OVATION Prime-2013: Extension of auroral precipitation model to higher disturbance levels

OVATION Prime (OP) is an auroral precipitation model parameterized by solar wind driving. Distinguishing features of the model include an optimized solar wind-magnetosphere coupling function (d Phi(MP)/dt) which predicts auroral power significantly better than Kp or other traditional parameters, the separation of aurora into categories (diffuse aurora, monoenergetic, broadband, and ion), the inclusion of seasonal variations, and separate parameter fits for each magnetic latitude (MLAT) x magnetic local time (MLT) bin, thus permitting each type of aurora and each location to have differing responses to season and solar wind input-as indeed they do. We here introduce OVATION Prime-2013, an upgrade to the 2010 version currently widely available. The most notable advantage of OP-2013 is that it uses UV images from the GUVI instrument on the satellite TIMED for high disturbance levels (d Phi(MP)/dt > 1.2 MWb/s which roughly corresponds to Kp = 5+ or 6-). The range of validity is approximately 0 < d Phi(MP)/dt <= 3.0 MWb/s (Kp about 8+). Other upgrades include a reduced susceptibility to salt-and-pepper noise, and smoother interpolation across the postmidnight data gap. The model is tested against an independent data set of hemispheric auroral power from Polar UVI. Over the common range of validity of OP-2010 and OP-2013, the two models predict auroral power essentially identically, primarily because hemispheric power calculations were done in a way to minimize the impact of OP-2010s noise. To quantitatively demonstrate the improvement at high disturbance levels would require multiple very large substorms, which are rare, and insufficiently present in the limited data set of Polar UVI hemispheric power values. Nonetheless, although OP-2010 breaks down in a variety of ways above Kp = 5+ or 6-, OP-2013 continues to show the auroral oval advancing equatorward, at least to 55 degrees MLAT or a bit less, and OP-2013 does not develop spurious large noise patches. We will also discuss the advantages and disadvantages of other precipitation models more generally, as no one model fits best all possible uses.