Variability in our oceans influences the variability in our atmosphere through complex air-sea interactions. We need to fully understand variability in both the ocean and the atmosphere to fully understand our climate. This sounds obvious, but has previously been fraught with difficulty. If we want to say something robust about variability then we need observations, preferably at a high spatial resolution and over a long period of time.
Millions of oceanographic observations have previously been taken all over the globe. Many different research institutes in many different countries store these observations so accessibility becomes problematic. These institutes may record simple information like dates, times, latitudes and longitudes differently so working with several databases becomes challenging. Above all, few databases cover a long enough period to truly study climate variability.
The research community lacks a data source to analyze both temporal and regional variability. I went to speak to Alexander Korablev to find out more about what is being done to remedy this. Alexander explained to me that ‘previous large-scale attempts at gathering observations have led to databases containing in-situ observations or climatological fields available through centers such as NODC or ICES. However, there are still strong demands for regional products which allow for ocean climate studies over long periods for time and at high temporal-spatial resolution’.
Alexander and his colleague Alexander Smirnov from the Arctic and Antarctic Research Institute (AARI, St. Petersburg, Russia) have spent many years collecting and compiling oceanographic data into the new Climatological Atlas Of the Nordic Seas and Northern North Atlantic, which extends all the way back to 1900.
This new Atlas will contain gridded fields of water temperature, salinity and density at different depths. Temporal and spatial resolution will be down to one month and 0.25 degrees respectively for some parameters. We will also be able to access time-depth diagrams of thermohaline parameters for selected areas with repeated observational programs. Now we can really analyze variability.
Alexander and his colleagues have constructed the Atlas with observations from over 500 000 ‘stations’ (see all the little black dots in the figure above). Alexander explains that a ‘station’ can consist of a single oceanographic profile on a single day, or data from a single point can be repeated many times spanning a longer time period. As a rule stations contain only temperature and salinity profiles but sometimes additional parameters including chemistry characteristics. The uneven distribution of the observations means that Alexander must apply a specially designed, state-of-the-art method (Data Interpolating Variational Analysis) to produce the gridded fields. All the challenges for compiling such an atlas must be met with modern and advanced solutions.
Alexander tells me that the new data can initiate and verify numerical models, and alleviate the study of previous variability. Alexander explains that ‘the Atlas contains two variants of mean climatological fields. The first variant is based on analysis of all the data regardless of observation date. This method is popular but regards interannual and decadal variability as noise. The second variant is based on analysis of monthly means. This method treats climatic signals more carefully although it reduces the area of the resultant fields.’
Alexander continues to explain that these options give more flexibility since the user can chose the most appropriate variant for his/her task. For example, we can now study events like the Great Salinity Anomaly (Dickson, 1988) in more detail.
The Great Salinity Anomaly
During the 1960’s, atmospheric circulation patterns over the Arctic favored southward ice transport out of the Arctic, along the East of Greenland towards Iceland. During these ‘Ice-Years’ great amounts of sea ice accumulated along the north coast of Iceland. The ice continued its journey south and melted in the process. The melt water caused a huge input of fresh water to the upper layers of the ocean. Ocean currents transported this fresh water – the Great Salinity Anomaly – into the North Atlantic. Because fresh water increases stability, this layer did not mix into the upper layers. The anomaly circulated in the Atlantic and re-entered the Nordic Seas.
Alexander eagerly displays the map showing the temporal development of upper level fresh water content as the Great Salinity Anomaly returned to the Nordic Seas. The figure clearly shows higher fresh water content in 1976 relative to the long-term mean. The data in the new Atlas lets us watch the Great Salinity Anomaly propagate through time. The Atlas also lets us extract time series for in-situ locations. With this data, Alexander and his colleagues have corroborated a theory that Blindheim and others proposed in 2000. The theory states that a new fresh water anomaly from the north magnified the GSA as it propagated back into the Nordic Seas. You can see in the figure that a fresh water peak in the East Iceland Current (EIC) (black circle) occurs at roughly the same time as the GSA signal enters the Nordic Seas again at FSS (blue circle and arrow). The data in the new Atlas confirms Blindheim’s theory.
Alexander tells me that this analysis was previously impossible. It’s a perfect example of how the improved in-situ dataset and derived climatological fields can enhance temporal and regional analysis of variability in the Nordic Seas. Better understanding of variability will give more clues about the driving processes and forces.
Alexander is clear that many challenges remain. He tells me that observational systems in the Nordic Seas fail to capture key variations such as fresh water export from the Arctic. Shelf areas along the Greenland coast are often inaccessible due to sea-ice, so reliable climatologies are impossible to compile. Alexander is excited about the new era, where underwater gliders will start routine observations under the ice. He mentions other technological advances such as 3D geo-statistical modeling and the inclusion of advection. These developments are already becoming a reality, and the new data will undoubtedly be included in future updates of the Atlas.
The Atlas will be published at the National Oceanographic Data Center (NODC, Washington, USA) and will be publicly available later this year. The in-situ dataset was extended and considerably updated as part of a new release of the World Ocean Database under Global Ocean Data Archeology and Rescue Project (GODAR).
The total size of the Atlas is roughly 200 GB, but separate fields and data will be available for users in convenient NetCDF format.
To contact Alexander, write to firstname.lastname@example.org