Oceanography

Oceanography covers the physical, chemical, biological, and geological study of the ocean. My interest is almost entirely physical oceanography, and specifically the part that sets the lower boundary condition for the atmosphere.

For Galapagos climate the separation between ocean and atmosphere is not sustainable. The archipelago’s aridity, its cloud regime, and its rainfall are all consequences of ocean state, and an account of its climate that starts in the atmosphere has already assumed away the explanation.

Why the ocean dominates here

The ocean has roughly a thousand times the heat capacity of the atmosphere per unit volume and moves far more slowly. On climate timescales this makes it the memory of the system — the atmosphere responds within days, the ocean integrates over months to decades. Where atmosphere and ocean are strongly coupled, as in the tropics, the ocean sets the terms.

In the Galapagos the chain runs concretely. The Humboldt Current brings cool water from the south, and equatorial upwelling and the Pacific Equatorial Undercurrent shoaling against the archipelago bring cold water up from depth. The resulting cold pool holds SST well below the threshold for deep convection, which suppresses rainfall while supporting a stratocumulus deck beneath a sharp inversion. That deck is what produces the garúa and the fog interception that sustains the highlands.

Change the ocean and the whole structure changes. That is what ENSO does: warm the surface, deepen the thermocline, and an arid archipelago becomes a wet one within weeks.

What this means methodologically

The ocean’s depth and slowness are why it is so much harder to observe than the atmosphere. There is no oceanic equivalent of the radiosonde network, and the subsurface record before Argo floats is sparse to the point that many questions about earlier decades are simply unanswerable. Transports are reported in sverdrups and carry uncertainties that quoted precision tends to hide.

For regional atmospheric modelling the practical consequence is that SST is usually prescribed rather than simulated. That is defensible for short runs and progressively less so as the timescale lengthens — see AMIP for the same trade-off made explicit, and for why it is worst in the tropics.

See also: SST, the thermocline, upwelling, ENSO, and the ocean-atmosphere coupling map.