Photospheric and heliospheric magnetic fields

The magnetic field in the heliosphere evolves in response to the photospheric field at its base. This evolution, together with the rotation of the Sun, drives space weather through the continually changing conditions of the solar wind and the magnetic field embedded within it. We combine observations and simulations to investigate the sources of the heliospheric field from 1996 to 2001. Our algorithms assimilate SOHO/MDI magnetograms into a flux-dispersal model, showing the evolving field on the full sphere with an unprecedented duration of 5.5 yr and temporal resolution of 6 hr. We demonstrate that acoustic far-side imaging can be successfully used to estimate the location and magnitude of large active regions well before they become visible on the solar disk. The results from our assimilation model, complemented with a potential-field source-surface model for the coronal and inner-heliospheric magnetic fields, match Yohkoh/SXT and KPNO/He 10830 Angstrom coronal hole boundaries quite well. Even subject to the simplification of a uniform, steady solar wind from the source surface outward, our model matches the polarity of the interplanetary magnetic field (IMF) at Earth similar to83% of the time during the period 1997-2001 ( independent of whether far-side acoustic data are incorporated into the simulation). We find that around cycle maximum, the IMF originates typically in a dozen disjoint regions. Whereas active regions are often ignored as a source for the IMF, the fraction of the IMF that connects to magnetic plage with absolute flux densities exceeding 50 Mx cm(-2) increases from less than or similar to 10% at cycle minimum up to 30-50% at cycle maximum, with even direct connections between sunspots and the heliosphere. For the overall heliospheric field, these fractions are less than or similar to1% to 20-30%, respectively. Two case studies based on high-resolution TRACE observations support the direct connection of the IMF to magnetic plage, and even to sunspots. Parallel to the data assimilation, we run a pure simulation in which active regions are injected based on random selection from parent distribution functions derived from solar data. The global properties inferred for the photospheric and heliospheric fields for these two models are in remarkable agreement, confirming earlier studies that no subtle flux-emergence patterns or field-dispersal properties are required of the solar dynamo beyond those that are included in the model in order to understand the large-scale solar and heliospheric fields.