Effects of Uncertainties in Electric Field Boundary Conditions for Ring Current Simulations

• Published in: Journal of geophysical research. Space physics 2018-01, Vol.123 (1), p.638-652
• Main Authors: Chen, Margaret W., O'Brien, T. Paul, Lemon, Colby L., Guild, Timothy B.
• Format: Article
• Language: English
• Subjects: Boundary conditions; Computer simulation; Convection modes; Correlation analysis; cross‐polar cap potentials; Defense programs; DST Index; Electric fields; Empirical analysis; inner magnetospheric electric field; Meteorological satellite program; Meteorological satellites; Polar caps; Proton density (concentration); Recovery; ring current; Ring currents; Simulation; Statistical analysis; Statistical analysis of data; Storms; Time series; Uncertainty analysis
• ISSN: 2169-9380; 2169-9402
• DOI: 10.1002/2017JA024496

Physics‐based simulation results can vary widely depending on the applied boundary conditions. As a first step toward assessing the effect of boundary conditions on ring current simulations, we analyze the uncertainty of cross‐polar cap potentials (CPCP) on electric field boundary conditions applied to the Rice Convection Model‐Equilibrium (RCM‐E). The empirical Weimer model of CPCP is chosen as the reference model and Defense Meteorological Satellite Program CPCP measurements as the reference data. Using temporal correlations from a statistical analysis of the “errors” between the reference model and data, we construct a Monte Carlo CPCP discrete time series model that can be generalized to other model boundary conditions. RCM‐E simulations using electric field boundary conditions from the reference model and from 20 randomly generated Monte Carlo discrete time series of CPCP are performed for two large storms. During the 10 August 2000 storm main phase, the proton density at 10 RE at midnight was observed to be low (< 1.4 cm−3) and the observed disturbance Dst index is bounded by the simulated Dst values. In contrast, the simulated Dst values during the recovery phases of the 10 August 2000 and 31 August 2005 storms tend to underestimate systematically the observed late Dst recovery. This suggests a need to improve the accuracy of particle loss calculations in the RCM‐E model. Application of this technique can aid modelers to make efficient choices on either investing more effort on improving specification of boundary conditions or on improving descriptions of physical processes. Key Points We constructed Monte Carlo (MC) discrete time series model of errors in empirical model of cross‐polar cap potentials (CPCP) with measurements We performed 20 simulations with electric field boundary conditions specified from MC CPCP time series for two magnetic storms During storm main phase, disturbance Dst index is bounded by the MC simulated values