In solutions forced by climatological winds, the NECC develops a discontinuity (right panels) in the central Pacific that is not present in the observations. The character
of the error suggests that it arises from the near-equatorial (5S--5N) zonal wind stress, $\tau ^{x}$, being relatively too strong compared to the y-derivative of the wind stress curl term, (curl~${\bf \tau })_{y}$, associated with the Intertropical Convergence Zone. This is confirmed in solutions forced by interannual winds, which exhibit a wide range of responses from being very similar to
the observed NECC to being extremely poor, the latter occurring when
near-equatorial $\tau ^{x}$ is relatively too strong. Our results show
further that the model NECC {\it transport} is determined mainly by the
strength of (curl~${\bf \tau })_{y}$, but that its {\it structure} depends
on near-equatorial $\tau ^{x}$; thus, NECC physics involves equatorial as
well as Sverdrup dynamics. Only when the two forcing features are properly
prescribed do solutions develop a NECC with both realistic spatial
structure and transport.
