The mission of the Oceanography Section (OS) is to understand the large-scale ocean circulation and the dynamics of climate through studies of the important processes in the ocean and sea-ice, in air-sea-ice interactions, and in coupled systems.
Global Ocean and Sea-Ice Modeling
Every member of the OS has spent a good fraction of their time in the development of the third version of the Community Climate System Model (CCSM). A number of improvements and changes have been made to the ocean and sea-ice components for this version that will be used for the International Panel on Climate Change (IPCC) version of the model. This work is documented in the Ocean Model Working Group and Polar Climate Working Group reports in the CCSM portion of this report. In addition, Bill Large (OS) and Gokhan Danabasoglu (OS) led and performed the large majority of the sensitivity studies made with the CCSM2 to address the causes of the errors in the tropical Pacific Ocean and in the stratus regions off the western boundaries of continents. These sensitivity studies are well documented in the second paragraph of the Ocean Working Group report under the CCSM.
Danabasoglu and Large showed that the incorrect equatorial Pacific sea surface temperature (SST) seasonal cycle in the CCSM2 can be eliminated if both wind stress components are replaced with observed or uncoupled atmosphere component fields. The current hypothesis is that due to some positive feedbacks, the diabatic heating terms in the atmosphere component become substantially different in a coupled configuration compared to an atmosphere-only integration, resulting in incorrect winds.
Stephen Yeager (OS) and Large are leading a development effort to configure a new low resolution version of the CCSM3 ocean model. It was prompted by the revelation that the low resolution CCSM2 could not maintain an acceptable Meridional Overturning Circulation (MOC) in the North Atlantic. Since this model is to be used in circumstances where computing resources are limited, or very long integration times are required, the number of grid points could not be increased. Therefore, the same number of horizontal grid locations were concentrated more in the North Atlantic, which allowed the Northwest Passage through the Canadian Archipelago to be opened. Also, the distribution of vertical grid points placed more points in the upper 1000 meters than in CCSM2. Preliminary results show that the North Atlantic MOC now stabilizes at about 12 Sverdrup. A slightly larger value would give a better match to observational estimates, but further development is on hold pending the availability of a low resolution (T31) version of the CCSM3 atmospheric component.
A 1000-year control simulation of the present day climate has been completed without flux adjustments using the second version of the CCSM. Jeffrey Kiehl (CCR) and Peter Gent (OS) have written an overall assessment of this control run. In particular, several aspects of the simulation, especially in the arctic region, are much improved over those obtained in the first version of the Climate System Model (CSM-1). Other aspects, such as the tropical Pacific region, have not been improved compared to those in the CSM-1. Priorities for further model development are discussed in the conclusions section of a manuscript that has been submitted to the Journal of Climate.
Gent and Danabasoglu have begun a study of heat uptake in the ocean component of the CCSM-2. They are using results from the 1000-year control run, the 1% per year increasing CO2 run, and the doubled CO2 run. They are analyzing heat content changes in the upper 300m and the upper 3km to be consistent with an analysis of ocean observations published by Levitus in Science in 2000. They want to answer the question of how confidently can the increased ocean heat content be detected at a particular point in time in the 1% CO2 run compared to the natural variability of this quantity in the long control run.
Danabasoglu performed a 10,000-year synchronous control integration using the uncoupled CCSM ocean component subject to realistic, time-dependent forcing to quantitatively determine how well its solution is reproduced by two accelerated, but otherwise identical, equilibrium integrations. To our knowledge, this is the longest synchronous integration of such an Ocean General Circulation Model (OGCM). The two accelerated cases use tracer time steps increasing with depth and unequal momentum and tracer time steps, respectively. After accelerated equilibration, these cases are extended synchronously to achieve final equilibration (defined as an annual- and global-mean potential temperature trend below 10-5°C/year-1). The synchronous control integration achieves equilibrium in about 3500 years. The accelerated equilibration times are 685 and 3000 surface years for the two cases, respectively. Their synchronous extensions require about 100 and 700 surface years. The accelerated solutions do differ from each other and from those of the control experiment. However, for practical purposes, many aspects of the accelerated and control solutions are more similar than otherwise. Because of severe non-conservation issues associated with tracer time step variations with depth, this technique is not recommended. Instead, unequal momentum and tracer time steps should be used. Compared to the control case, this method of acceleration provides a factor of 2.5 reduction in the computational cost (accelerated + synchronous equilibration) with the possibility of further reductions. Any accelerated integration must be followed by a synchronous extension to recover the correct seasonal cycle and to eliminate any spurious oscillatory behavior present in the accelerated phase. If the incremental gains with further synchronous extensions are judged to be minimal or of minor significance, shorter integrations (<700 years) may suffice. Among other factors, the equilibration times are affected by the choice of surface forcing method.
Yeager and Large have completed an investigation of the origins of density-compensating anomalies of temperature and salinity (spice) using a model forced with the most realistic surface products available over the forty years 1958-1997. In this hindcast, the largest interannual spiciness anomalies appear in the Pacific near the isopycnal s0=25.5, where deviations as great as 1.2°C and 0.6 psu are generated equatorward of winter outcropping in the eastern subtropics in both hemispheres. These source regions are characterized by very unstable salinity gradients and low mean density stratification in winter. Two related signatures of this mixing are density that is well mixed deeper than either temperature or salinity, and subsurface density ratios that approach 1. All ocean basins in the model are shown to have regions with these characteristics and signatures, but the largest positive anomalies are found on different isopycnals ranging from s0=25.0 in the Northeast Pacific (NEP) to s0=26.5 in the South Indian Ocean. A detailed examination of the Southeast Pacific (SEP) finds that large positive anomalies are generated by diapycnal mixing across subducted isopycnals. A salinity budget on s0=25.5 reveals that negative anomalies form by a steady isopycnal advection, moderated by vertical advection and heave. There is considerable interannual variability in the strength of anomalies and in the density on which they occur. Historical observations are consistent with the model results, but insufficient to verify all aspects of the hindcast, including the processes of anomaly generation in the SEP. It was not possible to relate deep isopycnal variability signals to surface forcing of any kind at either the outcrop region, or the anomaly source region. A complex scenario involving basin wide circulation of both the ocean and atmosphere, especially of surface water through the subtropical evaporation zones, is put forward to explain this model result. As expected, Pacific anomalies generated on s0=25.5 can be traced along mean geostrophic streamlines to the western boundary, where they produce about twice as large decadal salinity variations (order ± 0.1 psu) at ˜ 7°S than are seen at 12°N, where there is more variance at shallower isopycnals. At least portions of the s0=25.5 signals appear to continue along the boundary to a convergence at the equator, suggesting that the western boundary exchange window includes the upper pycnocline of the SEP and NEP source regions.
Daisuke Tsumune (OS), a visitor from the Central Research Institute of Electric Power Industry (CRIEPI) in Japan, Keith Lindsay (OS), Scott Doney, Woods Hole Oceanographic Institution (WHOI), Frank Bryan (OS), and Matthew Hecht, Los Alamos National Laboratory (LANL) have continued to examine the variability in the North Pacific using chlorofluorocarbon (CFC) ages and ideal age in a CCSM Parallel Ocean Program (POP) experiment with realistic forcing data from 1958 to 2000. The simulated CFC concentrations are in good agreement with World Ocean Climate Experiment (WOCE) observations in shallow layers. The CFC tracer ages are computed using the partial pressure approach and the ratio approach. The relationships of pCFC-11, pCFC-12 age, ratio age and ideal age are as expected from previous studies by simple models. Temporal changes and the spatial distribution of the difference between these age measures have been described for the first time. The CFC ages have increased between 1958 and 2000 due to non-linear mixing effects, while ideal age does not. The CFC ages are often used as proxies to detect the change of oceanic conditions, such as subduction rates. In contrast to the mean age, interannual variabilities of CFC ages are in good agreement with those of ideal age. Interannual variabilities of the CFC ages and ideal age are larger than seasonal, monthly and decadal variabilities in this calculation.
Bryan continues the analysis of a series of North Atlantic basin integrations at
resolutions between 0.4 and 0.1 degrees that are being carried out in collaboration with
Richard Smith (LANL) and Hecht (LANL). A synthesis of many integrations in this
series reveals the tight coupling between the separation of the Gulf Stream from the
boundary at Cape Hatteras and the strength of the cyclonic Northern Recirculation Gyre
(NRG). Multiple mechanisms including downstream advection of cyclonic vorticity
generated at the boundary and baroclinic-topographic interactions appear to be involved in
maintaining the NRG, and adequate representation of each of them a prerequisite to proper
simulation of Gulf Stream separation.
A major new effort in high resolution (0.1 degree) global ocean modeling was begun in collaboration with oceanographers at CRIEPI, LANL and the Naval Postgraduate School (NPS), using computational resources at the Earth Simulator in Yokohama, Japan. A baseline integration for the period 1990-2000 was run using National Centers for Environmental Prediction (NCEP) based atmospheric forcing. To address some of the biases that were observed in this initial integration, a series of sensitivity experiments surveying a broad range of parameter space was carried out. While the North Atlantic Basin cases described above have provided some guidance in optimizing the model configuration, not all of the results translate directly to the global configuration. In particular, the Gulf Stream separation is not as well simulated in the global model for the same level of dissipation as used in more successful North Atlantic Basin experiments. Analysis and further experiments are ongoing in order to prepare an ocean component of a very-high resolution coupled climate system model.
Smith and Gent have developed an anisotropic generalization of the Gent-McWilliams (GM)
parameterization for eddy-induced tracer transport and diffusion in ocean models. They
focus on its application in high-resolution eddy-permitting and eddy-resolving models,
where it provides an adiabatic alternative to the more commonly used biharmonic horizontal
diffusion operator. A series of numerical simulations of the North Atlantic Ocean are
conducted at 0.2 deg resolution using anisotropic viscosity, anisotropic GM, and
biharmonic mixing operators to investigate the effects of the anisotropic forms, and to
isolate changes in the solutions specifically associated with anisotropic GM. They then
conduct a high resolution 0.1 deg simulation using both anisotropic forms and compare the
results with a similar run using biharmonic mixing. Modest improvements are seen in
the mean wind-driven circulation with the anisotropic forms, but the largest effects are
due to the anisotropic GM parameterization which eliminates spurious diapycnal diffusion
and leads to significant improvements in the model thermohaline circulation. These
improvements include the meridional heat transport, overturning circulation, and deep
water formation and convection in the Labrador Sea. A manuscript on this work has been
submitted to the Journal of Physical Oceanography.
Ocean Biogeochemical Modeling
Oceanic biogeochemical processes, as they relate to the global carbon cycle and climate system, are under study by Lindsay and Doney. Two paths currently being pursued are: (a) global ocean biogeochemical and tracer modeling and (b) the study of the relationship between the ocean carbon cycle and the global carbon cycle. Work the former has primarily consisted of: (1) a preliminary investigation into the impact of various tracer advection schemes on biogeochemical tracers and (2) the implementation of a new passive tracer infrastructure in the CCSM POP ocean model. The preliminary advection investigation has revealed previously unnoticed behavior in the advection scheme currently used in CCSM POP. This work will continue in the coming year and various advection schemes will be assessed for how they perform on a variety of tracers. The new infrastructure has been designed to make it easier to incorporate passive tracer modules into CCSM POP. Passive tracers modules that have currently been implemented are ideal age and CFC's
With respect to (b) above, the ocean carbon cycle has been coupled to a terrestrial land carbon cycle, atmospheric transport of CO2, and atmospheric radiation computations using the prognostic carbon-cycle model developed in previous years. This has been implemented in the CSM-1 framework. A sequential spin-up strategy has been used to prevent large drifts in the carbon cycle when the models are coupled. The resulting coupled carbon cycle climate model is now being used to study the impact on the climate of the inclusion the carbon cycle.
The Synthesis and Modeling Project (SMP), the last phase of the U.S. Joint Global Ocean Flux Survey (JGOFS), which is coordinated by Doney and Jorge Sarmiento (Princeton University), is winding down to its final year. The SMP effort strives to synthesize data collected during the U.S. JGOFS field studies into a set of models that reflect the current understanding of the oceanic carbon cycle, and includes 65 projects funded by NSF, NASA, NOAA, and DOE, and over 100 investigators. Joan Kleypas (ESIG) worked on this project as an Associate Scientist within OS until April 2003. Kleypas actively coordinated SMP activities through instigation and organization of meetings, collection and preparation of project data, editing of SMP special issues, and overall facilitation of investigator interactions, particularly between observationalists and modelers. This year, Doney and Kleypas organized the annual SMP meeting at WHOI, and edited the second SMP Special Issue in Deep Sea Research II, which is slated for publication in December 2003. Kleypas also managed the ongoing collection and dissemination of SMP data between NCAR and WHOI (David Glover). Additional information about the SMP is maintained at http://usjgofs.whoi.edu/mzweb/syn-mod.htm.
Air-sea-ice Interactions
The influence of the North Atlantic Oscillation-Arctic Oscillation (NAO-AO) on northern hemisphere sea-ice variability has been investigated in the 1000-year control integration of the CCSM2 by Marika Holland (OS). The model reasonably simulates the spatial structure and variance of the sea level pressure (SLP), surface air temperature, and precipitation associated with the NAO-AO. The sea-ice response to the NAO-AO compares well to observations. Variations in the spatial structure of the SLP anomalies associated with the NAO-AO over the course of the integration are present and part of the natural variability of the system. However, the magnitude of the observed trends over the last 40 years in the NAO-AO are never realized in the model simulations, suggesting that these trends may be associated with changes in anthropogenic forcing. A paper documenting this research was published in J. Climate (Holland, 2003).
The mechanisms forcing northern hemisphere polar amplification in coupled climate models have been investigated by Holland and Cecilia Bitz (University of Washington). The polar amplification varies considerably between the 15 different coupled climate models that were analyzed. It was found that conditions of the control climate sea-ice, including the extent and thickness, have a significant influence on the location and amplitude of polar warming at 2ŚCO2 conditions. Additionally, changes in cloud cover and ocean heat transport into the Arctic are significantly correlated to the polar warming. A manuscript discussing these results has been published in Climate Dynamics (Holland and Bitz, 2003).
Additional analysis of the CCSM control integration has been performed to examine variations in the southern hemisphere ice cover. This is collaborative work between Holland, Bitz and Elizabeth Hunke (LANL). The dominant mode of ice variability exhibits a dipole pattern with positive anomalies in the Pacific associated with negative anomalies in the Atlantic. This variability compares well to the observations. The mechanisms driving this variability and the atmosphere and ocean conditions associated with this variability have been investigated. The influence of ENSO variability and the southern annular mode on variations in southern ocean sea-ice extent and thickness have also been considered. A manuscript documenting this research is in preparation for submission to the Journal of Climate.
The fresh water budgets of the Arctic Ocean and the mechanisms that drive variability in the budget terms, including the transport of fresh water from the Arctic to the North Atlantic are being investigated within the CCSM2 control integration. This work is a collaborative effort between Holland, and investigators at the University of Colorado (Wanli Wu, Joel Finnis, Amanda Lynch, Mark Serreze). The primary goal of this project is to explore the progressive integration of the impacts and processes of arctic freshwater cycling on the local, pan-arctic, and global scales. This includes the study of processes that modify the arctic freshwater budgets in the terrestrial, atmospheric and oceanic environments; the integration of these processes to produce the exchange of freshwater between the Arctic and North Atlantic; and ultimately the influence of these processes on northern North Atlantic deep-water formation.
A major update to the ocean forcing datasets used in the OS has been completed by Large and Yeager. They now include consistent surface radiation, both long and short wave that were acquired from William B. Rossow of the NASA Goddard Institute for Space Studies (GISS). All the datasets now cover at least the period 1983 through 2000. Ancillary data sets of higher quality have been used to objectively correct the surface winds, air temperature and humidity to produce a globally balanced heat flux with only minimal adjustment to the radiation. The precipitation is a blend of two available data sets (Global Precipitation Climatology Project and the Climate Prediction Center Merged Analysis of Prediction - CMAP), where global ocean model solutions were used to determine which to use as a function of latitude. In response to an expressed community need to force ocean and sea-ice with a repeating annual cycle of forcing, these data sets have been processed into such a form. The criteria were to preserve the 43-year mean fluxes, to have realistic storm events, and to transition smoothly from December 31 to January 1, all without changing the standard forcing infrastructure.
Investigations are still continuing to study the transfer of wind energy to the ocean through resonant inertial forcing. One aspect of this work is in collaboration with Robert Stockwell, Colorado Research Associates (CoRA) and involves characterizing the inertial component of the wind. Moored data buoy vector winds have been examined for the occurrence of inertial oscillations. Local spectral analysis, using the S-Transform, finds considerable energy in the inertial frequency ranges, typically in the form of strong (10/s) yet short lived (order 1 day) events. The other research path is in collaboration with Patrice Klein, French Research Institute for Exploitation of the Sea (IFREMER) and Hecht and investigating the role of ocean meso-scale eddies in extracting wind energy by broadening the resonant frequency band. A pair of simulations of a 0.1 degree North Atlantic model have been performed, both forced by QuikScat (QSCAT) scatterometer winds at 0.5 degree resolution. In one case, as much as possible of the full inertial content of the winds is maintained in the 6-hourly wind fields. In the second case, much of the inertial component, depending on latitude, is filtered out by using daily averaged winds. The impact of this high frequency wind forcing on the energetics and hydrography of the North Atlantic ocean is being examined. The results of this work may enable parameterization of the eddy-dependent extraction of wind energy in non-eddy-resolving ocean climate models.
Ocean Observations
Large and Danabasoglu have been collaborating withscientists at the University of California at Los Angeles,Scripps Institute of Oceanography and WHOI on estimating the climatological three dimensional velocity field in the global ocean. The study makes use of velocity data from sub-surface floats, surface drifters and currentmeters combined with hydrographic profiles, sea surface heightmeasurements and geoid estimates based on satellite gravimetry from the Gravity Recovery and Climate Experiment (GRACE) mission. The estimates employ a generalized version of optimal interpolation that has been used extensively in meteorology and oceanography. This method requires a priori statistics, in the form of space-time covariances and/or empirical orthogonal functions. The primary NCAR contribution is associated with the low-frequency (non-mesoscale) variability and mean (ensemble) part of these a priori statistics and is based on analyses of CCSM POP global ocean circulation model simulations. The interannual and low frequency variability are estimated from a suite of integrations using atmospheric forcing from the 40 year NCEP reanalysis that has been adjusted by Large and Yeager for use in the CCSM simulations. This integration for 5 repeats of the 40 year atmospheric forcing for a total of 200 years is just underway. To estimate the ensemble statistics of the mean circulations, a set of 20 runs will be carried out with repeats of an annual cycle of atmospheric forcing. Each run will be carried out for 50 years. The 20 annual cycles are constructed from yearly values of atmospheric forcing based on modern observations (based on extensive satellite observations and atmospheric reanalyses starting in 1983). The last two weeks values for each year are tapered to produce smoothly varying annual signals. This 20 member ensemble of mean circulation and the 200 year run forced with interannual variability will then be analyzed to produce both local in space covariances and Empirical Orthogonal Functions (EOFs) for use in the global estimation project. We also expect that these simulations will be analyzed to characterize the variability of the model simulations on their own. The first of these ensemble integrations has been just completed.
Over the past year Aquarius, a new NASA satellite mission, designed to measure sea surface salinity (SSS) successfully passed the Final Risk Mitigation Phase (RMP) Review with the Earth System Science Program (ESSP) RMP Executive Board. The review board unanimously congratulated the Aquarius team for "answering the mail" and doing an excellent job preparing the task closure reports and addressing their questions. As a member of this team, Large, in collaboration with Yeager, addressed a question posed by earlier reviews: will Aquarius SSS accuracy of 2 psu be sufficient to infer anything about the hydrological cycle in midlatitudes where the SSS signal due to freshwater fluxes is diluted by deep mixed layers, and contaminated by other processes? From a series of ocean model integrations forced with different global precipitation products, they were able to show that monthly SSS estimates over the 3 year mission lifetime would be able to discriminate between the various products, because their differences are so great.