5.1 Sources of energy

Different physical mechanisms for heating the corona probably govern active regions, closed loops in the quiet corona, and the open field lines that give rise to the solar wind (see other reviews by Marsch, 1999Hollweg and Isenberg, 2002Jump To The Next Citation PointLongcope, 2004Gudiksen, 2005Aschwanden, 2006Klimchuk, 2006). The ultimate source of the energy is the solar convection zone (e.g., Abramenko et al., 2006bMcIntosh et al., 2007Jump To The Next Citation Point). A key aspect of solving the “coronal heating problem” is thus to determine how a small fraction of that mechanical energy is transformed into magnetic free energy and thermal energy above the photosphere. It seems increasingly clear that loops in the low corona are heated by small-scale, intermittent magnetic reconnection that is driven by the continual stressing of their magnetic footpoints. However, the extent to which this kind of impulsive energy addition influences the acceleration of the solar wind is not yet known.

Intertwined with the coronal heating problem is the heliophysical goal of being able to make accurate predictions of how both fast and slow solar wind streams are accelerated. Empirical correlation techniques have become more sophisticated and predictively powerful (e.g., Wang and Sheeley Jr, 19902006Arge and Pizzo, 2000Leamon and McIntosh, 2007Cohen et al., 2007Vršnak et al., 2007) but they are limited because they do not identify or utilize the physical processes actually responsible for solar wind acceleration. There seem to be two broad classes of physics-based models that attempt to self-consistently answer the question: “How are fast and slow wind streams heated and accelerated?”

  1. In wave/turbulence-driven (WTD) models, it is generally assumed that the convection-driven jostling of magnetic flux tubes in the photosphere drives wave-like fluctuations that propagate up into the extended corona. These waves (usually Alfvén waves) are often proposed to partially reflect back down toward the Sun, develop into strong MHD turbulence, and dissipate over a range of heights. These models also tend to explain the differences between fast and slow solar wind not by any major differences in the lower boundary conditions, but instead as an outcome of different rates of lateral flux-tube expansion over several solar radii as the wind accelerates (see, e.g., Hollweg, 1986Jump To The Next Citation PointWang and Sheeley Jr, 1991Matthaeus et al., 1999Jump To The Next Citation PointCranmer, 2005Suzuki, 2006Jump To The Next Citation PointSuzuki and Inutsuka, 2006Cranmer et al., 2007Jump To The Next Citation PointVerdini and Velli, 2007Jump To The Next Citation PointVerdini et al., 2009).
  2. In reconnection/loop-opening (RLO) models, the flux tubes feeding the solar wind are assumed to be influenced by impulsive bursts of mass, momentum, and energy addition in the lower atmosphere. This energy is usually assumed to come from magnetic reconnection between closed, loop-like magnetic flux systems (that are in the process of emerging, fragmenting, and being otherwise jostled by convection) and the open flux tubes that connect to the solar wind. These models tend to explain the differences between fast and slow solar wind as a result of qualitatively different rates of flux emergence, reconnection, and coronal heating at the basal footpoints of different regions on the Sun (see, e.g., Axford and McKenzie, 1992Jump To The Next Citation Point1997Fisk et al., 1999Jump To The Next Citation PointRyutova et al., 2001Jump To The Next Citation PointMarkovskii and Hollweg, 2002Jump To The Next Citation Point2004Jump To The Next Citation PointFisk, 2003Jump To The Next Citation PointSchwadron and McComas, 2003Jump To The Next Citation PointWoo et al., 2004Fisk and Zurbuchen, 2006).

It is notable that both the WTD and RLO models have recently passed some basic “tests” of comparison with observations. Both kinds of model have been shown to be able to produce fast (v > 600 km ∕s), low-density wind from coronal holes and slow (v < 400 km ∕s), high-density wind from streamers rooted in quiet regions. Both kinds of model also seem able to reproduce the observed in situ trends of how frozen-in charge states and the FIP effect vary between fast and slow wind streams.

The fact that both sets of ideas described above seem to mutually succeed at explaining the fast/slow solar wind could imply that a combination of both ideas would work best. However, it may also imply that the existing models do not yet contain the full range of physical processes – and that once these are included, one or the other may perform noticeably better than the other. It also may imply that the comparisons with observations have not yet been comprehensive enough to allow the true differences between the WTD and RLO ideas to be revealed.

Several recent observations have pointed to the importance of understanding the relationships and distinctions between the WTD and RLO models. The impulsive polar jets discussed in Section 4 may be evidence that that magnetic reconnection drives some fraction of the fast solar wind (see also Fisk, 2005Jump To The Next Citation PointMoreno-Insertis et al., 2008Pariat et al., 2009). Also, direct observations of Alfvén waves above the solar limb indicate the highly intermittent nature of how kinetic energy is distributed in spicules, loops, and the open-field corona (De Pontieu et al., 2007Tomczyk et al., 2007Tomczyk and McIntosh, 2009). Spectroscopic observations of blueshifts in the chromospheric network have long been interpreted as the launching points of solar wind streams, but it remains unclear how nanoflare-like events or loop-openings contribute to the interpretation of these diagnostics (He et al., 2007Aschwanden et al., 2007McIntosh et al., 2007). Even out in the in situ solar wind – far above the roiling “furnace” of flux emergence at the Sun – there remains evidence for ongoing reconnection (Gosling et al., 2005Gosling and Szabo, 2008). There is also evidence that the dominant range of turbulence timescales measured in interplanetary space (i.e., tens of minutes to hours) is related to the timescale of flux cancellation in the low corona (Hollweg, 19902006).

Determining whether the WTD or RLO paradigm – or some combination of the two – is the dominant cause of global solar wind variability is a key prerequisite to building physically realistic predictive models of the heliosphere. Many of the widely-applied global modeling codes (e.g., Riley et al., 2001Roussev et al., 2003Tóth et al., 2005Usmanov and Goldstein, 2006Feng et al., 2007) continue to utilize relatively simple empirical prescriptions for coronal heating in the energy conservation equation. Improving the identification and characterization of the key physical processes will provide a clear pathway for inserting more physically realistic coronal heating “modules” into three-dimensional MHD codes.

  Go to previous page Go up Go to next page