摘要：Our study statistically compares the total energy flux outputs of Newell et al.'s (2010a) oval variation, assessment, tracking, intensity, and online nowcasting (OVATION) Prime model, Hardy et al.'s (1991) Kp-based model, and a coupled Space Weather Modeling Framework ring current model to energy flux data obtained from 2198 Defense Meteorological Satellite Program (DMSP) satellite passes in the Northern Hemisphere. Our DMSP data set includes 28 days grouped into continuous 3 and 4 day periods between 2000 and 2008 and encompasses magnetic local times (MLTs) between 04:00 and 21:00. We obtain the most equatorward magnetic latitude coordinate, where a DMSP satellite energy flux measurement exceeds 0.4 erg/cm2/s, and use this point as a proxy for the equatorward boundary of the auroral oval in a particular MLT sector. We then calculate a prediction efficiency (PE) score by comparing the DMSP boundary coordinates to each model, using the same energy flux threshold to obtain a model's boundary location. We find that the PE for the OVATION Prime model is 0.55, and the PE for the Hardy Kp model is 0.51. When we accomplish the same analysis using a higher energy flux threshold equal to 0.6 erg/cm2/s, the OVATION Prime model's PE increases to 0.58, while the Hardy Kp model's score drops to 0.41. Our results indicate that more complex modeling techniques, like those used in OVATION Prime, can more accurately model the auroral oval's equatorward boundary. However, Hardy's discretized Kp model, despite its relative simplicity, is still a competitive and viable modeling option.
