[1] The current configuration in the auroral region is known to consist typically of downward and upward field-aligned current (FAC) sheets, connected in the ionosphere by meridional Pedersen currents, while divergence free electrojets (EJs) flow azimuthally as Hall currents. This configuration of the auroral current circuit was introduced by BostrĂśm (1964) and labeled as "Type 2," while he suggested also an alternative "Type 1" configuration, with filamentary FACs connected in the ionosphere by azimuthal Pedersen currents. By using an updated version of the recently developed ALADYN technique, we investigated the divergence of the auroral electrojets for a few FAST crossings over the auroral oval in the 20-22 MLT sector, two of which are presented in detail. Although a precise estimate of the electrojet divergence is difficult, because of several error sources, the results suggest that this divergence can be significant over certain latitude ranges, comparable with the FAC density. Direct FAC-EJ coupling appears to contribute to the ionospheric current closure not only during active times, as already known, but also during rather quiet periods. The quiet time FAC-EJ coupling is likely to be achieved in a mixed "Type 1/Type 2" configuration, with the FAC sheet (Type 1) azimuthally connected to the Pedersen component of the EJ (Type 2). This configuration requires a non-zero tangential component of the electric field, and is therefore more likely realized inside or near the Harang region. At the same time, the divergence of the Hall current is presumably negligible, and likewise the ionospheric polarization, consistent with statistical results published recently. During more active intervals and possible reconfigurations of the auroral current circuit, our results suggest that the FAC-EJ coupling could be also achieved by Hall currents. We conclude by exploring a tentative scenario for the integrated evolution of the ionospheric current closure and Cowling mechanism during the substorm cycle. A systematic examination of more experimental evidence is needed to validate this scenario.