YEAR-END REPORT: March 1999
Part A: Progress in each area of activity
1. Simulation of the NAO using a global atmosphere-only model
This project is based at McGill University and will be reported separately by Drs. Lin and Derome.
2. Simulation of the NAO using a global ocean-only model
Guoqing Li has now carried out this run using a coarse resolution global ocean model, including a parameterization of the mesoscale eddy field. What he does is to take the annual mean surface heat flux anomalies analyZed from COADS data by da Silva et al.(1994) for the years 1945 to 1994, and then use these anomalies to drive the ocean model. The results show a weakening of the North Atlantic thermohaline circulation (THC) after 1960, followed by an increase in strength beginning in the mid-1980's. It is not clear how significant this change is but it is in general accord with expectations based on what we think happened in the North Atlantic at that time. Guoqing has also carried out a run in which the da Silva et al. heat flux anomalies are randomly chosen in time, thus creating a random time series of heat fluxes with which to drive the model. In this run, the forcing has a white spectrum in time but has the spatial structure associated with observed heat flux anomalies. The North Atlantic THC in a 400 year run shows significant low frequency variability. The spectrum of the THC time series is red. Analysis of this, and the previous run, is continuing. In addition to the above work, Dr. Drew Peterson has been helping me to develop ideas on how to parameterize mesoscale eddies in ocean circulation models and we have a manuscript (Peterson and Greatbatch) we are about to submit on this work. I also have several papers with scientists at Los Alamos National Laboratory (Drs. Dukowicz and Nadiga) that have arisen from ongoing collaboration on the eddy parameterization problem.
3. Similation of the NAO using a coupled atmosphere-ocean model
Work on this project is being done partly at McGill and partly at Dalhousie University. The work at McGill is focussed on the Marshall/Molteni atmosphere model coupled to a coarse resolution global ocean model; the work at Dalhousie on the nature of the interdecadal variability in the GFDL coupled ocean/atmosphere model (the latter work is being done in collaboration with Dr. Tom Delworth).
In both models the atmospheric component appears to be the main source of the interdecadal variability, with the ocean responding in an essentially passive way. This, I believe, should be the ``null hypothesis" when studying interdecadal variability in the North Atlantic climate system, namely that the source of the low frequency variability is the atmosphere with the ocean responding passively. The challenge is then to unravel the complex nature of air/sea interaction over the North Atlantic to determine if and under what circumstances, the atmosphere is responsive to North Atlantic sea surface temperature (SST) anomalies. Understanding the nature of this response is necessary if we are to be successful in making long time scale (seasonal and beyond) predictions.
The work on the GFDL model is now submitted to Journal of Climate. We show that the interdecadal variability in the N. Atlantic THC of the GFDL coupled model is driven by the low frequency variation in the surface heat flux variability provided by the atmosphere model. We argue that the source of the low frequency variability is internal to the atmospheric model. A run in which
the annual mean surface fluxes from the coupled model are randomly rearranged in time gives similar variability to that found in the fully coupled model, strongly suggesting that the variability in the fully coupled model is not part of a dynamically coupled mode. Finally we show that the THC variability is driven by surface heat flux anomalies that bear a strong resemblance to the North Atlantic Oscillation (NAO).
4. Modelling studies of North Atlantic blocking patterns
In the proposal we asked for a student to work in this project. I have a new student arriving this summer who will likely work on this project. In addition, Dr. Drew Peterson is learning how to run the atmospheric model. The project has been hampered by the lack of adequate computer resources here at Dalhousie. On the positive side, using CICS funding for the North Atlantic Modelling project, a Silicon Graphics Origin 200 server with 4 processors and 1 Gbyte of memory is being installed at the time of writing. SGI equipment has the advantage of being compatible with MSC software and it should be possible to run the atmospheric model on this server. The project remains an important part to our plans because of the relationship between blocking patterns and the NAO (low/high NAO index being associated with more/less frequent blocking patterns).
5. Studies based on a primitive equations atmospheric model
Dr. Peterson has been working on the RPN atmospheric model code, a necessary preliminary to coupling. He is also in the process of finalising the global ocean model we shall use for the next round of coupled model studies. In May, Dr. Peterson and I shall be meeting with Drs. Derome, Lin (Charles), Hall and Lin(Hai) at McGill so that coupling with Nick Hall's atmospheric model can proceed. Plans are also proceeding to couple the global ocean model to the RPN atmospheric model.
Part B: Layman summary of scientific progress
The major concern of this research is to unravel the role, if any, played by the ocean in an atmospheric mode of variability known as the North Atlantic Oscillation (NAO). The NAO is the North Atlantic signature of a northern hemispheric mode of variability known as the Arctic Oscillation (AO) consisting of a systematic strengthening and weakening of the atmospheric westerly winds that circle the globe (in essence, the jet stream). When the NAO is in a high/low index phase, these westerly winds are stronger/weaker than normal and Europe experiences mild/cold winters while eastern Canadian winters are colder/warmer than usual. The key issue is whether the ocean plays a role in regulating the NAO. If the ocean does play a role, then it should be possible to make predictions several years into the future concerning the character of the winter weather in Europe and eastern Canada. So far our research is indicating that substantial variability can take place in the NAO that is unrelated to the ocean. On the other hand, recent model results from the UK appear to show the ocean dictating the changing phase of the NAO. This apparently
conflicting state of affairs is an indication that considerable more research is required before we fully understand the NAO. To address this issue we are developing coupled ocean-atmosphere models of varying complexity. We are also carrying out model experiments using the separate atmospheric and oceanic components of these models.
Part C: Report on progress being made by the collaborators
The project is particularly strong in its collaborative links with MSC in the Halifax/Dartmouth area (Dr. Hal Ritchie), McGill University (Drs. Derome and Lin) and with MSC scientists at both RPN and CCCMA in Victoria. Dr. Drew Peterson will be visiting Victoria during the summer to analyze the interdecadal variability in the CCCMA coupled ocean/atmosphere model, an activity not documented under the activities but one that has naturally developed through this project. Personally I find interacting with the meteorologists in this group very stimulating and highly beneficial to my research activity and my understanding of climate variability.
Part D: Relevant publications