A note on nonlinear corrections to the Ekman layer pumping velocity