Accelerated Climate Prediction Initiative (ACPI) Pilot Program
at the Scripps Institution of Oceaongraphy

Back to main page for pilot-ACPI  at Scripps

SIO Contribution to pilot-ACPI quarterly progress report, 06-27-2001

1. Objectives

The work done by SIO during this period represents new work incorporated into the pilot-ACPI project.  As such, it does not fall directly into any of the original "phases" of the project but rather represents our taking advantage of new opportunities for exploring issues related to pilot-ACPI's goals.

The objectives for this period were:

  1. Assemble the required data and interface routines needed for forcing LANL's biological ocean model with data obtained from the pilot-ACPI runs.  This will allow the biological model to forecast changes in the ocean's biology associated with anthropogenic forcing.
  2. Complete a surface heat budget analysis for the PCM historical runs.  The purpose of this analysis was to determine if the observed increase in the model's ocean temperatures was completely consistent with the idea that they were forced specifically by the increase in CO2.
Progress on these two objectives will now be described.

2. Assemble data and interface routines for LANL's biological model

The biological modeling will be handled by Mat Maltrud at LANL,  with the help of postdoc Joanne Lysne.  Mat determined that the following variables are needed, at daily intervals, to force the biological model:

Many of these quantities are directly saved in the daily pilot-ACPI runs.  However, wind stress was not retained, and the atmospheric quantities are on the model's lowest level, rather than at the 10m reference height level needed for calculating the surface fluxes from the bulk formulae.  After some discussion, we decided to adapt the PCM's code that calculates the ocean/atmosphere interface fluxes given the atmospheric model data to a stand-alone routine that will compute the surface fluxes we need. In practice, we used CCM3's flux calculations, which are essentially identical to PCM's.  An alterative method would have been to directly force the ocean model with surface heat flux components from the pilot-ACPI runs.  However, our previous experience with this technique during phase I of the project makes us think this would be ill-advised.  The advantages of calculating the fluxes using the PCM/CCM3 code are that they should be completely compatible with forcing the LANL biological model (which is based on POP, the same ocean model used in PCM), and calculating from the bulk formulae will automatically include the negative feedbacks needed to prevent the ocean model's state from wandering too far from reality.  This code also includes parameterizations to recenter the atmospheric data at the proper reference height, and to include the effects of stable or unstable near-surface temperature gradients in the flux calculations.

As of 06-27-2001, the PCM/CCM3 interface code required has been identified and modified as needed, but not yet tested to make sure it is working properly.  An additional issue that will need to be addressed is how to match the atmospheric model grid to the ocean model grid.  We will likely use SCRIP for that, which is a LANL product especially suited to remapping fields to POP's grid.  However, some details of how to handle land values, and atmosphere cells that overlap both ocean and land, remain to be worked out.

3. Surface heat budget calculations for anthropogenic warming

The PCM historical runs all show an increase in surface temperature and ocean heat content.  Since increases in anthropogenic gasses are the only difference in these runs from the control run (which does not show secular warming), it is fair to conclude that the gasses are forcing the warming.  However, the traditional paradigm of how CO2 will warm the climate makes specific predictions of how the increase in ocean heat content will be accomplished -- by an increase in longwave emission from the atmosphere that warms the ocean.  To our knowledge, this had not actually been tested in the PCM runs, just inferred from the results.  It seemed worthwhile to perform a surface ocean heat budget for the model to see if this did, in fact, account for the model's ocean warming.

The results for one ensemble member, B06.22, are shown below. (The budges for five experiments were computed, but all were similar).

The net surface heat flux, which is directly driving the increase in ocean heat content, is shown in gold.  The largest component driving the heat content increase is the increase in downward long wave radiation from the atmosphere (LW-DOWN), which is mostly but not completely compensated by increasing upward long wave loss from the ocean's surface (LW-UP) as the ocean SST increases.  There are also smaller components that contribute; both shortwave and latent heat fluxes drop, presumably due to increasing cloudiness and a drop in wind speeds, respectively.  On the other hand, the sensible heat flux increases, showing that the warmer atmosphere directly transfers some of its additional heat into the ocean.

The relative ranking of these sources is important, as it has bearing on the expected time relationship between the ocean's heat increases and atmospheric temperatures.  If the primary increase in ocean heat content were arising from an increase in sensible heat flux, this would impart a decadal lag between increases in atmospheric temperature and the subsequent forced increase in ocean heat content.  However, the largest component is actually the longwave flux, which drives a increase in ocean heat content that is simultaneous with the increase in atmospheric temperatures.

Finally, we note that the net increase in surface heat flux is a small residual between the increased longwave downwards from the atmosphere and the increased upwards from the ocean's surface.  To our knowledge this has not been specifically noted before, and might have impliciations for interpreting studies of changes in the atmosphere's energy balance.

Back to main page for pilot-ACPI at Scripps