Ocean Prediction - modelling for the future

The ocean is the Earth’s largest ecosystem. It plays a major role in regulating the weather and climate of the planet. In addition, the ocean moderates global warming through CO₂ absorption and its massive heat capacity. 

Figure 1: The ocean’s surface layer (0-40 metres) absorbs atmospheric carbon that gets transported to the deep ocean in specific areas near the poles (noted by Carbon sink) where the mixed layer and deep ocean interface. (Map by Robert Simmon, NASA adapted from the IPCC 2001 and Rahmstorf 2002).

The United Nations indicates that 40% of the world’s population – nearly 2.4 billion people – live within 100 km of coasts and estimates the size of coastal economies to be between US$ 3–6 trillion a year¹. Coastal areas host essential infrastructure such as ports, harbors, desalination plants, power plants, aquaculture pens, etc. The ocean provides food, enables trade and has a strong role in many indigenous cultures. Knowledge of its physical characteristics and of the biological life it contains contributes to tourism, fisheries, maritime transportation, renewable and non-renewable energy extraction, and much more. The ocean itself can be a source of minerals and medical ingredients.

Thus, the importance of marine and coastal safety and ocean resilience cannot be underestimated. Climate information is essential to guide future coastal development and to adapt existing infrastructure to mitigate the impacts of hazardous marine and ocean weather. Impact-based early warnings for natural hazards and climate prediction and projections help coastal communities and businesses to avoid risks and increase resilience.

The ocean itself is an essential Earth System component in climate change projections. Understanding the ocean is fundamental to understanding the planet and the changes that are being wrought upon it by human activity. Today the ocean ecosystem and physical characteristics are being impacted by human actions, and the repercussions will be felt by all. There can be no delay in furthering knowledge and comprehension of the ocean, its interaction with the atmosphere and the impact of humanity on the ocean.

History

Historically, the exploration and exploitation of the ocean has gone hand in hand with the growth in knowledge about it, and the atmosphere above it. It was at the initiative of the oceanographer, meteorologist and astronomer Matthew Maury, then a lieutenant in the U.S. Navy, that the First International Meteorological Conference was held in Brussels in 1853 to achieve a uniform system of meteorological observations at sea. The Conference paved the way for the creation some 20 years later of the International Meteorological Organization, the predecessor of WMO.

The same Matthew Maury was one of the first to publish met-ocean contributions in a book entitled “The Physical Geography of the Sea and its Meteorology” (Maury 1864). His marine environmental perspectives on winds and currents led, amongst others, to a decrease in transit times across the world’s ocean, resulting in economic and safety benefits. Indeed, his pioneering work laid the foundation for modern marine meteorology. Today, there is still an urgent need for oceanographers to continue Maury’s quest and to share their knowledge with those far and near with all who benefit from or are impacted by the ocean.

Surface drifts and currents of the oceans (Source: NOAA)

Operational oceanography

Operational oceanography can be described as the provision of routine oceanographic information needed for decision-making purposes. The core components of operational oceanographic systems are a multi-platform observation network, a data management system, a data assimilative prediction system and a dissemination/accessibility system. These are interdependent, necessitating communication and exchange between them, and together provide the mechanism through which a clear picture of ocean conditions, in the past, present and future, can be seen. As is the case with the atmosphere, ocean prediction spans multiple time scales, from hours to days up to monthly and seasonal predictions.

The advances in ocean observations and prediction systems over the last 20 years have made operational oceanography infrastructure critical to a wide range of marine activities. Predictions –ranging in timescales from the immediate, to support safety and for tactical decisions, through to seasonal and longer timeframes, to inform planning and resilience activities – all require operational oceanography.

Differences in ocean and atmosphere

Air and water have extremely different properties that are exemplified by comparing the atmosphere and the ocean. The weight of the top 10 m of the ocean is equivalent to the weight of the entire atmosphere above it. The heat capacity of the top 2.5 m of the ocean is equivalent to that of the entire atmosphere above it. Additionally, the top 2.5 cm of the ocean contains the same amount of water as the entire atmosphere above it. While both atmosphere and ocean are governed by the same equations of motion, their circulation characteristics, scales of motion and properties are markedly different. The interaction between these two domains is also one of the fundamental processes driving weather and climate on Earth.

From a WMO perspective, ocean and atmospheric prediction are intrinsically linked through physical processes that are increasingly taken into consideration by modellers in both domains. At time scales of less than a few days, the interaction between the ocean and atmosphere has a big influence on weather in certain locations, such as near coasts that experience ocean upwelling, particularly when upwelling is linked to sudden changes in ice cover. At time scales beyond a few days, the ocean-atmosphere interaction contributes to weather forecasts over all locations and its importance increases with greater lead times. At seasonal and climate prediction scales, the coupling of ocean and atmosphere in the prediction systems is crucial.

Weather forecasting, such as for tropical cyclone formation and intensity, and longer-term predictions, such as for seasonal precipitation, are reliant on temperature and current observations in the ocean (Weller et al 2019).

Status of Ocean Prediction

An example of a one-page state of the ocean summary from ocean reanalysis systems of the European Union’s Copernicus Marine Service (source: Annual Ocean State Report (von Schuckmann et al 2019)

To understand the status of ocean prediction, one must first get an overview of the current state of ocean prediction systems and the international network that unites them to get an outlook on future improvement of the science behind ocean prediction, the prediction system capacity and the potential for further integration of ocean systems into seamless Earth System models. The maturing of oceanographic observations, forecast systems and research, the core ocean forecasting disciplines of data assimilation, ocean modelling, forecast verification and observing system evaluation is now enabling new research and operational areas to flourish.

A key to any prediction system is the real-time availability of observations from surface and from space-based platforms. The important differences between the atmosphere and the ocean become very apparent here. From a remote sensing satellite perspective, the ocean is less transparent and therefore less measurable at depth than the atmosphere. Satellite-based information, therefore, is in large part only available for the very surface of the ocean. However, satellite-based altimeters that measure sea surface height are an oceanographic remote sensing strength. Ocean height is representative of the depth integrated processes between ocean surface and bottom, and satellite altimetry enables the definition of large-scale ocean eddies in real time in the prediction systems. Additionally, altimetry enables tracking of long-term changes in ocean depth, such as sea level rise.

The Tropical Pacific Observing System (TPOS), which measures long-term changes in ocean-atmosphere heat exchange, was designed in the 1980s to improve the scientific understanding of the El Niño–Southern Oscillation (ENSO) phenomenon in order to better predict ENSO events. It has since provided vital data contributing to improved ENSO forecasts for decisions about agriculture, for example (Hansen et al., 1998; Chiodi and Harrison, 2017). Ocean observing systems have been developed for the Atlantic (PIRATA) and Indian Ocean (RAMA) following the TPOS model. TPOS is also adapting to meet the observational, experimental and operational needs of today and the future.

Data Assimilation – Data assimilation schemes vary among ocean forecasting groups. The primary objective is to minimize the misfit of model results with the observations while respecting the rules of physics. Observations assimilated in ocean forecast systems now include altimetry, ocean colour, surface velocities, sea ice and data from emerging platforms such as ocean gliders. Many systems now employ multi-model approaches or ensemble modelling techniques. A key upcoming change in data assimilation will be the arrival of the Surface Water and Ocean Topography (SWOT) altimeter, which will provide a true two-dimensional picture of the ocean surface topography with roughly 2 km resolution, rather than along satellite track measurements, where the tracks are interspaced by 200 km and separated over time.

Short term prediction – Short term ocean predictions encompass timescales from the next few hours to ten days or more and are often referred to as forecasts. There has been significant progress in ocean forecasting in recent years (Bell et al. 2015, Davidson et al 2019). Improvements of forecasting systems have included increased resolution (horizontal and vertical), inclusion of tides, sea ice drift and thickness, ecosystem approaches, improvement to mixing biases and extending regional mode areas [e.g., polar regions and progress of coupled modeling (wave coupling, sea ice, hurricane models, etc.)].

Short-to-medium-term coupled ice–ocean–wave– atmosphere prediction is being used to improve weather forecasts on the timescale of three days to two weeks. This will enable safer at-sea and coastal operations through improved prediction of extreme weather and climate events such as tropical cyclones. Increased activities in the high latitudes are also driving the further development of operational ice and ocean prediction.

Sub-seasonal to Seasonal prediction – Unlike large-scale atmospheric events which evolve on daily time scales, large scale ocean events typically evolve on weekly to monthly time scales and include marine heat waves and variations in sea level that can cause fair-weather flooding and exacerbate the flood risks of tropical and extratropical storms.

Sub-seasonal to seasonal prediction, with a forecast range longer than two weeks but less than a season, is now routinely performed using coupled ocean-atmosphere models. At lead times longer than two weeks, coupling of the atmosphere to the ocean contributes, for example, to predictability of monsoon variations and the Madden Julian Oscillation (e.g., Woolnough et al., 2007). In addition, satellite observations suggest that midlatitude ocean mesoscale eddy–induced sea surface temperatures can influence the atmospheric planetary boundary layer, which may drive predictability of winter storm-tracks on sub-seasonal to seasonal timescales (Saravanan and Chang, 2019).

Sub-seasonal prediction of regional variations in sea surface temperature and near-surface currents is also of direct interest for a wide range of activities and enterprises including management of fisheries, offshore mining activities and ocean transportation.

Coastal Prediction – Along coasts, decision-makers looking after increasingly populated and urbanized coastal areas are benefiting from coastal operational oceanography. This is because operational oceanography is increasingly able to provide accurate information on phenomena such as coastal river plumes from sediments and nutrients, predicting the occurrence and evolution of harmful algae blooms, and coastal erosion.

Sea ice – Sea ice is also considered to be part of the coupled ocean system. Due to its insulating and reflective properties, sea ice regulates exchanges between the atmosphere and ocean. At the sub-seasonal to seasonal timescale, prediction systems increasingly account for sea ice, either to improve the forecasts themselves, or to provide dedicated sea-ice forecasts. Sub-seasonal prediction of sea ice has wide potential applications as well (for example ship routing) but these have not yet been fully harnessed (Chevallier et al. 2019).

Climate reanalyses and ocean reporting – In parallel with efforts by the climate community to generate “reanalyses” of past conditions, the state of the ocean analysis aims to recreate ocean conditions over the last 30 years at global and regional scales. Three-dimensional analyses of the past and present ocean state at global-to-coastal scales are being developed based on the same modelling and assimilation infrastructure used for ocean forecasting. This follows the approach of atmospheric reanalyses using available historical observations to generate physically-consistent data cubes. The annual Ocean State Report (von Schuckmann et al 2019) of the European Union’s Copernicus Marine Service is a premier example of careful analysis of a year’s worth of analysis data against a historical context. A summary graph provides large scale trends of the main ocean variables in various regions of the globe.

Communication and verification – The communication and dissemination of information to downstream users has improved. Nowadays, the dissemination of outputs from forecasting systems is akin to the approaches taken by WMO in the distribution of numerical weather prediction products. Most ocean forecasting systems are also now investing in verification, monitoring and validation efforts to be able to show the value of their products to their users.

Strengthening ocean predictions through partnerships

The United Nations Decade of Ocean Science for Sustainable Development provides an opportunity to further galvanize operational oceanography. The Decade brings momentum for international and national ocean communities to come together to extend the network and science essential for the generation of comprehensive ocean information. A key goal of the Decade is a predicted ocean where society has the capacity to understand current and future ocean conditions.

WMO and Intergovernmental Oceanographic Commission (IOC) of UNESCO have long recognized the value and need for ocean forecasting services and have worked together towards enablement and understanding of the full value chain on ocean prediction. In recent years, the Global Ocean Observing System (GOOS)² has emphasized this focus on ocean prediction to deliver relevant services for societal benefit. The international ocean forecasting community is collaborating across OceanPredict, GOOS, WMO, IOC, the Committee on Earth Observation Satellites (CEOS), and the Blue Planet initiative of the intergovernmental Group on Earth Observations (GEO). These partnerships reinforce the sharing of ideas and bring together the oceanographic and atmospheric science and modelling communities. Partnerships within national settings are also advancing ocean predictions for societal benefit. Examples from the Australian, Canadian and U.S. governments show the success of collaboration between meteorological and oceanographic institutions.

Bluelink Ocean Forecasting Australia

Bluelink is a partnership between the Australian Bureau of Meteorology (BOM), the Commonwealth Scientific and Industrial Research Organisation (CSIRO) and the Australian Department of Defence with collaborating partners that include the Integrated Marine Observing System, the Defence Science and Technology Group, National Computational Infrastructure, and the university sector.

The operational Bluelink ocean forecast system is used to transform physical oceanographic observations into coherent analyses and predictions. These analyses and predictions form the basis for information services about the marine environment and its ecosystem and can provide boundary data for weather predictions. Bluelink information services are available to marine industries – commercial fishing, aquaculture, shipping, oil and gas, renewable energy – government agencies – search and rescue, defence, coastal management, environmental protection – and other stakeholders – recreation, water sports, artisanal and sport fishing – who depend on timely and accurate information about the marine environment.

At its core, Bluelink consists of three inter-connected component systems at global, regional and near-shore (littoral zone) scales. The key scientific objective is to deliver reliable, operational ocean forecasts and reanalyses of the ocean mesoscale (global system), sub-mesoscale (regional system) and nearshore circulation (littoral zone system) at timescales from days to weeks. Beyond the traditional short-term forecasting of physical ocean properties (temperature, salinity, surface height, currents, waves), marine activities such as water quality and habitat management as well as climate monitoring increasingly rely on operational oceanographic data and products.

Typical example of (a) Bluelink reanalysis fields and (b) observed fields for the Tasman Sea. Colour shows SST, arrows show surface velocity. (This comparison is adapted from: www.marine.csiro.au/ofam1/bran1/br3p5_EAC_tv12/20120111. html.)

US Partnership in Ocean Modelling for an Earth System

GOFS forecast sea surface temperature, Day 7-1/2 (Source: U.S. Naval Research Laboratory)

Within the U.S., the National Oceanic and Atmospheric Administration (NOAA) and the Department of the Navy have partnered for well over a decade to develop and implement operational ocean predictions. The output of these models provides the basis for a variety of met-ocean forecasting services to support safe maritime operations, including tropical cyclone predictions, search and rescue, response to marine environmental emergencies, such as oil spills, and operations near the marginal sea-ice zone.

The U.S. Navy began running global ocean circulation models in 1999 (Rhodes et al. 2002). The present version of the Navy Global Ocean Forecast System (GOFS), operational in 2020, couples the Hybrid Coordinate Ocean Model with the Community Ice Code Sea Ice (CICE) model. In 2021, the Navy will operate a 1/25 degree GOFS with CICE and tides.

NOAA implemented its global Real-Time Ocean Forecasting System (RTOFS) in 2011. Initially based on the U.S. Navy’s development of GOFS, NOAA also incorporated the Navy Coastal Ocean Data Assimilation System (NCODA) into RTOFS. RTOFS Version 2.0, implemented in December 2020, incorporated an upgraded NOAA Ocean Data Assimilation (DA) system, RTOFS-DA.

In 2017, as part of the Hurricane Forecast Improvement Project, NOAA implemented a new regional scale coupled ocean-weather model. The Hurricanes in a Multi-scale Ocean coupled Non-hydrostatic (HMON) model provides forecasters with intensity and track guidance from 0 to 5 days to support the official warning and forecast products from the National Hurricane Center/ Regional Specialized Meteorological Centre (RSMC) Miami. Like HMON, the operational Hurricane Weather Research and Forecast model also uses coupled ocean states prescribed using initial and boundary conditions from RTOFS.

Global RTOFS forecast sea surface temperature anomaly, Day 8 (Source: NOAA)

Efforts such as these, representing just a subset of the U.S. operational ocean modelling efforts, are shaping the development of fully-coupled Earth system prediction capability. As part of a national effort codified in legislation as the Weather Research and Forecasting Innovation Act of 2017, NOAA is collaborating across the nation’s weather enterprise – government agencies, academia and the private sector – to improve its numerical weather prediction. NOAA will accomplish this through an Earth Prediction Innovation Center (EPIC), engaging the enterprise to accelerate scientific research and modelling contributions into the Unified Forecast System (UFS). As a community-based Earth system data assimilation and prediction system, the UFS will, over the next five years, lead to full Earth systems coupling – ocean, atmosphere, land, sea ice and the biosphere – for weather and climate applications. EPIC will facilitate this broad collaboration with a cloud development environment, code repository, observations and tools, and community support and engagement.

Canadian Partnership in Atmosphere- Ocean-Ice Monitoring, Coupled Prediction and Ocean Services

The Canadian Operational Network of Coupled Environmental PredicTion Systems (CONCEPTS) is a collaboration between three Federal Departments: Fisheries and Oceans Canada (DFO), Environment and Climate Change Canada (E3C) and the Department of National Defence (DND). The network works to develop and implement computer models that support ocean-ice forecasting advancements. The aim is to take advantage of breakthroughs in ocean modeling and new real-time global oceanographic observation systems to produce oceanographic forecast products and improve seasonal to inter-annual climate forecasts. This core network is leveraging multiple collaborations with academic institutions, the private sector and institutions abroad such as Mercator Ocean International.

An example of the prediction systems cascading approach used to generate products and services at high spatial and temporal resolution, for the Bay of Fundy, Canada. (Paquin et al 2019)

In order to facilitate collaboration across government departments and with external CONCEPTS partners, a three thrust strategy was implemented in 2009:

1. the collection and dissemination of measurements of physical properties of marine environments for assimilation in models to improve forecasts from environmental (weather, ice, wave and ocean) prediction systems in Canada
2. the development of coupled environmental prediction systems to improve analyses and forecasts from environmental (weather, ice, wave and ocean) prediction systems in Canada
3. the availability of CONCEPTS products and services for end users including:

(a) providing feedback to monitoring and prediction systems for continuous improvements

(b) enabling collaboration within and outside of CONCEPTS through the development and provision of discovery, visualization and accessibility systems of observation and model output.

The current suite of coupled atmosphere-ocean-ice prediction systems comprise:

1. Global Ice Ocean Prediction System (GIOPS) running at ¼ degree resolution³
2. Regional Ice Ocean Prediction System (RIOPS) running at 1/12 degree resolution over the North Pacific, the Arctic and the North Atlantic^4,5
3. Coastal Ice Ocean Prediction System (CIOPS) running at 1/36 degree over the Northwest Atlantic and the Northeast Pacific⁶
4. Great Lakes Water Cycle Prediction System (atmosphere, ocean, ice, hydrology) running at 2 km horizontal resolution
5. Hydrodynamic modelling over the St. Lawrence River from Montréal to Québec city

An example of product for E-Navigation application generated by a high resolution estuarian river 2D prediction system run by the Canadian Meteorology and Environmental Prediction Center H2D2 (Matte et al 2017)

S2S Project

Sea-ice thickness, mixed layer depth, and sensible heat fluxes over the ocean from the ECMWF extended-range control forecast starting on 2 December 2019 and verified on 16 January 2020 (45-day lead time).

To bridge the gap between medium-range weather forecasts and seasonal forecasts, the WMO World Weather Research Programme (WWRP) and the World Climate Research Programme (WCRP)⁷ launched the Sub-seasonal to Seasonal prediction project (S2S). The main goal of the project is to improve forecast skill and understanding of the sub-seasonal to seasonal timescale, and to promote the uptake of S2S predictions by operational centres and exploitation by the applications communities (www.s2sprediction. net). The first phase of the S2S ran from 2013 to 2017 and the second phase started in 2018 and will end in 2023. One research focus of Phase 2 is the sub-seasonal to seasonal predictability and prediction of ocean and sea-ice. S2S works in coordination with the Working Groups on Subseasonal to Interdecadal Prediction (WGSIP), on Data Assimilation and Observation Systems (DAOS), and on Predictability, Dynamics, Ensemble Forecasting (PDEF) to promote improved sub-seasonal predictions though better initialization of the ocean/sea-ice state and depiction of key ocean and sea-ice processes that provide predictability at sub-seasonal timescales.

A major achievement of the first phase of S2S was the establishment in 2015 of the S2S database containing near real-time sub-seasonal forecasts (up to 60 days) and re-forecasts (sometimes known as hindcasts) from 11 operational centres. Most of the S2S models are coupled ocean/sea-ice/atmosphere models, and the list of parameters available from the S2S database has always included sea-surface temperature and sea-ice cover.

Since January 2020, nine new ocean and sea-ice parameters have been added to the S2S database, consisting of 20° C isotherm depth, mixed-layer thickness, salinity and potential temperature in the top 300 m, surface currents, salinity, sea-surface height and sea-ice thickness. The availability of this extensive set of ocean and sea-ice variables substantially increases the power of the database for S2S coupled system research and to address key science questions. Currently, the new variables are available from four models – European Centre for Medium-Range Weather Forecasts (ECMWF), E3C, Chinese Meteorological Administration (CMA) and Météo-France – three weeks behind real-time forecasts (from January 2020) and the corresponding re-forecasts, which are produced in near real-time. In the coming year, the new variables will become available from an expanded set of S2S models.

As an example of what is already being done, Zampieri et al. (2018) evaluated the forecast skill of several models from the S2S database and found that some of them displayed significant skill in predicting sea-ice cover up to a month in advance. This important result suggests that state-of-the-art sub-seasonal to seasonal forecasts could be potentially useful for applications such as ship routing in Arctic regions. The availability of the new ocean variables should trigger new research studies on the predictability of high-impact ocean weather, such as heatwaves, which will provide insights on the possible use of these sub-seasonal forecasts for applications such as fishery. The image below provides an example of a possible use for this ocean data in ocean weather maps. In this example, sea-level anomalies relative to the climate are issued for a forecast lead time of 3 to 4 weeks.

The addition of the new parameters will also make the S2S database more apt to lead to a better understanding of air-ice-ocean interactions at the sea-ice margin, as illustrated in below. It will also help diagnose the evolution of ocean drift with forecast lead time in the S2S forecasts.

In order to coordinate these activities, S2S Phase 2 includes an ocean subproject, which will develop a protocol for coordinated case studies that can be conducted by centres doing S2S prediction of specific ocean extreme events and air-sea interactions, for example the onset of ENSO. Examples could include a predictability study of a prominent coral bleaching event in 2017, and the intra-seasonal air-sea interaction at the onset of the 2015/2016 El Niño event.

Organization of international ocean prediction

Recent literature^8,9 has documented the strength of the value cycle approach (Day, 1999¹⁰) in the research to operations to services technology transfer. In particular, Ruti et al. (2020) provide a description of this cycle in the current meteorological context and as a key component to enable the Earth Systems approach. This value chain links the production and delivery of these services to user decisions and to the outcomes and values resulting from those decisions. User feedback is then fed back to Research and Operations in order to further improve services. Similar thinking is taking place in other disciplines.

The image on the right (adapted from Schiller et al 2019) depicts the marine value chain. There are two key international initiatives which support ocean predictions: OceanPredict and GOOS. For over two decades, OceanPredict and its predecessors have been focused on research and operational implementation of ocean forecasting systems. GOOS has provided ocean observations for initializing and validating ocean predictions, driven by the collaboration between the WMO and IOC.

OceanPredict

In the late 1990s, the international Global Ocean Data Assimilation Experiment (GODAE) was launched to (i) demonstrate the feasibility and utility of ocean monitoring and forecasting on the daily-to-weekly timescale and (ii) contribute to building a global operational oceanography infrastructure (Smith and Lefebvre, 1997; Schiller et al., 2018). Building on its success, GODAE OceanView was established in 2009 (Bell et al., 2009) to define, monitor and promote actions aimed at coordinating and integrating research associated with multi-scale and multi-disciplinary ocean analysis and forecasting systems.

In 2019, GODAE OceanView became OceanPredict which continues to expand its activities with an added emphasis on ocean prediction as part of the broader network of international initiatives linked to operational oceanography. OceanPredict is thus developing close partnerships with international entities including the WMO, the IOC, GOOS and GEO Blue Planet.

OceanPredict is supported by 14 countries. However scientific and technical participants from any country are welcome in OceanPredict and its task teams. Participants from 41 countries attended the OceanPredict19 symposium (Vinaychandran et al 2020), including scientists from operational prediction centres, government agencies, academia and private consortiums/companies.

Most OceanPredict groups are integrally linked to or are a part of numerical weather and environmental prediction centres such as NOAA, E3C, MétéoFrance and the Japan Meteorological Agency to name a few. In fact, seven of the nine WMO World Meteorological Centres (WMC’s) are members of OceanPredict. Additionally, all of the prediction centres have academic collaborators that support some of their research. OceanPredict coordinates research and development activities in ocean data assimilation, ocean system evaluation, marine ecosystem prediction, coastal ocean prediction, atmosphere ocean coupled prediction systems, and intercomparisons and validations of ocean prediction systems.

Collaboration on these themes includes academic researchers, researchers from operational prediction agencies, and development teams that support development, operations and dissemination at prediction agencies. Through international workshops, under the leadership of a dedicated Science Team, OceanPredict brings the various communities together to advance the science and applications of ocean prediction. Leading experts from the WMO community are keynote speakers, and some workshops are joint events with WMO partners such as ECMWF. The OceanPredict Science Team has three core objectives:

• assessments of forecast system and component performance combined with component improvements
• initiatives aiming to exploit the forecasting systems for greater societal benefit
• evaluations of the dependence of the forecasting systems and societal benefits on the components of the observation system.

Outlook

As ocean forecast models progress, it will become increasingly important to define and project what type of events will be predictable by ocean prediction systems and by coupled atmosphere ocean prediction systems with useful accuracy and confidence intervals.

An aspect to keep in mind is the ability of systems and users to consume prediction products. A good example of this is E-navigation to support maritime safety, where new file standards and methodologies will enable ship bridge-embedded or hand-held navigation systems to fully exploit numerical output from ocean and atmospheric prediction systems in real time. This will enable advanced route planning software, but also enable numerical engineering models (digital twins) of a ship, translating environmental prediction information into ship impact information. It is important to note that for most at-sea activities, the user wants complete and coherent marine environmental information and predictions, which could include variables like wind, waves, ice conditions, atmospheric temperature, atmospheric pressure, water level, water temperature and water salinity.

Links to meteorology and WMO

In moving forward, the link between operational oceanography and operational meteorology needs to strengthen. More specifically, the full value chain in operational oceanography will require both international and national frameworks to deliver overall end user value. There is already significant interaction with oceanography groups in the WMO community, such as evidenced by the emerging use of the WMO Information System (WIS) for ocean observations. In moving forward, while most weather prediction centres include ocean prediction in their activities, it will be important to strengthen the ocean-weather relationship from observations through to prediction and end use.

Leveraging WMO Systems in the future of Ocean Prediction

The WMO has well-developed frameworks covering the full value chain for meteorological services that evolves to meet demands while enabling new work and information flows across the whole Earth System value chain. Three of these frameworks are discussed in this section with respect to the ocean prediction value chain.

Figure 4: Data coverage by the three major components of the Global Observing System (GOS) (i.e., ships, drifting buoys, and moored buoys) based on information received by Météo-France through the GTS in 2018 (see legend for symbol details). (from Front. Mar. Sci., 30 August 2019 | https://doi.org/10.3389/ fmars.2019.00419 )

WIGOS

The WMO Integrated Global Observing System (WIGOS) provides an overarching framework for integrating the various sources of observations that contribute to WMO application areas. WMO Rolling Review of Requirements (RRR) compares observational user requirements with observing systems capabilities to determine how the design of WIGOS needs to evolve. Together with impact studies for the identification of observational gaps, the RRR is used to prioritize the evolution of the global observing systems and to recommend key actions to WMO Members and other significant programmes to address gaps.

The WIGOS framework provides a systematic approach that can enable ocean prediction groups to implement systematic evaluations of observed impacts against forecasts in order to appraise performance across the whole ocean prediction value chain. In particular, implementation of a Rolling Review Process on the Ocean Forecast side would better connect the full oceanographic value chain, and ensure that investment in ocean observations provide the best value for money with respect to better ocean prediction information services.

WIS

The WMO Information System (WIS) join NMHSs and regions together for data exchange, management and processing. At present most ocean observations used in near real-time prediction systems are transmitted via the WMO Global Telecommunications System (GTS) (See figure below for coverage of GTS transmitted observations in 2018).

GTS has also proven to be an effective channel for Tsunami Warnings by delivering messages with a delay that is, in most cases, less than two minutes. WMO is evolving WIS/GTS to use new technologies for data exchange, and WIS 2.0 will provide better means to subscribe to data streams and effective ways to deliver warning messages.

In the future, enabling ocean data on the WIS will have many dividends. WIS provides the global infrastructure for the exchange of data and information between all NMHSs and incorporates the long-established GTS for the delivery of real-time observational data (and increasingly those metadata needed to make best use of the real-time data) needed for their operational requirements. While the GTS remains the standard method of global data exchange between NMHSs and fulfills their operational requirements and applications, the academic community and the public have a clear need for a more streamlined and consolidated data management architecture, which should provide access to data and metadata in a common format.

GDPFS

The GDPFS centres responsible for weather forecasting up to 30 days (upper) and for long-range and climate forecasting (lower).

The WMO Global Data-Processing and Forecasting System (GDPFS) enables all NMHSs to make use of advances in numerical weather prediction (NWP) by providing a framework for sharing data related to operational meteorology, hydrology, oceanography and climatology. The GDPFS is a cascading process that brings the NWP strength of WMO’s global centres (WMCs) down to regional centres (RSMCs) then to NMHSs in a coordinated way. RSMCs enable the delivery of harmonized services, including for marine and ocean matters. More than 40 RSMCs have responsibility to support ocean related services including for marine meteorology, ocean wave prediction, severe weather and tropical cyclones.

As previously mentioned, seven of the nine designated WMCs have ocean prediction systems encompassed in the OceanPredict initiative that run in coupled or uncoupled prediction modes as part of their day-to-day operations. Encompassing ocean prediction systems within such a framework has many advantages, including integration of scientific advances in ocean predictions and applying new observation systems (ie. SWOT) into operational ocean/environmental prediction systems.

Such weather/ocean collaboration is already underway, as evidenced by an upcoming ECMWF and OceanPredict meeting on data assimilation in May 2021. This collaboration needs to flourish. It is anticipated that, under the UN Decade of Ocean Science for Sustainable Development, a framework can be put in place for the full operational oceanographic value chain, akin to that of the GDPFS. By the end of the Decade, a fully integrated value chain for Marine Environmental Prediction (Operational Oceanography and Meteorology) is envisaged, however, an open question remains: how should its ocean component be built? The options are to build a full ocean value chain framework first or to build ocean components into the existing elements of the meteorological value chain put together by the WMO.

In moving forward, the link between operational oceanography and operational meteorology needs to strengthen. More specifically, the full value chain in operational oceanography will require both international and national frameworks to deliver overall end user value. There is already significant interaction with oceanography groups in the WMO, such as evidenced by the use of GTS for ocean observations. In moving forward, while most weather prediction centres include ocean prediction in their activities, it will be important to strengthen the ocean weather relationship from observations through to prediction and end use products and services.

Partnerships for the future

The vision of the WMO, as stated in the WMO Strategic Operational Plan 2020-2023, is “to see a world where all nations, especially the most vulnerable, are more resilient to the socioeconomic consequences of extreme weather, climate, water and other environmental events and underpin their sustainable development through the best possible services, whether over land, at sea or in the air.” To achieve this, WMO is embracing an Earth System approach that will enable access to and use of numerical analysis and Earth System prediction products at all temporal and spatial scales from the WMO Seamless GDPFS.

In order to continuously improve products and services, all the key components of the Earth System need to be integrated into seamless data assimilation and prediction systems, leveraging WIGOS, WIS 2.0 and the WMO Seamless GDPFS. Within the WMO community the WMO Reform has provided the framework to achieve the integration of disciplines required to achieve this goal. It also provides mechanisms to better partner with key relevant national and international organizations, academic institutions and the private sector. One single entity cannot achieve this by itself, and resource pressures are such that replication of existing infrastructure will not be possible. In addition, the required high-performance computing, storage and telecommunication will likely exceed what individual Nations can afford.

Public, academic and private partnerships are therefore essential. Leveraging existing global, regional and national infrastructure will allow all communities to benefit from the information available to feed their decision-making systems. WMO and its partners firmly believe that together we will achieve the grand challenges facing humanity today – as underlined in the United Nations Sustainable Development Goals – and we will be better prepared to find solutions for those to come in the future.

Authors

Fraser Davidson, Co-Chair, OceanPredict Science Team

Andrew Robertson, Co-chairs of the WWRP/WCRP S2S project

Frédéric Vitart, Co-chairs of the WWRP/WCRP S2S project

Anthony Rea, WMO Secretariat

Michel Jean, President, WMO INFCOM

Andreas Schiller, CSIRO Oceans and Atmosphere

Thomas J. Cuff, Director, Office of Observations, National Weather Service, NOAA

Sarah Grimes, WMO Secretariat

Eunha Lim, WMO Secretariat

Estelle de Coning, WMO Secretariat

Peiliang Shi, WMO Secretariat

Footnotes

¹ https://www.un.org/sustainabledevelopment/wp-content/uploads/2017/05/Oc…

² Co-sponsored by IOC, WMO, United Nations Environment Programme (UNEP) and International Science Council (ISC)

³ G. Smith et al., QJRMS, Volume142, Issue695, January 2016 Part B, Pages 659-671

⁴ JF Lemieux et al., 2016, QJRMS, Volume142, Issue695, January 2016 Part B, Pages 632-643

⁵ Smith, G.C., Liu, Y., Benkiram, M., Chikhar, K., Surcel Colan, D., Testut, C.E., Dupont, F., Lei, J., Roy, F., Lemieux, J.F., and Davidson, F., 2020. The Regional Ice Ocean Prediction System v2: a pan Canadian ocean analysis system. Geoscientific Model development Discussions, pp1-49.

⁶ Paper in preparation

⁷ Co-sponsored by WMO, IOC and ISC

⁸ WMO, 2015: Valuing weather and climate: Economic assessment of meteorological and hydrological services. WMO-1153, 308 pp., https://library.wmo.int/doc_num.php?explnum_id=3314.

⁹ Ruti et al., Advancing Research for Seamless Earth System Prediction, Bull. Amer. Met. Soc. 2020 DOI: https://doi.org/10.1175/BAMS-D-17-0302.1

¹⁰ Day, G. S., 1999: The Market Driven Organization: Understanding, Attracting, and Keeping Valuable Customers. Free Press, 285 pp

References

Bell, M. J., , A. Schiller, , P.-Y. LeTraon, , N. R. Smith, , E. Dombrowsky, , and K. Wilmer- Becker, 2015: An introduction to GODAE OceanView. J. Oper. Oceanogr., 8, s2–s11, doi:10.1080/1755876X.2015.1022041. (CHECKED)

Brassington, G., and Coauthors 2015: Progress and challenges in short- to medium-range coupled prediction. J. Oper. Oceanogr., 8, s239–s258, doi:10.1080/1755876X.2015.1049875.

European Commission, 2015: Marine Environment Monitoring Service fact sheet. [Available online.]

Fu, L.-L., , and R. Ferrari, 2008: Observing oceanic submesoscale processes from space. Eos, 89, 488–489, doi:10.1029/2008EO480003.

Gehlen, M., and Coauthors, 2015: Building the capacity for forecasting marine biogeochemistry and ecosystems: Recent advances and future developments. J. Oper. Oceanogr., 8, s168–s187, doi:10.1080/1755876X.2015.1022350.

GODAE OceanView Science Team, 2014: GODAE OceanView Strategic Plan 2015–2020. UK Met Office, 48 pp.

Legler, D., and Coauthors, 2015: The current status of the real-time in situ global ocean observing system for operational oceanography. J. Oper. Oceanogr., 8, s189–s200, doi:10.1080/1755876X.2015.1049883.

Maury, M. F. (1864). The Physical Geography of the Sea and its Meteorology. Mineola, NY: Courier Corporation.

Matte, Pascal & Secretan, Yves & Morin, Jean. (2017). Hydrodynamic Modeling of the St. Lawrence Fluvial Estuary. I: Model Setup, Calibration, and Validation. Journal of Waterway, Port, Coastal and Ocean Engineering. 143. 10.1061/(ASCE)WW.1943-5460.0000397.

Oke, P., and Coauthors, 2013: Evaluation of a near-global eddy-resolving ocean model. Geosci. Model Dev., 6, 591–615, doi:10.5194/gmd-6-591-2013.

Paquin, J.P.,  Y. Lu, S. Taylor, H. Blanken, G. Marcotte, X. Hu, L. Zhai, s. Higginson,. S. Nudds, J. Chanut, G.C. Smith and N. Bernier 2019, High-resolution modelling of a coastal harbour in the presence of strong tides and significant river runoff, Ocean Dynamics, https://doi.org/10.1007/s10236-019-01334-7

Rhodes, Robert C, et al. “Navy Real-Time Global Modeling Systems.’ Oceanography, vol. 52, no. 1, 2002, pp 29-43.

Ryan, A. G., and Coauthors, 2015: GODAE OceanView class 4 forecast verification framework: Global ocean inter-comparions. J. Oper. Oceanogr., 8, s98–s112, doi:10.1080/1755876X.2015.1022330.

Ryabinin V, Barbière J, Haugan P, Kullenberg G, Smith N, McLean C, Troisi A, Fischer A, Aricò S, Aarup T, Pissierssens P, Visbeck M, Enevoldsen HO and Rigaud J (2019) The UN Decade of Ocean Science for Sustainable Development. Front. Mar. Sci. 6:470. doi: 10.3389/fmars.2019.00470 February 5, 2021, 5:46 AM

Schiller, A., and Coauthors, 2015: Synthesis of new scientific challenges for GODAE OceanView. J. Oper. Oceanogr., 8, s259–s271, doi:10.1080/1755876X.2015.1049901.

Smith, N., , and M. Lefebvre, 1997: The Global Ocean Data Assimilation Experiment (GODAE). Proc. Monitoring the Oceans in the 2000s: An Integrated Approach, International Symposium, Biarritz, France.

Tonani, M., and Coauthors, 2015: Status and future of global and regional ocean prediction systems. J. Oper. Oceanogr., 8, s201–s220, doi:10.1080/1755876X.2015.1049892.

Speich S et al 2019m  Editorial : Oceanobs’19 : An Ocean of Opportunity, Front. Mar. Sci., 06 September 2019 | https://doi.org/10.3389/fmars.2019.00570

Schiller et al 2016    https://doi.org/10.1175/BAMS-D-15-00102.1

Weller RA, Baker DJ, Glackin MM, Roberts SJ, Schmitt RW, Twigg ES and Vimont DJ (2019) The Challenge of Sustaining Ocean Observations. Front. Mar. Sci. 6:105. doi: 10.3389/fmars.2019.00105

Chevallier, C, F. Massonnet, H. Goessling, V. Guémas, and T. Jung, 2019: The Role of Sea Ice in Sub-seasonal Predictability. In A. W. Robertson & F. Vitart (Eds.), Sub-seasonal to seasonal prediction: The gap between weather and climate forecasting (pp. 1–15). Amsterdam, Netherlands: Elsevier.

Saravanan, R., and P. Chang, 2019: Ocean-Atmosphere Interaction and Its Relevance to S2S Prediction.  In A. W. Robertson & F. Vitart (Eds.), Sub-seasonal to seasonal prediction: The gap between weather and climate forecasting (pp. 1–15). Amsterdam, Netherlands: Elsevier.

Woolnough, S. J., F. Vitart and M. A, Balmaseda, 2007: The role of the ocean in the Madden-Julian Oscillation: Implications for MJO prediction Quart. J. Roy. Meteor. Soc., 133, 117-128.

Vinayachandran, Pn & Davidson, Fraser & Chassignet, Eric. (2019). Toward Joint Assessments, Modern Capabilities, and New Links for Ocean Prediction Systems. Bulletin of the American Meteorological Society. 101. 10.1175/BAMS-D-19-0276.1.

Visbeck, M. (2018). Ocean science research is key for a sustainable future. Nat Commun. 9:690. doi: 10.1038/s41467-018-03158-3

Karina von Schuckmann ((Editor)), Pierre-Yves Le Traon ((Editor)), Neville Smith (Chair) ((Review Editor)), Ananda Pascual ((Review Editor)), Samuel Djavidnia ((Review Editor)), Jean-Pierre Gattuso ((Review Editor)), Marilaure Grégoire ((Review Editor)), Glenn Nolan ((Review Editor)), Signe Aaboe, Eva Aguiar, Enrique Álvarez Fanjul, Aida Alvera-Azcárate, Lotfi Aouf, Rosa Barciela, Arno Behrens, Maria Belmonte Rivas, Sana Ben Ismail, Abderrahim Bentamy, Mireno Borgini, Vittorio E. Brando, Nathaniel Bensoussan, Anouk Blauw, Philippe Bryère, Bruno Buongiorno Nardelli, Ainhoa Caballero, Veli Çağlar Yumruktepe, Emma Cebrian, Jacopo Chiggiato, Emanuela Clementi, Lorenzo Corgnati, Marta de Alfonso, Álvaro de Pascual Collar, Julie Deshayes, Emanuele Di Lorenzo, Jean-Marie Dominici, Cécile Dupouy, Marie Drévillon, Vincent Echevin, Marieke Eleveld, Lisette Enserink, Marcos García Sotillo, Philippe Garnesson, Joaquim Garrabou, Gilles Garric, Florent Gasparin, Gerhard Gayer, Francis Gohin, Alessandro Grandi, Annalisa Griffa, Jérôme Gourrion, Stefan Hendricks, Céline Heuzé, Elisabeth Holland, Doroteaciro Iovino, Mélanie Juza, Diego Kurt Kersting, Silvija Kipson, Zafer Kizilkaya, Gerasimos Korres, Mariliis Kõuts, Priidik Lagemaa, Thomas Lavergne, Heloise Lavigne, Jean-Baptiste Ledoux, Jean-François Legeais, Patrick Lehodey, Cristina Linares, Ye Liu, Julien Mader, Ilja Maljutenko, Antoine Mangin, Ivan Manso-Narvarte, Carlo Mantovani, Stiig Markager, Evan Mason, Alexandre Mignot, Milena Menna, Maeva Monier, Baptiste Mourre, Malte Müller, Jacob Woge Nielsen, Giulio Notarstefano, Oscar Ocaña, Ananda Pascual, Bernardo Patti, Mark R. Payne, Marion Peirache, Silvia Pardo, Begoña Pérez Gómez, Andrea Pisano, Coralie Perruche, K. Andrew Peterson, Marie-Isabelle Pujol, Urmas Raudsepp, Michalis Ravdas, Roshin P. Raj, Richard Renshaw, Emma Reyes, Robert Ricker, Anna Rubio, Michela Sammartino, Rosalia Santoleri, Shubha Sathyendranath, Katrin Schroeder, Jun She, Stefania Sparnocchia, Joanna Staneva, Ad Stoffelen, Tanguy Szekely, Gavin H. Tilstone, Jonathan Tinker, Joaquín Tintoré, Benoît Tranchant, Rivo Uiboupin, Dimitry Van der Zande, Karina von Schuckmann, Richard Wood, Jacob Woge Nielsen, Mikel Zabala, Anna Zacharioudaki, Frédéric Zuberer & Hao Zuo (2019) Copernicus Marine Service Ocean State Report, Issue 3, Journal of Operational Oceanography, 12:sup1, S1-S123, DOI: 10.1080/1755876X.2019.1633075

Zampieri, L., Goessling, H. F., & Jung, T. (2018). Bright prospects for Arctic sea ice prediction on sub-seasonal time scales. Geophysical Research Letters, 45, 9731–9738. https://doi.org/10.1029/2018GL079394.