Fig. 2

Similar to Fig. 1, but for \(({\mathcal {O}}\text {-}{\mathcal {F}})_{\dot B}\) of the SV. In case II (dashed line), \(({\mathcal {O}}\text {-}{\mathcal {F}})_{\dot B} = {\mathcal {O}}(1)\) for much of the assimilation period before decays gradually in the last 20 years, implying that there is no similarity between the forecasted SV \({\dot b}_{l}^{m(f)}\) and the observed SV \({\dot b}_{l}^{m(o)}\). But its magnitude is much smaller in case III (solid line), and it decays monotonically in time, indicating that the SV assimilation accelerates the spin-up process