Abstract/Summary

Accurate reconstruction of global solar-wind structure is essential for connecting remote and in situ observations of solar plasma, and hence understanding formation and release of solar wind. Information can routinely be obtained from photospheric magnetograms, via coronal and solar-wind modelling, and directly from in situ observations, typically at large heliocentric distances (most commonly near 1 AU). Magnetogram-constrained modelling has the benefit of reconstructing global solar-wind structure, but with relatively large spatial and/or temporal errors. In situ observations, on the other hand, make accurate temporal measurements of solar-wind structure, but are highly localised. We here use a data assimilative (DA) approach to combine these two sources of information as a first step towards producing a solar-wind “reanalysis” dataset that optimally combines model and observation. The physics of solar wind stream interaction is used to extrapolate in heliocentric distance, while the assumption of steady-state solar-wind structure enables extrapolation in longitude. The major challenge is extrapolating in latitude. Using solar-wind speed during the interval of the first perihelion pass of Parker Solar Probe (PSP) in November 2018 as a test bed, we investigate two approaches. The first is to assume the solar wind is two-dimensional and thus has no latitudinal structure within the ±7∘ bounded by the heliographic equatorial and ecliptic planes. The second assumes in situ solar-wind observations are representative of some (small) latitudinal range. We show how observations of the inner heliosphere, such as will be provided by PSP, can be exploited to constrain the latitudinal representivity of solar-wind observations to improve future solar-wind reconstruction and space-weather forecasting.