The zeros and signs of δ¯ρFB actually tend to coincide with those of ρ¯0zz (not shown) because δρ≈-aδT+bδS, the vertical gradients of a and b are small, and hence δρδρ itself approximately obeys (7). Furthermore, the amplitudes in corresponding left and right panels are similar, although the actual anomaly tends to be somewhat weaker than that computed from (8), indicating that the growth of the anomalies has reduced below a linear trend within a year, owing to the onset of the adjustment processes discussed in Section 3.2.3. The exception is within 3° of the southern and northern boundaries where density anomalies
are small and the contours of the total density field are nearly horizontal (top panels), a consequence of find more the restoration of temperature and salinity toward prescribed values in those regions (Section 2.1). If density depends only on temperature, only on salinity, or if T and S have the same vertical structure (and the vertical gradients of ∂ρ/∂T∂ρ/∂T and ∂ρ/∂S∂ρ/∂S are negligible), a diffusion anomaly, δκbδκb, will only generate a dynamical anomaly, that is, it will shift isopycnals vertically with no change in spiciness. In the real ocean
AZD2281 nmr (and in our control run), however, T and S have different structures, and δκbδκb generates both spiciness and dynamical anomalies ( A). Fig. 4b illustrates the generation of both anomaly types in Solution FB, comparing δTFBδTFB (top-left) to its parts due GPX6 to dynamics δ′TFBδ′TFB (middle-left) and spiciness δ″TFBδ″TFB (bottom-left) during year 1. The pattern of the total temperature anomaly δTFBδTFB is very similar to that of δρFBδρFB in Fig. 4a (top-left panel) and it is largely explained by δ′TFBδ′TFB, especially in the tropics. In the subtropics, however, δ″TFBδ″TFB is significant because there are prominent salinity signals associated with subsurface waters, namely, high-salinity North Pacific Tropical Water ( Tsuchiya, 1968 and Suga et al., 2000, and references therein) and South Pacific Subtropical Lower Water ( Wyrtki, 1962) overlying
deeper, lower-salinity waters ( Wyrtki, 1962 and Talley, 1985). Fig. 4b also plots temperature anomalies where (7) is assumed to hold individually for both temperature and salinity (right panels). They show that the general patterns of all the anomaly fields in Solution FB are well predicted by the one-dimensional model. After the initial and local responses, solutions adjust toward a new equilibrium state through wave radiation, advection, and mixing. Locally-forced dynamical anomalies are associated with an unbalanced pressure field that excites baroclinic waves. In contrast, spiciness anomalies do not affect pressure (they are a passive tracer), and hence respond only to advection and mixing.