For example, the pathway from the tropical North Pacific in our Experiments NE and NW suggests that spiciness anomalies can enter the Indonesian Seas. Since there is strong selleck screening library vertical mixing in the Indonesian Seas, such subsurface spiciness signals may impact SST there (Ffield and Gordon, 1992, 1996).] Below the mixed layer, temperature anomalies along the equator are a superposition of dynamical and spiciness components. Their structure generally depends on the strength and spatial patterns of the signals in the regions where they are locally generated and on the processes by which they spread to the equator. Forcing near and at the equator (Regions ESE, ESW, ENE, ENW, EQE, and EQW), however,

has common influences on the equatorial temperature structure. It generates positive dynamical anomalies (deepening of isopycnals) in the lower pycnocline and weaker negative dynamical anomalies in the upper pycnocline;

it also generates negative spiciness anomalies in the pycnocline, which partially cancel Pirfenidone purchase the positive anomalies due to dynamical signals in the lower pycnocline (Fig. 8b, Fig. B.3b and Fig. B.4b). An assumption underlying our split of the domain into subregions is that the ocean’s response to δκbδκb is (approximately) linear, that is, the total response is close to the sum of the individual responses. Linearity should hold in the limit of small δκbδκb, since δTδT will then be well approximated by the first-order term in the Taylor expansion of T   with respect to δκbδκb. To test this property, we compared the sum of the temperature anomalies (∑eδTe)∑eδTe to δTFBδTFB (see Sections 2.2 and 2.3) along a few representative meridional sections (not shown). The two solutions are very similar at year 1, consistent with the fact that not much signal has yet propagated from each forcing region to other regions. At year 20, the large-scale patterns of ∑eδTe∑eδTe and δTFBδTFB are still similar (by the eye). On the ZD1839 chemical structure other hand, ∑eδTe∑eδTe is much noisier with strong mesoscale features superimposed on the large-scale signals. This difference suggests that mesoscale disturbances caused by δκbδκb

in one region are not much attenuated in other regions in regional experiments because κbκb is small outside their respective forcing regions, whereas they are attenuated by δκbδκb everywhere in Experiment FB. This difference can be interpreted as a nonlinear effect due to terms like δκbδTezzδκbδTezz in Experiment FB. Although we have restricted forcing by δκbδκb to be depth-independent, a number of studies point toward the importance of its vertical structure. For example, Sasaki et al. (2012) increased the background vertical diffusivity, κbκb, only above the center of the equatorial pycnocline in the equatorial Pacific ( analogous to our Regions EQE and EQW), in order to simulate the enhanced mixing recently found there (Richards et al.