This is a preliminary effort to document the QFlux feature in EdGCM. I'm hoping to use the feedback from the forum to create a good FAQ entry on QFlux. Please request clarification if you need it.
EdGCM includes three ocean modes, specified sea surface temperature (SST), prescribed sea surface temperature (QFlux), and prescribed sea surface temperature with deep ocean (QFlux + deep ocean). These ocean modes form a hierarchy from total fixed boundary conditions through a simple ocean model. EdGCM does not offer a full ocean model, but the QFlux scheme used can mimic many features of a "real" ocean.
18.374, 15.949, 15.459
25.575, 25.674, 25.145
28.058, 27.508, 27.106
28.354, 27.894, 27.274
27.853, 27.701, 27.809
Figure 1, Specified SST, part of a global dataset
The Specified SST is a set of monthly ocean temperatures for each grid cell arraigned much like an Excel spreadsheet (Figure 2). The specified ocean cannot change since it is a set of temperatures for the ocean surface. The problem with this ocean mode is that changes to the atmosphere are constrained by the ocean. Even if you double CO2 the specified ocean temperatures will prevent most of the warming from occurring. Clearly some kind of solution is needed since you can't really do global warming with Specified SST.
Figure 2, Specified SST sample temperatures
The Prescribed SST ocean mode or QFlux allows the ocean to adjust to the surface air temperature. Each grid cell in the ocean is given a temperature and a depth for the mixed layer (Figure 3). The mixed layer is the surface part of the ocean the is mixed by currents and winds so it interacts with the atmosphere on a daily basis. The QFlux ocean grid cell has temperature and thickness, therefore it has a heat capacity. There are no ocean dynamics, so the mixed layer cells cannot move mass, only energy.
Figure 3, QFlux ocean temperatures
In QFlux oceans, energy is moved from one grid cell to another to mimic ocean currents (Figure 4). In the real world the gulf stream and other currents move warm tropical water further north, that's why England is warm in the winter. Since QFlux cannot move mass it moves energy to accomplish the same thing; each grid cell has energy added or removed to simulate the effect of the ocean currents. The result of these manipulations of the ocean by QFlux are implied ocean transports.
Figure 4, QFlux ocean adjustments
The result of QFlux is a surface ocean that reacts to the atmosphere much like a real ocean would (Figure 5). If the ocean waterer is warmer than the air then energy is transfered from the ocean to the atmosphere, and the mixed layer cools. If the air is warmer than the ocean then the mixed layer warms. The energy added or subtracted from the mixed layer allows implied ocean transports of energy. The heat storage capacity of the ocean in QFlux is sufficient to slow warming for decades, but eventually the whole mixed layer warms and the simulation reaches equilibrium.
Figure 5, QFlux ocean adjustments
The deep ocean adds a heat sink below the mixed layer of the QFlux ocean. The heat sink allows the ocean to warm more slowly since energy is diffused downwards into the deep ocean. Since the eddy diffusion is parameterized from observed values the deep ocean only exists for the modern. It can take thousands of years for the deep ocean to reach equilibrium.
The mixed layer depth alters the heat storage capacity of QFlux. Each grid cell in the ocean has a monthly mixed layer depth and a percent ocean (Figure 6). Together the two numbers determine the heat storage capacity of the ocean grid cell in QFlux mode. The mixed layer depth changes on a monthly basis to reflect the observed mixing of the surface ocean based on currents and winds. An unfortunately side effect of the monthly mixed layer depth is that when the mixed layer thins the sea ice and freeze all the way to the bottom, which causes the model to crash.
Figure 6, Mixed layer depth
QFlux and deep ocean are what Jim Hansen calls "no surprise" oceans. Since the implied ocean transports cannot change the ocean can warm or cool, but the gulf stream cannot slow down or speed up. The use of prescribed conditions is a limitation, but most experiments can be performed without need of a dynamic ocean.
This ends the first part of the QFlux tutorial. Next week I'll look at what you need to do to collect QFlux.
Last edited by mshopsin; December 14th, 2005 at 10:51.
|Thread||Thread Starter||Forum||Replies||Last Post|
|Climate sensitivity||tomrees||Science and Modeling Questions||37||December 1st, 2005 03:06|
|Questions on paleoclimate setups||galen.mckinley||Science and Modeling Questions||4||November 4th, 2005 15:53|
|Bad Input File: "Deep Ocean" Qflux simulations||mshopsin||Announcements||0||October 11th, 2005 12:58|