Multi-Spacecraft Predictions of Space Weather Fronts at Earth:Predictions Using Four L1 Spacecraft and Dynamic Time Warping

Predictions of the solar wind at Earth primarily consist of forward extrapolations of measurements from two spacecraft, Wind and the Advanced Composition Explored (ACE). Using Wind or ACE, OMNI predicts solar wind conditions at Earth from L1 assuming that each measurement corresponds to a plane propagating in the ecliptic. Relying on one spacecraft at a time for solar wind predictions at Earth is not ideal, since the structure of the solar wind varies even on Earth-L1 scales. In addition to Wind and ACE, two other spacecraft, the Solar and Heliospheric Observatory (SOHO) and Deep Space Climate Observatory (DSCOVR), exist at L1. All four L1 spacecraft regularly observe the proton spectra from the solar wind at an approximately one minute cadence. DSCOVR, ACE, and Wind also possess magnetometers that have a time resolution less than 16s. Using the magnetic fields and the proton parameters from these four spacecraft, we create a mathematically unique prediction of the solar wind properties at Earth by assuming each propagating front is a plane with an arbitrary propagation vector. Instead of using χ² minimization to find one time offset for an entire time range, we use a Dynamic Time Warping (DTW) algorithm. DTW allows for each observation in the solar wind of a given spacecraft to map to a single spacecraft. As such, we obtain continuous series of time offsets, thus a series of propagating solar wind planes. Here we present an initial set of time periods around solar wind discontinuities and compare the four spacecraft DTW solution to OMNI predictions and near-Earth observations from THEMIS B and C. We find the DTW solution does a better job of predicting the observations in THEMIS B and C, especially when a front contains multiple discontinuities.