is almost perfect! Density is within range. Magnetic field variation has right amplitude but sign and phase differ. lg T [K] lg T [K] BX [nT] BX [nT] BY [nT] BY [nT] BZ [nT] BZ [nT] Hours from 00:00 Oct 29 6

Need for improvement Coronal model was based on a reduced, spatially varying adiabatic index for the heating and acceleration of the plasma. In good agreement with the ambient solar wind observed at 1AU, but it is not self-consistent and distorts the physics, especially at CME shocks. CME is not self-consistent, but a superposed flux rope model.

Lower Corona Improved thermodynamics: Electron heat conduction Optically thin radiative losses Exponential heating function, in active regions transition to heating proportional to magnetic field strength Chromosphere boundary (T=20000K, ne=1012) Resolve transition region Ability to synthesize line-of-sight images in the EUV and X-ray regime

EUV and X-ray LOS images for CR1913 EIT 171Å EIT 195Å EIT 284Å SXT AlMg Old SC model synthesis New LC model synthesis Observation: Aug 27, 1997

Streamer structure for CR1913 PFSS model Old SC model New LC model Top view Earth Centered view Closed field lines structures are stressed by heating near surface and subsequent redistribution via heat conduction

SXT response for CR1913 3D topology of T=1.6MK LOS synthesis of SXT AlMg response Connection of the hot, closed field line regions overlay the inversion lines on the surface and correspond to the x-ray emission.

Alfvén Wave Turbulence Alfvén wave turbulence has been suggested in the past as a possible mechanism both to heat and accelerate the solar wind plasma. Hinode observations suggest that the energy input is sufficient for heating and acceleration (e.g. de Pontiue et al. 2007). Wind acceleration: work done by wave pressure force Coronal heating: Alfvén wave dissipation at the ion cyclotron resonance.

4D Alfvén wave model Wave kinetic approach for describing the transport of Alfvén waves in a background MHD plasma. The intensity of the (low frequency) Alfvén wave turbulence is a 4D (3D space, 1D frequency) problem: Advection in Space Advection in frequency Dissipation at Ion Cyclotron 2-way coupling of wave turbulence to MHD plasma background: Wave pressure deposit momentum to the background flow background solution determines the wave propagation and spectral evolution.

Model input Magnetogram driven potential field extrapolation. Spectrum: user input, but in current work we assume a Kolmogorov spectrum as the initial condition, i.e. Specify Alfvén wave pressure at the solar base: Given solar wind expansion factor from PFSS model, Given velocity at 1AU based on WSA model, Impose energy conservation along the field lines to determine the Alfvén wave pressure at 1RS (Suzuki, 2006). In-going Alfvén waves at 1RS are absorbed.

Result for CR2029 Wave dissipation not yet in place, so that: (1) Solar wind speed too high (2) Conversion rate of wave energy into heating is too low Compared to previous SC model (spatially reduced adiabatic index), we can now describe the corona self-consistent, without distorting the shock wave compression ratio

Convection Zone Atmosphere with coronal heating and radiative losses (Abbett, 2007) OPAL EOS Vertical velocity at photospheric level shown in gray scale Magnetic field concentrated in the intergranular lanes

Flux Emergence In the Corona Buoyant flux rope initially at 1500km below the photosphere 3D view at 2 minutes after initialization

Flux Emergence In the Corona Convective down flow tries to drag flux rope downwards, up flow helps the flux to emerge, thus segmentation of rope

Outlook New radiation MHD package for Convection Zone Coupling of Convection Zone to Global Heliosphere Allows physically consistent CME eruptions into a physically consistent solar wind