Long-term data: Why, where, how?

An understanding of physical, chemical and biological processes in the oceans requires regular and long-term observations. Only with such temporally extensive data sets can we separate real long-term trends in environmental drivers (e.g. a rise in temperature or decreasing pH) and their long-term biological consequences from the random (natural) variation that is a key feature of all complex systems.

Oceanic and coastal time series are therefore an invaluable tool for marine scientists and have been established in many locations around the world since the beginning of the 20th century, with the majority having been initiated from the 1950s onwards. These time series were initially concentrated in coastal regions as they were operated by marine stations (e.g. Plymouth, the Biological Station Helgoland and the Stazione Zoologica Anton Dohrn, to name but a few). These time series generate a range of data usually including temperature and salinity as a minimum, but in addition many also collect biological samples to study the abundance and diversity of marine organisms, such as phyto-and zooplankton, fish or benthic organisms. In combination, these data facilitate studies not only of how these assemblages are affected by their physical and chemical environment but also how this environment shapes biological interactions in the long-term.

With respect to station-based time series there are many good examples in the North Sea and North Atlantic, for instance the Helgoland Roads time series in the German Bight which was established in 1962, the Western Channel Observatory off Plymouth, established in 1903, or Hydrostation ‘S’ which started its measurements in the Sargasso Sea near Bermuda in 1957 (although operated by a marine station this is not really a coastal time series). All of these collect physical and chemical data as well as measuring different biological parameters (for more examples see list below).

An excellent example of a both spatially and geographically very extensive time series is the Continuous Plankton Recorder survey.  It operates CPR units on ships of opportunity and by now its time series, the oldest of which started in 1937, are covering a large part of the North Atlantic, and more recently the Pacific Ocean and other parts of the world ocean. There are also a number of cruise programmes that follow the same track on a regular basis over long periods of time, to collect time-series data on large spatial scales. The Atlantic Meridional Transect (AMT), which has been running since 1995, is a prime example of this type of spatially- and temporally-extensive sampling programme. 

Increasingly, as technology facilitates the installation of sensors that can operate largely unattended for extended periods of time, arrays of measuring devices are also being established in offshore locations and in the deep sea. The deep sea observatory at the Hausgarten for instance has now been operating for almost 14 years. Several projects are now devoted to the installation and management (including data management) of global networks of automated measuring devices including OceanSites for the deep sea and ARGO. Although these automated measuring devices relieve scientists of the need for labour and time intensive manual counts and measurements, their operation is nevertheless very costly (ship time and the equipment itself). Thus, such technologically advanced equipment is beyond the means of most developing countries. Capacity building, through investment in new technologies for developing countries, as well as training in the use of this equipment, aims to redress this imbalance between the developed and developing worlds.

The ongoing anthropogenic climate change has certainly increased the scientific interest in time series analyses and their importance is increasingly also recognized at a political level. Many data centres globally are now concerned with the long-term preservation of newly generated data and the retrieval of historic data sets that have not yet been properly curated. The challenge for scientists around the world is now the integration and analysis of these often complex data sets.

Below is a preliminary list of projects involving fixed-point time-series measurements with on-line data, and links to the relevant websites.