The paper of Pulkkinen et al. [1995] discusses the interesting question on how an MHD simulation matches the observed field, approximated by data-based models. However, due to an error in the numerical calculations, the currents implied by the empirical model were significantly underestimated. Adding to this confusion, the quiet-time empirical model field was compared with that measured in the tail lobe
during a disturbed period with unusually high solar wind pressure. Taken together, these two factors led to a gross exagerration of the inaccuracy of the T89 model.