Today we looked at the observed relationship between the mid-latitude atmosphere and the SST of the mid-latitude ocean. We have discovered that the tropical atmosphere and ocean are closely coupled, and that the mid-latitude atmosphere is influenced by tropical dynamics (PDO). We find that the atmosphere and ocean aren2t as closely coupled in the mid-latitudes as they are in the tropics, and that it is much more difficult to change the SST in the mid-latitudes via atmospheric forcing (no Bjerkness feedback). From the thermal wind equation we can see that a local change in SST in the mid-latitudes will result in weaker circulation than in= the tropics (due to geostrophy).

What patterns do atmosphere-forced SST anomalies have in the mid-latitudes? To answer this question we looked at figures from a paper by Wallace et al. [1990] that compared patterns of atmospheric circulation anomalies in the mid-latitude North Pacific with SST anomalies. The patterns of variability are very PNA-like with the phase of the PNA forcing the SST =2E When the Aleutian low is anomalously strong the westerly winds at around 40N are stronger and result in anomalously low SST. Similarly in this phase of the PNA, the high pressure at about 30N is also stronger and results in warming of the SST under the high and anomalously low SST2s south of the high where the winds (and evaporation) are greater. The opposite pattern exists when the low and high pressure regions are weaker than usual. It is important to note that the atmosphere should be doing this despite ENSO forcing just from orography and the land-sea heat capacity contrast. A couple of questions arise from these patterns:

1. How much of the variability is due to the tropics?
2. How do the SST anomalies themselves effect these patterns?

The second question leads us into an investigation of the thermal coupling of the mid-latitude ocean and atmosphere. We saw that we didn2t need coupling for the PNA structure to exist. One characteristic of the ocean that may allude to the nature of the coupling is that it has a much redder spectrum than the atmosphere due to its high heat capacity (factor of 30 greater than the atmos). Thus, the ocean responds to low frequency thermal forcing and filters out the high frequencies. In a coupled system, we find that the atmosphere responds to the slow modulation of the ocean, and the modulation of the ocean is due to the low frequency forcing of the atmosphere. In this view of thermal coupling, we assume that the atmosphere is forced purely stochastically at both high and low frequencies and that the mid-latitude ocean has no intrinsic interannual variability. We can view the ocean2s effect on the atmospheric variability as a "subwoofer" effect. A complication to this simple linear feedback mechanism is ocean currents.

In general, we find that coupling of the mid-latitude ocean and atmosphere enhances the natural variability of the atmosphere at lower frequencies (year to year and decade to decade variability) by about a factor of 2.5 over the water (Manabe and Stouffer, 1996). The role of ocean circulation in SST anomalies in the mid-latitudes can be seen when a coupled ocean-atmos model using a mixed layer representation of the ocean is compared to a coupled model using a ocean slab model. The results of these coupled runs show that ocean circulation tends to damp the SST anomalies except in the eastern tropical Pacific and the far southern oceans. We also saw that coupling increased the PNA signal by 25% when an atmospheric GCM was forced by the warm phase of ENSO.

We continued our analysis of the mid-latitude coupling by investigating a stochastically forced coupled atmos-ocean model. The geometry of the model was a column of the ocean-atmosphere, with the atmosphere represented as a 2-layer gray body and the ocean as a slab mixed layer. The coupled heat equations included terms involving the net short-wave heat fluxes, long-wave heat fluxes, surface fluxes, and atmospheric dynamics captured in a stochastic forcing term (F). We linearize these equations about the climatology and ignore small anomalies in the short-wave forcing to arrive at a more simplified set of equations. To close the system, we assume that the surface air temperature is linearly related to the atmosphere temperature. After several more steps where we take a Fourier decomposition of the set of equations (assume wave-like solution) and decompose the stochastic forcing into a deterministic and a random part (F=DL+N), we arrive at our final set of equations:

is T_a=3D-aT_a+bT₀+N

ib s T₀=3DcT_a-dT₀

The terms are:

N =3Dnoise inherent to atmospheric dynamics

bT₀ =3Dforcing of the atmosphere by the ocean (feedback where not all heat comes back locally)7b represents both dynamic and therm= odynamic response.

-aT_a =3D damping term

cT_a =3D changes in ocean due to changes in air temperature

-dT₀=3D loss term due to turbulent exchange back to the atmosphere

The direct effects of these thermal anomalies are:

changes in storminess (u¢ v¢ ) Û changed barotropic flow