OBSERVATIONS – E-SURFMAR
Coordinate, optimise, and progressively integrate European activities for surface marine observations in support of Numerical Weather Prediction and climate monitoring.
"Surface marine observations globally account for 3.5% of the total error reduction on Day 1 forecasts achieved by all types of observation ingested in real-time by NWP models*."
– Procure drifting buoys, arrange deliveries in a few suitable locations, coordinate deployments in the EUMETNET area thanks to voluntary participating ships, and arrange for data processing and delivery.
– Compensate participating members according to their activities in surface marine observations.
– Realize savings in communication costs for conventional VOS ships not equipped with Shipborne-Automatic Weather Station (S-AWS).
– Operate a small fleet of S-AWS to patch coverage gaps in the EUMETNET area.
– Facilitate the exchanges of best practices and technical information between participants.
– Manage the development and implementation of a EUropean common Automatic Weather Station (EUCAWS), procured under common specifications.
– Monitor the availability, timeliness, and quality of the data produced by each component of the service and take appropriate measures to correct the problems.
– Work closely with several task teams from the two international organisations involved in surface marine observations: JCOMM Data Buoy Cooperation Panel (DBCP) and Ship Observations Team (SOT).
– Maintain an operational network of about 100 drifting buoys measuring air pressure and sea surface temperature in the North Atlantic, with data made available in real time onto the Global Telecommunication System of WMO.
– Support financially and technically the operation of 4 moored buoys as a baseline for the calibration and the validation of satellite-derived wind and wave observations.
– Support the activities of Voluntary Observing Ships (VOS) through better coordination and harmonisation of practices, and compensation of participating members for the observations and the communications related to this component.
– Operate a fleet of Shipborne Automated Weather Stations (S-AWS) to patch coverage gaps in the EUMETNET area and facilitate the automation of the observation aboard ships through the development of such stations.
The Coordinating Member for this Operational Service is Météo-France and its Manager is Paul Poli. The Data Buoy Manager is Gilbert Emzivat and the VOS Coordinator is Jean-Baptiste Cohuet.
*Estimates are based on ECMWF Operations observation feedback for all observations assimilated between May 2015 and June 2016, considering the first day of each month.
The Operational Service was established during 2003 as an optional element of the EUMETNET Observation Programme (previously called EUCOS), on the basis of a detailed Operational Service proposal.
Supported by 15 countries, it had been initially defined with a period of four years divided into two stages of two years each under the responsibility of Meteo France-France. During stage 1 (2003-2004), a comprehensive design study was carried out and given to the EUMETNET Council. This study was accepted by the EUMETNET Council in September 2004. Stage 2 (2005-2006) mainly consisted in the implementation of the designed network.
In 2006, the Council renewed its confidence in Meteo France to manage the Operational Service during its next phase (2007-2011). Like other Operational Services, this phase was extended to one year (2012).
In November 2012, the EUMETNET EIG Assembly decided to continue the E-SURFMAR Operational Service for 5 years from the 1st January 2013 to the 31st December 2017, under the responsibility of Meteo-France.
Since its creation, the E-SURFMAR Operational Service has seen significant reorganisation in the management of marine observing programmes within Europe:
– the COSNA group (Composite Observing System for the North Atlantic) was disbanded in August 2003;
– a VOS Technical Advisory Group (VOS-TAG) was established in September 2003. Before its forming, no forum or organisation actually existed in Europe to co-ordinate technical and operational strategies for European voluntary observing ships;
– the European Group on Ocean Stations (EGOS) reformed as the Data Buoy Technical Advisory Group (DB-TAG) of the E-SURFMAR Operational Service in January 2005. A Data Buoy manager was appointed by the EUMETNET Council (Meteo France responsibility).
The E-SURFMAR design study was driven by the main EUCOS aim: to improve the quality of numerical and general forecasts over Europe. It showed that the most suitable parameter required by regional Numerical Weather Prediction (NWP) which cannot be provided by the space segment, is air pressure.
The recommended was to increase the density of observations for this parameter as a matter of priority. It also showed that this could be achieved through the use of more drifting buoys and Voluntary Observing Ships (VOS) reporting hourly data from sensitive areas: in the North Atlantic (north of 35°N) and in the Mediterranean Sea.
Some parameters measured by satellites, such as Sea Surface Temperature (SST), wind and waves, require calibration and validation. However, reliable in situ SST data are already provided by drifting buoys. So, an increase of these platforms would naturally contribute to improve the quality of satellite data calibration for SST.
For wind and waves, the E-SURFMAR design study recommended the use of four existing moored buoys which would be upgraded to provide directional wave spectra and 10-minutes wind data, and possibly being re-located in more suitable positions.
A considerable amount of work was carried out during the first ten years of the Operational Service. Results include:
– the direct management of a network of about 100 drifting buoys and 26 Shipborne Automated Weather Stations;
– full integration of the former EGOS group activities;
– a significant decrease in costs of observations carried out by automated systems operated by the Operational Service management but also by individual NMSes;
– a significant increase in the volume of observations reported by these systems;
considerable savings on data transmission costs thanks to the use of Iridium and data compression techniques as and when possible;
– improvements in the timeliness of drifting buoy data thanks to the use of Iridium instead of Argos for these platforms;
– the development and the use of a global metadata database for ships;
– the drafting of common technical specifications for a European Shipborne AWS to be used in a EUMETNET call for tender (EUCAWS project);
– the establishment of financial arrangements to compensate national activities that are suitable for the Operational Service;
– co-operations, most of them informal, with third parties: NOAA (US), MSC (Canada), Puertos des Estado (Spain), MOON Community, GHRSST group…;
– the set up of tools available on the Web to monitor all operational surface marine observation networks in the world.
The Operational Service also saw a significant increase in the volume of observations reported by national shipborne S-AWS.
In 2010, ECMWF conducted an OSE study which showed a clear positive impact of buoy data on numerical weather predictions. In the North Atlantic, the present density of additional buoys – compared to the situation before E-SURFMAR started –, has a visible but moderate impact in terms of forecasts scores. The impact of E-SURFMAR contribution was slightly positive on storms Klaus and Xynthia for the 24h forecasts but remarkable for 48h and 96h. ECMWF recommended not to decrease the present density of sea level pressure observations in the North Atlantic.
Since mid-2007, between 90 and 105 drifting buoys fully or partially funded by E-SURFMAR were permanently operating. This number significantly decreased in 2011 due to wrong batches of buoys delivered by the two usual manufacturers. The average number of operational buoys was below 75 buoys till mid-2012.
Since, more than 100 E-SURFMAR drifting buoys are permanently operating in the EUMETNET area. In parallel, about 45 non-project buoys in average – from NOAA, Meteo-France, LOCEAN, Marlin -, are operating in the area. E-SURFMAR monitor their data as it does for those funded by the Operational Service.
Drifting buoys are deployed from different harbours thanks to local partners (e.g. Port Meteorological Officers): Charleston, Fos-sur-Mer, Halifax, Helsingborg, Kirkenes, Le Havre, Liverpool, Norfolk, Rotterdam, Reykjavik, Southampton…
European Union Horizon 2020 Research and Innovation
E-SURFMAR takes part in the European Union (EU) Horizon 2020 (H2020) research and innovation programme under grant agreement no. 633211 within the project AtlantOS.
AtlantOS aims to improve and innovate the Atlantic observing by using the Framework of Ocean Observing to obtain an international, more sustainable, more efficient, more integrated, and fit-for-purpose system. AltantOS is a large-scale EU Horizon 2020 research and innovation project that contributes to the Trans-Atlantic Research Alliance and GEO.
The participation of E-SURFMAR in this project is to lead the surface drifter network component. Three axes of enhancement are being explored: enhance the geographical coverage, enrich the variety of variables measured at low cost by drifters, and study whether vertical sampling of the upper ocean layer by drifters would assist in understanding and predictability. In addition, E-SURFMAR is working within this project to improve the general data access and data integration, for all categories of users of surface drifter data.
Started on 1 April 2015, the project is due to last for 4 years. The EU funding in AtlantOS enables annual deployment of 13 drifters measuring currents, sea-surface temperature, and atmospheric surface pressure in the Tropical Atlantic. In addition, a prototype drifter to measure sea-surface salinity at an affordable cost is also being developed by partners.
The operations of four moored buoys are presently compensated by the Operational Service: K-pattern buoys K5, M6 and Lion and SeaWatch buoy Cabo Sillero. The first three are operated by Met Office, Met Eireann – in association with MRI – and Meteo France, respectively. The last one is operated by Puertos del Estado. These buoys were chosen for their theoretical ability to perform accurate wind and wave measurements (spectra).
During a long time, Cabo Silleiro was the only buoy to provide directional wave spectra and Lion buoy was only providing omnidirectional spectra. Since mid-2008 and October 2013 respectively, K5 and Lion moored buoys have also been reporting directional wave spectra. M6 does not measure this parameter yet.
In general, the current quality of moored buoy data is among the best we can get from surface marine stations. For instance, the RMS of air pressure differences with model outputs was 0.7 hPa in 2013 for the 4 moored buoys (against 0.6 hPa for drifting buoys, 0.7 hPa for S-AWS and 1.3 hPa for conventional VOS). For wind data, the RMS of differences with model outputs was 2.6 m/s in 2013 (against 3.9 m/s for S-AWS and 4.5 m/s for conventional VOS).
These four buoys are part of national networks which are also monitored by the Operational Service. The aim is here to get, in real time, as many quality observations as possible.
Besides moored buoy networks operated by E-SURFMAR participating members or historical partners (Puertos del Estado and MRI), other networks exists. The Operational Service seeks for cooperation with the institutes which operate these networks and their closest NMS in order to have their data fed into the GTS in real time. Thus, since mid-2010, the Greek Poseidon network, the Italian Ron network as well as, more recently, moored buoys operated by BSH (Germany), have been fed into the GTS and carefully monitored. The close cooperation between E-SURFMAR and these institutes also allowed to improve the measurements carried out by these buoys which now are close to reach WMO standards.
In 2013, each month about 370 European conventional VOS reported 270 observations per day in average from the EUMETNET area of interest. This is 20% less observations than in 2002 (for 30% less ships), before E-SURFMAR started, but still slightly higher than the target of 250 observations which should not be overpassed. During the same period, each month about 125 European S-AWS reported 1,700 observations per day in average from the EUMETNET area. This is 5 times more than in 2002.
For different reasons, the timeliness of conventional VOS data – 86% of the messages were received within 50 minutes in 2013 -, remained below the target of 90% for the eleventh consecutive year of the Operational Service. At the opposite, the timeliness of S-AWS data climbed well above the target of 90%: 94% in 2010.
Early in the Operational Service, it was noted that the quality of pressure values reported by conventional VOS was below that of automated systems. The most common errors are due to a wrong correction of the height of the barometer above the waterline. Despite the set up of daily updated blacklist, the situation did not change much during the past years. It however must be noticed that a slight improvement appeared in 2013 : the RMS of differences between observations and model outputs was 1.3 hPa (against 0.7 hPa for S-AWS), for a target of 1.0 hPa.
By the end of 2013, 10 BaTos and 18 BaRos AWS funded by the Operational Service were in operation (E-SURFMAR S-AWS fleet). The BaRos is normally a simple (autonomous) system reporting air pressure only (see picture here below). Moreover, three systems out of the 18 have been upgraded in BaRos+ AWS by Meteo France (air temperature, air humidity and wind measurements added). In 2013, several installations were done thanks to MOON members on ships plying in the Mediterranean Sea. The maintenance of this network of AWS is not an easy job. Ships are often changing their programmes and may be sold or de-constructed without any lead time.
A call for tender for the procurement of common S-AWS was issued in 2013. Three prototypes funded by the Operational Service are under construction and should be delivered to three E-SURFMAR participants on mid-2014 for evaluation. Later, series should be purchased by a few NMSs for their own purposes as well as by the Operational Service for E-SURFMAR objectives.
In 2011, KNMI started the implementation of the “half compression” technique, developed by the Operational Service, on their conventional ships. This technique allows significantly savings on communication costs. The GTS data transmission is presently ensured by Meteo France but could be shared with other NMSs. The savings are more significant if all NMSs do the same. DWD and Met Office, which operate the larger conventional VOS fleet, should follow.
A set of quality control tools was developed by Meteo France to monitor E-SURFMAR observation networks. Mainly based on comparisons with model outputs, they may be actually used to monitor any surface marine observation platform in the world reporting onto the GTS – not only those of the Operational Service.
Every month, the performances of the networks in matter of data availability, timeliness and quality, are assessed and compared to previous months and targets.