Advances in global wave modelling

At MetOcean Solutions, we continuously improve our models to ensure the highest possible performance.

Our science team has recently made great improvements in global wave hindcasting thanks to using more accurate historical winds and studying the effect of icebergs and ocean currents in ocean waves.

MetOcean’s Southern Ocean Programme in partnership with Defence Technology Agency has been collecting wave data in the Southern Ocean over the last 2 years (find out more at www.metocean.co.nz/southern-ocean). The area presents the highest modelling errors, and the data gathered is helping to reduce that. This is a crucial achievement due to the energetic swells constantly generated in this part of the ocean that have far reaching effects. Consequently, it will result in better wave prediction over coastal areas.

MetOcean Solutions’ physical oceanographer Dr Jorge Perez, responsible for improving wave hindcasting and forecasting capabilities, says the analysis undertaken and the historical reconstruction has made it possible to minimize errors in wave data from deep waters, allowing better boundary conditions for high-resolution grids in coastal regions.

“Lessons learned such as the importance of currents and icebergs will result in advancements to our wave forecasting operational systems and therefore, historical data and wave predictions of the highest quality to end users.”

Clear gains are apparent, with an approximate 30% improvement in model skill demonstrated overall. The resulting improvements for the year of 2015 are shown as an example in the figure below.

 
fig 1-top.png
 
 
Example of bias comparison with model improvements (top) and reference simulation (bottom) for the year 2015.

Example of bias comparison with model improvements (top) and reference simulation (bottom) for the year 2015.

 

MetOcean’s advancements in global wave modelling were presented by Dr Perez at Spanish Conference on Coastal and Port Engineering, XV Jornadas Españolas de Ingeniería de Costas y Puertos, held last week in Málaga, Spain.

The conference is a biennial scientific-technical event, gathering experts and decision makers to facilitate knowledge exchange between all sectors engaged in coastal and port activities. For more information, visit www.costasypuertos2019.com

MetOcean Solutions is a division of New Zealand’s National Meteorological Service.

The full abstract is provided below.


Avances en modelado de oleaje global

Pérez, Jorge* Rapizo, Henrique* Guedes, Rafael* y Durrant, Tom*

*Metocean Solutions, New Zealand Meteorological Service.

1.       Introducción

El modelado de oleaje a escala global ha experimentado un rápido desarrollo en los últimos años dando lugar a reconstrucciones históricas cada vez más precisas (e.g., Durrant et al., 2014; Pérez et al., 2017). No obstante, incluso las bases de datos más recientes presentan errores significativos en ciertas regiones. Esto plantea una doble problemática. Por un lado, la creciente internacionalización de los intereses de empresas e instituciones hace evidente la necesidad de contar con datos de calidad en todo el mundo. Por otro lado, los errores en una ubicación específica a menudo son consecuencia de errores a miles de kilómetros de distancia, en la zona de generación o propagación. Por ejemplo, el Océano Antártico es la región que actualmente presenta mayores errores; en parte porque es un área muy compleja desde el punto de vista del modelado y en parte por la tradicional escasez de medidas instrumentales en el hemisferio sur. No obstante, es crucial reducir dichos errores, ya que constituye la zona de generación de swells muy energéticos que alcanzan las costas de regiones mucho más pobladas.

La forma de más obvia de reducir errores es utilizar forzamientos (i.e., viento, hielo y corrientes) de mayor calidad. En este análisis se han utilizado y comparado las bases de datos más recientes para obtener la combinación de forzamientos que resulta en menores errores en el modelado de oleaje a escala global. Como es habitual en han analizado vientos y cobertura de hielo, en este caso del reanálisis CFSR generado por NCEP-NCAR y del reanálisis ERA5 generado recientemente por el centro europeo para predicciones de medio plazo (ECMWF). Adicionalmente, se han analizado el efecto de la probabilidad de icebergs derivada de imágenes de satélite y de las corrientes oceánicas de tres bases de datos: CFSR, HYCOM y GLORYS. La evaluación de los resultados se ha basado principalmente en mediciones de satélite, pero se ha complementado con boyas y drifters recientemente desplegados en Australia y Nueva Zelanda.

2.       Resultados y conclusiones

La comparación entre experimentos con distintos forzamientos se ha basado en el modelo numérico WaveWatch III en su versión 5.16 utilizando los términos fuente ST4. La configuración de referencia es una malla global de 0.5 grados por 0.5 grados, forzada con vientos y cobertura de hielo de CFSR, sin icebergs ni corrientes. La mejora de mayor magnitud respecto a esta configuración se obtiene al sustituir CFSR por ERA5, lo que reduce notablemente el sesgo y el error cuadrático en la mayor parte del mundo. Estos resultados indican que las mejoras en resolución, asimilación de datos, o modelado de ciclones tropicales respecto al anterior reanálisis del ECMWF han conseguido que actualmente ERA5 sea la opción más adecuada para forzar modelos globales de oleaje. La inclusión de corrientes también produce mejoras a nivel global pero de una magnitud menor. En concreto las tres bases de datos de corrientes ayudan a reducir los sesgos, siendo GLORYS la que produce mejores resultados a pesar de tener menor resolución espacial que HYCOM y menor resolución temporal que CFSR. Es especialmente notable la reducción del sesgo positivo en el Océano Antártico, principalmente por el efecto de la corriente circumpolar antártica, que reduce la transferencia de energía del viento al oleaje. No obstante, incluso con la introducción de corrientes sigue existiendo un sesgo positivo. Este sesgo se reduce aún más con la inclusión de icebergs, que aumentan el bloqueo de energía, pero no llega a desaparecer por completo. La comparación entre el sesgo de la configuración de referencia y el de la configuración óptima para el año 2015 se muestra en la figura 1. La comparación de errores cuadráticos (no mostrado) indica mejoras globales en torno al 30%.

fig 1-top.png
Fig. 1. Sesgo respecto a datos de satélite de la simulación global forzada con vientos de ERA5, icebergs, y corrientes GLORYS (panel superior) y la simulación de referencia, forzada con vientos de CFSR y sin icebergs ni corrientes (panel inferior) para el año 2015.

Fig. 1. Sesgo respecto a datos de satélite de la simulación global forzada con vientos de ERA5, icebergs, y corrientes GLORYS (panel superior) y la simulación de referencia, forzada con vientos de CFSR y sin icebergs ni corrientes (panel inferior) para el año 2015.

Este análisis a escala global y la reconstrucción histórica resultante ha permitido minimizar los errores en los datos de oleaje en aguas profundas y disponer de mejores condiciones de contorno para las mallas de detalle en zonas costeras. Adicionalmente, lecciones aprendidas de este análisis, como la importancia de corrientes y icebergs, van a resultar en mejoras en los sistemas operacionales de predicción de oleaje de Metservice. Esto permite proporcionar a los usuarios datos históricos y predicciones de oleaje de la mayor calidad posible.

Agradecimientos

Se agradece el apoyo a este estudio tanto de la armada de Nueva Zelanda (NZ Navy) como de la oficina de investigación naval (Office of Naval Research, ONR) por medio de la subvención NOOO14-17-S-B001.

Referencias

DURRANT, T., GREENSLADE, D., HEMER, M. y TRENHAM, C. (2014). “A Global Hindcast focussed on the Central and South Pacific”. CAWCR Technical Report , 46.

PEREZ, J., MENENDEZ, M. y LOSADA, I. J. (2017). “GOW2: A global wave hindcast for coastal applications”. Coastal Engineering, 124 , 1-11.

MetOcean Solutions’ ocean forecasting system presentation at OceanPredict’19 Symposium in Canada

Next week, Dr João Marcos Souza and Prof Moninya Roughan will be at OceanPredict ’19 Symposium in Halifax, Canada.

Metocean Solutions’ physical oceanographer Dr João Marcos Souza will present “New Zealand’s national ocean forecast system - present and future”, showcasing MetOcean’s sophisticated operational ocean forecasting capability. Based on international best practices with the current state-of-the-art science, the system combines a number of different ocean models and data dissemination platforms. It is designed for rapid deployment of high-resolution model domains and portability between different platforms.

 
General architecture concept of MetOcean’s operational system.

General architecture concept of MetOcean’s operational system.

 

“We will present some of the advances we are making in ocean circulation modelling, an overview of MetOcean’s operational system and capabilities, together with our ongoing developments and future plans,” says João. “It is a great opportunity to present the latest advances in New Zealand’s operational oceanography and engage with best practices implemented around the world.”

At the conference, Metocean Solutions’ Head of Research Partnerships Prof Moninya Roughan will be presenting the Moana Project. The Moana Project, led by Prof. Roughan, is a cross-institutional programme involving all the oceanographic research organisations in New Zealand, in collaboration with international experts from Australia and the United States. The project will shed new light on the performance of New Zealand’s oceans to support the seafood sector.

The OceanPredict ’19 Symposium, hosted by GODAE OceanView, is being held 6-10 May at Halifax Convention Centre, Canada. The event brings together oceanographic science, research and end-user communities to increase awareness of current ocean modelling capabilities, and to explore and define the direction of future operational ocean forecasting.

MetOcean Solutions is a division of New Zealand’s National Meteorological Service.

For more information, visit the conference website: oceanpredict19.org

The full abstracts are provided below.


New Zealand ocean forecast system - present and future

Azevedo Correia de Souza, Joao*, Soutelino, Rafael*, Durrant, Tom*, Couto, Phellipe*

New Zealand’s maritime domain is one of the largest on the planet, with an exclusive economic zone of approximately 4,300,000 km2 – about 15 times its land area. The seafood sector alone brings $4.18B to NZ annually. Offshore oil and gas exploration provides about 30% of the country’s consumption, from 21 petroleum licenses in the Taranaki basin. Moreover, tourism is a growing industry accounting for about 5.9% of the GDP and often related to the country’s coastal landscapes. Therefore, having a reliable ocean forecast system is of critical importance to the country’s economy and to the safety and resilience of the community and environment. This includes the capability to model and forecast ocean processes at a range of spatial and temporal scales. To accomplish this, a sophisticated system including different ocean models and data dissemination platforms has been developed. The system is designed for rapid deployment of high-resolution model domains, kept up to date with state-of-the-art techniques, and portability between different platforms. At the present, this system is mainly based on downscaling of global models (except for ocean waves) and a series of local nested model grids. A mix of “Regional Ocean Modeling System” (ROMS) and “Semi-implicit Cross-scale Hydroscience Integrated System Model” (SCHISM) domains are used to evaluate and predict ocean circulation and state properties, while “WAVEWATCH III” (WW3) and “Simulating Waves Nearshore” (SWAN) are used for simulating surface gravity waves down to harbour scales. A micro-service architecture based on docker and controlled by a built-for-purpose distributed workflow scheduler ensures a stable, highly-available system. New developments underway include the use of un-structured model grids, 4DVar data assimilation of global and local observations on a national scale, waves-circulation coupling, and the use of cloud-based computational resources. Focusing mainly on the ocean circulation modelling, a general description of the system and capabilities at Metocean are presented together with ongoing developments and future plans.

*MetOcean Solutions, division of Meteorological Service of New Zealand


The Moana Project: Seafood sector support for ocean data collection to improve ocean prediction in New Zealand

Roughan, Moninya*

New Zealand derives wealth and wellbeing from the ocean, including a seafood sector worth $4.18B annually, and yet, their oceans are very poorly understood. NZ lags other developed nations that have integrated ocean observing and modelling programmes, and cannot comprehensively measure, observe or predict the state of their Exclusive Economic Zone (EEZ).  Ocean circulation drives the transport of larvae, determines population connectivity and impacts fisheries recruitment and abundance, all of which are being impacted by ocean warming and changes in circulation patterns.

Embracing ‘the Internet of Things’ concepts, we are developing a low-cost smart ocean sensor to be deployed throughout NZ’s EEZ by the seafood sector. With our industry partners; Seafood NZ, Deepwater Group, Paua (Abalone) and Rock Lobster Industry Councils, iwi (indigenous) and recreational fishing communities, we will revolutionise ocean data collection. The temperature profile data will be returned in near real time via the cell phone network (or satellite) and ingested into data assimilating ocean prediction models, leading to an open-access nationwide Ocean Analysis and Prediction System, delivered by the Meteorological Service. This disruptive technology approach is an exemplar for other marine nations with strong seafood sectors and under investment in the marine observing and modelling space. We show the benefit of partnering with end users to collect and return research quality datasets that are relevant for industry needs.

This project will provide a more complete picture of ocean temperatures, circulation and dynamics, and the relationships with fishery recruitment variability, aiding prediction. This project will underpin operational efficiencies, biosecurity protection, risk mitigation and economic growth for NZ’s seafood sector ensuring long-term sustainability.

*MetOcean Solutions, division of Meteorological Service of New Zealand



Wave forecast model upgrades

MetOcean Solutions has released an upgrade in all regional and local scale operational wave models. This upgrade brings improvements to model skill throughout the forecast horizon.

Wave models are used to simulate the physical processes occurring in wave growth, wave breaking and wave propagation. The processes involved in describing the input from the wind and the dissipation from wave breaking are collectively referred to as the model source terms.

“New source terms (called ST6) have recently been implemented in the official release of the Simulating WAves Nearshore (SWAN) spectral wave model,” says MetOcean Solutions’ Senior Physical Oceanographer Dr Rafael Guedes. “These terms are based on field observations and incorporate some important new physical features, including airflow separation under strong wind forcing, swell dissipation and a better description of breaking dissipation.

“Our operational services rely extensively on SWAN. The new source terms represent a great improvement in the physical representation of wave generation and dissipation within our regional and local scale wave models.

“We have carefully calibrated and validated all our regional operational domains with the new physics,” continues Rafael. “Comparison against satellite altimeters and in-situ wave observations showed consistent improvements in our models across all major areas.”

Figure 1 and Figure 2 below show percentage changes in Root-Mean-Square-Deviation (RMSD) and scatter index (SI), two commonly-used measurements of wave model skills. Overall improvements are apparent as highlighted by the blue colours, with up to 30% decrease in RMSD and 11% decrease in SI with ST6 in some of these areas. The improvements are shown in more detail for MetOcean’s 5-km SWAN grid in Australia Northwest Shelf in Figures 3 and 4.

 
Figure 1. Percentage changes in Root-Mean-Square-Deviation (RMSD) between operational SWAN domains run with old and new SWAN physics. Blue and red indicate reduction and increase in RMSD respectively.

Figure 1. Percentage changes in Root-Mean-Square-Deviation (RMSD) between operational SWAN domains run with old and new SWAN physics. Blue and red indicate reduction and increase in RMSD respectively.

 
 
Figure 2. Percentage changes in Scatter Index (SI) between operational SWAN domains run with old and new SWAN physics. Blue and red indicate reduction and increase in SI respectively.

Figure 2. Percentage changes in Scatter Index (SI) between operational SWAN domains run with old and new SWAN physics. Blue and red indicate reduction and increase in SI respectively.

 
 
Figure 3. Validation against satellite altimeters of MetOcean Solutions’ 5km Australia Northwest Shelf SWAN domain using the old physics source terms. Overall scatter diagram and scatter density are shown at the top. Model bias and RMSD are presented at the bottom.

Figure 3. Validation against satellite altimeters of MetOcean Solutions’ 5km Australia Northwest Shelf SWAN domain using the old physics source terms. Overall scatter diagram and scatter density are shown at the top. Model bias and RMSD are presented at the bottom.

 
 
Figure 4. Validation against satellite altimeters of MetOcean Solutions’ 5km Australia Northwest Shelf SWAN domain using the new ST6 physics source terms. Overall scatter diagram and scatter density are shown at the top. Model bias and RMSD are presented at the bottom.

Figure 4. Validation against satellite altimeters of MetOcean Solutions’ 5km Australia Northwest Shelf SWAN domain using the new ST6 physics source terms. Overall scatter diagram and scatter density are shown at the top. Model bias and RMSD are presented at the bottom.

 

At MetOcean Solutions, a division of MetService, we continuously improve our models with the current state-of-the-art science to ensure the highest possible performance.



An operational high resolution hydrodynamic forecast model for Port Phillip Bay

MetOcean Solutions has recently operationalised a high resolution hydrodynamic forecast model that allows simultaneous simulation of waves, currents and their interaction for Port Phillip Bay, Australia.

This new capability was developed as part of a larger project funded by the Australian Cooperative Research Centres Projects initiative. The project, a partnership between OMC International, Pivot Maritime International, University of Melbourne and MetOcean Solutions will provide an integrated modelling system for predicting under-keel clearance to support port and shipping services in tidal inlets.

“This project was conceived through an industry-research partnership and has leveraged the technical expertise in all partners,” says MetOcean Solutions’ Development Manager Dr Tom Durrant. “By bringing cross-sector experience, it is possible to develop science value-added solutions designed to increase safety and assist informed decisions at sea.”

In this particular project, the circulation model (SCHISM) is coupled with an upgraded wave physics implemented in the Wind Wave Model (WWM-III). The ‘improved WWM-III’ combined with the state-of-the-art unstructured hydrodynamic model SCHISM develops a forecast model capable of simultaneous simulation of waves, currents and their interaction.

“This is of great importance in this part of the world,” says MetOcean Solutions’ physical oceanographer Phellipe Couto.

“Port Phillip Heads, connecting Port Phillip Bay and Bass Strait is a notorious stretch of water that has claimed many ships and lives. Strong tidal currents interacting with waves combine to create significant challenges to ship navigation. Explicit accounting of this interaction, combined with unstructured model grids allowing the complex features of the main channels to be resolved at much higher resolution than previously, offer significant improvements in our ability to accurately forecast both waves and currents in the heads.

“Working closely with the University of Melbourne has provided a great opportunity to rapidly transition cutting edge science into operational systems.”

The operational high resolution hydrodynamic forecast model developed will provide input into the under keel clearance system operated by OMC international, strengthening the offerings available through the Metocean Solutions and OMC partnership (find more information here).

Operational SCHISM is MetOcean Solutions’ powerful new capability in high resolution coastal hydrodynamics, improving forecast by well representing complex nearshore bathymetries. The forecast model was also operationalised for Tasman and Golden Bay, New Zealand. Click here for more information.

The SCHISM model for Port Phillip Bay is freely available at MetOceanView.

For more information visit www.metoceanview.com or contact us at enquiries@metocean.co.nz

An operational hydrodynamic forecast model for Tasman and Golden Bay

MetOcean Solutions has recently operationalised a high resolution hydrodynamic model for Tasman and Golden Bay, New Zealand.

The underlying forecast data is produced by a state-of-the-art unstructured hydrodynamic model (SCHISM), with offshore 3D boundary conditions sourced from a 3-km ROMS implementation of the central NZ region.  This new capability was developed as part of the Sustainable Seas Project together with the  Cawthron Institute and NIWA and will provide valuable information necessary to manage contamination risk in the aquaculture industry and beach water quality forecasts relevant to regional councils and recreational beach users.

General Manager MetOcean Solutions Dr Brett Beamsley says MetOcean Solutions’ science team has many years of experience with the SCHISM model (previously SELFE); applied primarily in high value consultancy services or research projects, with the unstructured domain capability key to representing complex nearshore bathymetries in a computationally efficient manner.

“This particular project has leveraged the strong scientific capabilities in all three research partners (NIWA, Cawthron and MetOcean Solutions) and illustrates what can be achieved when working together collaboratively.”

"SCHISM is a valuable addition to our operational hydrodynamic forecast system,” says MetOcean Solutions’ physical oceanographer Phellipe Couto. “It allows our model applications to account for an even better representation of topographic features (e.g. islands, embayments, navigation channels and tidal inlets) and engineering structures (e.g. ports and breakwaters) that pose critical aspects in the modulation of the hydrodynamic regime surrounding nearshore and coastal waters.

“In practical terms, this enable us to resolve multi-scale geophysical processes such as tides, river plume dispersion and storm surge with an extra degree of accuracy and therefore provide better forecast solutions to the end user.

“The impact of storm surges on coastal areas has become highly topical particularly in the last year and the rapid deployment of this type of operational modelling infrastructure has the potential to more accurately predict coastal nearshore water levels.

SCHISM model grid resolution from approximately 10 m nearshore to 1.5 km offshore.

SCHISM model grid resolution from approximately 10 m nearshore to 1.5 km offshore.

“In this particular project, we developed a model grid with resolution varying from 10 m in the nearshore to approximately 1.5 km offshore, defining estuaries, intertidal areas, channels, streams, major rivers and relevant beaches. The model is a full 3-dimensional implementation with atmospheric and oceanic initial and boundary conditions provided by high resolution in-house models developed for the Central New Zealand oceanic domain encompassing North and South Islands’ coastal areas around the Cook Strait. We also included fluvial discharges from 11 different rivers forecasted by NIWA’s hydrological modelling capability (TOPNET) as an important forcing to our model.”

“SCHISM presents a powerful new capability for Metocean Solutions in high resolution operational coastal hydrodynamics,” says MetOcean Solutions’ Development Manager Dr Tom Durrant. “This is the first of several planned implementations.”

The project ‘Near real-time forecasting using operational oceanographic forecasting of contamination risk to reduce commercial shellfish harvest and beach closures’ is a collaborative effort of experts from the Cawthron Institute, NIWA and MetOcean Solutions. A project to build connected land-river-sea models and provide a timely risk assessment of contamination to beaches and shellfish growing areas. For more information on Sustainable Seas National Science Challenge click here.

The SCHISM model for Tasman and Golden Bay is freely available at MetOceanView.

For more information visit www.metoceanview.com or contact us at enquiries@metocean.co.nz


Forecasting Gita - extreme storm surge and wave heights

When tropical storm Gita passed over New Zealand on 20-21 February this year, high winds and low pressures combined with energetic ocean swells caused significant storm surges along the New Zealand West Coast.  

Storm surge is the abnormal rise of water generated by a storm, over and above the predicted astronomical tides. When accompanied by high waves, this surge can cause significant damage to coastal areas, including flooding and accelerated erosion.

“The storm offered an opportunity to validate our inhouse wave and surge models,” explains Dr Séverin Thiébaut, senior oceanographer at MetOcean Solutions. “Validation is when we compare the model results directly with in situ measurements, and provides us with a clear indication of how good our models are at predicting the timing and the magnitude of extreme storms. For Gita, we used data from a tide gauge at Charleston on the West Coast and a wave buoy in offshore south Taranaki.”

The forecast model predictions of storm surge and wave height during the storm are shown in Figures 1 and 2, respectively. Modelled wind and rainfall are also presented in these figures. These operational models are produced by MetOcean Solutions for a range of applications in NZ waters. The models are tuned to replicate the typical conditions, so verifying the predictions under an extreme storm is a powerful test.

 
Figure 1: Predicted storm surge progress: wind and rain (left) and storm surge (right) coincide as tropical storm Gita passes over New Zealand.

Figure 1: Predicted storm surge progress: wind and rain (left) and storm surge (right) coincide as tropical storm Gita passes over New Zealand.

 
 
Figure 2: Predicted wave progress: wind and rain (left) and wave height (right) forecast as tropical storm Gita passes over New Zealand.

Figure 2: Predicted wave progress: wind and rain (left) and wave height (right) forecast as tropical storm Gita passes over New Zealand.

 

In Figures 3 and 4 we present the time series comparison of measured and forecast storm surge and wave heights.

“The test shows our model slightly underestimated the storm surge as measured by the nearshore tide gauge,” continues Séverin. “This is likely due to the geometry of the small bay where the tide gauge is located plus the effects of wave setup on the measured water levels. Near the shore, waves will produce a localised increase in water level, which magnifies the observed storm surge. However, we are delighted the timing of the predictions was good and the magnitude acceptable for an open-coast extreme.”

“The Charleston tide gauge was located on the margin of the main storm surge effect; the model predicted coastal water level elevations up to 0.50 m in the northern parts of the South Island but unfortunately no open coast tide gauges are located in that area.”

Figure 3: Time series of the storm surge as measured by the Charleston tide gauge data (Source:  LINZ ) and the MetOcean Solutions forecast values. The tide gauge reads higher values due to wave setup and the close proximity to shore.

Figure 3: Time series of the storm surge as measured by the Charleston tide gauge data (Source: LINZ) and the MetOcean Solutions forecast values. The tide gauge reads higher values due to wave setup and the close proximity to shore.

“The comparison for offshore wave height was excellent for a such rapidly moving system. At the offshore wave buoy, the forecast timing and height of the waves were very well correlated with the measured storm values. The highest measured significant wave height was 8.8 m, while the forecast value was 8.4 m. “It is very encouraging to see the models perform with such confidence under these extreme storm conditions,” notes Séverin. “For context, the largest significant wave height ever recorded on the West Coast of NZ was 10.4 m, measured in May 1977 near the Maui A platform.”

Figure 4: Time series of the measured and modelled significant wave height at the offshore wave buoy (Source:  OMV ).

Figure 4: Time series of the measured and modelled significant wave height at the offshore wave buoy (Source: OMV).

Through our web portal MetOceanView, MetOcean Solutions provide automated storm watch services for coastal and offshore operators. These are automatically generated when preset conditions are identified within the forecast horizon. During the lead up to Gita’s landfall, significant wave heights of 8-9 m were predicted more than 3 days ahead (Figure 5). Warnings are sent via email alerts for operational decision-making - increasing safety and efficiency for people working at sea. These warnings are based on an ‘ensemble’ of  forecasts of the winds and waves, which captures the inherent uncertainty over a 7-day forecast horizon and the chaotic nature of a revolving tropical storm. Probabilistic guidance allows effective management decisions to be made within a robust, quantitative framework.

 
Figure 5: Gita was forecast to be wild! Example of the ensemble forecast warning system that provided guidance on the range of possible outcomes for wind speed (top), significant wave height (middle) and wave direction (bottom) in the offshore Taranaki area.

Figure 5: Gita was forecast to be wild! Example of the ensemble forecast warning system that provided guidance on the range of possible outcomes for wind speed (top), significant wave height (middle) and wave direction (bottom) in the offshore Taranaki area.

 

MetOcean Solutions is a wholly owned subsidiary of state-owned enterprise, Meteorological Service of New Zealand (MetService). MetService is New Zealand’s national weather authority, providing comprehensive weather information services, to help protect the safety and well-being of New Zealanders and the economy.

For more information on our forecast models, contact us at enquiries@metocean.co.nz.

Smart Ideas funding for MetOcean Solutions

MetOcean Solutions has received research funding for a project applying innovative technology to weather forecasting. The research, entitled ‘Machine learning for convective weather analysis and forecasting’, was funded in the Ministry of Business, Innovation and Employment (MBIE) 2017 Endeavour Round.

Led by MetOcean Solutions, the project is a collaboration with the Knowledge, Engineering & Discovery Research Institute (KEDRI) of Auckland University of Technology, and the New Zealand MetService.

“We are very pleased to receive funding for this exciting project,” states project lead Dr David Johnson.

The research will use machine learning to predict convective weather events. Convective weather produces heavy rain, lightning and strong winds, significantly impacting the safety, efficiency and well-being of New Zealanders, and recent severe events have caused loss of life and extensive damage to property.

“Convective weather is localised in time and space and can develop quite quickly and sometimes without much warning,” continues Dr Johnson. “These events are always associated with cloud masses - like big thunderstorm clouds - which can be seen from Earth-observing satellites. Currently, human forecasters detect these events by looking at satellite imagery, numerical model guidance, and rain radar. Trained forecasters are good at doing this, but humans cannot look everywhere all the time. Recent advances in machine learning mean that computers are now exceptionally good at identifying and labelling features in images. Our research will train a machine-learning algorithm to analyse satellite imagery, possibly combined with other inputs such as numerical model guidance or rain radar, and predict where and when heavy rain, lightning or wind squalls might occur. As this is a machine process it can potentially be fully automated and then used to send alerts on a phone app.

“There is a good chance that the algorithm will be better than a human at consistently making the correct predictions, which means that all New Zealanders will benefit from better weather forecasts. An automated forecasting system also allows for greater customisation and localisation for individual needs. If successful, the research could lead to apps that makes it possible for you to request a phone alert if heavy rain is likely at your location - your phone already knows where you are - allowing you to take in your washing, seek shelter or postpone your drive home. Many industries are weather-dependent, and accurate local forecasts of strong winds or heavy rainfall will help anyone working or organising events outdoors, including the forestry, fisheries, construction and transport industries, all of whom have different weather thresholds to ensure safety and efficiency. Human forecasters could never manage to serve all the myriad of end-user needs at different locations and times.

“The key to success is the collaboration with KEDRI, who are world-leading machine learning experts, and MetService who carry out day-to-day severe weather forecasts for New Zealand. MetOcean Solutions brings our track-record of applying state of the art science and technology to provide end-user tools and services.”

MBIE's 2017 Endeavour Round funded 68 new scientific research projects from 408 applications.

Click here for the MBIE press release on the funding.

logos.png

A real-time data solution supports drilling operations in Wellington Harbour

Griffiths McMillan JV (GMJV) is tasked by Wellington Water and the Greater Wellington Regional Council with drilling for a new water source for Wellington City, with the goal of providing an alternative supply in the event of a severe earthquake. GMJV approached Heron Construction Company to supply the jackup barge and support tug for the project. The target aquifer, however, lies beneath Wellington Harbour - just north of the Miramar Peninsula. A marine drilling rig is required for the operations, which means that operations are very sensitive to the wave conditions.

A marine drilling rig.

A marine drilling rig.

MetOcean Solutions has been supporting Heron Construction in their specialist marine activities with high resolution forecasting for the last decade. Managing director Greg Kroef says, “We rely on the forecast guidance from MetOcean Solutions wherever we are working in Australia and New Zealand. So, when the need arose in Wellington, we approached them for the best possible weather forecasting system for this site.”

The drilling job in the harbour requires very accurate predictions of wave height. However, the oceanographic conditions in Wellington Harbour present significant challenges for forecasting, with ocean swells entering from the south along with local seas generated by the infamous capital city winds.
 
Oceanographer Dr Peter McComb and his science team extensively studied the harbour dynamics for the recent harbour deepening project. “The drilling location is affected by both the southerly ocean swells and the short local wind seas,” explains Peter. “Our forecasting model has to deal with multiple and sometimes simultaneous sources of wave energy, plus discern how those waves refract and transform due to the shape of the seabed and the harbour tidal hydrodynamics.”

The wave buoy, framed by Wellington City

The wave buoy, framed by Wellington City

The operational SWAN wave model for Wellington Harbour has a spatial resolution of 80 m, and is one of nearly 200 forecasting domains that MetOcean Solutions run four times per day for various parts of the planet. This high-resolution model is coupled with the tidal hydrodynamics to capture the effects of the ebb and flood flows through the entrance on the waves, and it has spectral boundaries prescribed by a 3-stage nest into our global WAVE WATCH III wave forecast model. The model system produces the hour-by-hour wave conditions for the coming 7 days.  

Forecast model result from MetOceanView showing the 13 July 2017 southerly storm waves penetrating Wellington Harbour. The X denotes the wave buoy location.

Forecast model result from MetOceanView showing the 13 July 2017 southerly storm waves penetrating Wellington Harbour. The X denotes the wave buoy location.

“However, just having a wave forecast model is sometimes not enough to ensure safe operations - especially in Wellington which is notorious for rapid changes in weather conditions,” continues Dr McComb. “So we decided to deploy one of our directional wave buoys near the drilling site to provide real-time monitoring of the wave conditions along with instant verification of the forecast accuracy.”

The wave buoy sends data ashore every hour so everyone involved can monitor the sea conditions in real time. See for yourself at:  http://wavebuoy.metocean.co.nz/wellington 

“When ingested into the MetOceanView forecasting system, we co-plot the measured and the forecast wave heights. This is the most honest representation of forecast accuracy, and allows users to gain confidence in the timing and the magnitude of the wave predictions. The recent storm from 13 July was a very energetic event, but we captured it perfectly from 3-4 days ahead.”   

The real-time wave buoy data shows the waves resulting from the storm in mid-July.

The real-time wave buoy data shows the waves resulting from the storm in mid-July.

Accurate wave forecast – the significant wave heights measured by the buoy were very close to those forecasted for the storm in mid-July.

Accurate wave forecast – the significant wave heights measured by the buoy were very close to those forecasted for the storm in mid-July.

MetOceanView ingests observations from over 10,000 sites every day, including 252 locations around New Zealand. 

For queries about ingesting site-specific data, contact enquiries@metocean.co.nz.

Storm watch helps safe management

Forecasts of wind and wave conditions are vital for effective and safe management of near- and offshore operations. MetOcean Solutions has developed a warning system based on ensemble forecasting, which allows us to provide warning of upcoming severe wind and wave conditions for up to 7-day horizons, thereby helping operators plan for safe and efficient management.

Accurate high-resolution forecasts of wave conditions at local scales can be done using nested or grid-refining models such as SWAN (Simulating WAves Nearshore). Such forecasts capture local wave transformation and dissipation, and can be done at a reasonable computational cost. However, at horizons beyond 2-3 days a forecast inevitably suffers from the onset of chaotic uncertainty that is a physical characteristic of the atmosphere-ocean system. 

Ensemble forecasting is a method that quantifies this uncertainty in longer-range forecasts, providing probabilistic guidance for management decisions. MetOcean Solutions can customise a forecast warning system for any offshore or nearshore location based on the 7-day forecast wind speed, significant wave height (Hs) and maximum individual wave height (Hmax). The method used to estimate extreme Hmax accounts for the long-term uncertainty of the severity of the environment and the short-term uncertainty of the severity of the maximum wave of a given sea state, complying with the International Organization for Standardization (ISO standards).

Example of Storm Watch programme. T indicates present time.

Example of Storm Watch programme. T indicates present time.

We use such ensemble forecasts to provide the best possible site-specific warning to clients of upcoming severe conditions, using pre-set thresholds to define the level of alert. 

Up to 60 wind and wave ensemble forecasts can be used to support a storm watch programme.  

An example of email alert is illustrated below. All components of the email can be customised upon request.
 

Example of Storm Watch alert email. 

Example of Storm Watch alert email. 

Storm watch guidance. Orange shading indicates when the alerting thresholds are exceeded. Wsp: windspeed. Hmax: maximum wave height. Hs: significant wave height. 90 PCTL: 90th percentile. 

Storm watch guidance. Orange shading indicates when the alerting thresholds are exceeded. Wsp: windspeed. Hmax: maximum wave height. Hs: significant wave height. 90 PCTL: 90th percentile. 

Wave height warnings. Red shaded area represents the incoming wave direction for which thresholds are exceeded.

Wave height warnings. Red shaded area represents the incoming wave direction for which thresholds are exceeded.

MetOcean Solutions can set up a storm watch programme for any location. For more information, or to discuss your site, contact us at enquiries@metocean.co.nz.

Geraldton Port - improving long wave predictions with real-time wave buoy data

“MetOcean Solutions has worked exceptionally hard to assist the Port of Geraldton with its surge problem. The product they now deliver is heavily relied on when scheduling shipping. The efficiencies gained in reduced labour costs to port users are significant, but is the contribution to safety that represents the greatest gain. In 2005 there were 242 parted ships lines at Geraldton. Following the developments in forecasting Long Wave effects at Geraldton, last year there were only 30 parted ships lines. This was the lowest number of parted lines since records have been kept. MetOcean Solutions is keeping Geraldton Port efficient, and port workers safe.”
Captain Ross Halsall, Acting Harbour Master, Port of Geraldton.

Observational wave data can help ports manage the onset of long wave events. In recent work for Geraldton Port (Australia), we improved our harbour long wave forecasts by integrating wave buoy data from two upstream locations into the short-range predictions.

Long (or infragravity) waves are created when swell waves interact with the coast and result in water level oscillations with periods much longer than the original swell waves. Long wave periods of 60-120 seconds are typical, and these are very problematic because their wavelength is similar to the size of ships. When such waves enter a harbour the moored vessels are energised, causing dangerous surging which can break lines and put personnel in danger. As a result, ports often have to close when long wave heights exceed a safe threshold.
 
Accurate forecasts of long wave height can help ports increase safety, reduce unwarranted closures and effectively plan for reopening. For the past 11 years, MetOcean Solutions has provided a specialist long wave forecasting service that ports and harbours throughout New Zealand and Australia have come to rely upon.   
 

Numerical simulation showing that the sea-swell climate in the vicinity of the Port of Geraldton is extremely complex, mainly due to the reef near the Port entrance.

Numerical simulation showing that the sea-swell climate in the vicinity of the Port of Geraldton is extremely complex, mainly due to the reef near the Port entrance.

“Our scientific research over many years provides the basis for a robust prediction system,” explains Senior Oceanographer Dr Severin Thiebaut. “We use a semi-empirical technique to establish the relationship between the offshore wave spectra and the long wave height at each berth in a port. It's a method that we have tested at more than 30 locations worldwide”. 
 
The forecast wave spectra is derived from the suite of global and regional numerical models run by MetOcean Solutions. Geraldton Port have been using these predictions to guide the harbour operations since 2007. Predictions have proved very reliable, but sometimes the arrival of a swell front may differ from the forecast by a few hours as the exact timing is difficult to resolve perfectly with a spectral wave model. To supplement the long wave forecasts, we developed a new technique using real-time wave buoy data to improve the short term predictions and better detect occasions with a sharp rise in long wave energy inside the harbour. 
 
‘“The goal is to ensure that our port clients are never surprised by weather events,” says Dr Thiebaut.  

The Department of Transport maintains wave buoys at Rottnest Island and at Jurien Bay, some 370 km and 170 km to the south of Geraldton, respectively. The live data from these buoys, updated at 30 minute intervals, are used to track the progression of swell up the west coast of Australia and provide harbour long wave predictions some 5-6 hours ahead.

Data from wave buoys at Rottnest Island and Jurien Bay help improve long wave predictions for Geraldton Port.

Data from wave buoys at Rottnest Island and Jurien Bay help improve long wave predictions for Geraldton Port.

“Using the live buoy data provides additional confidence to the standard forecasts. In particular, it allows us to accurately capture any rapidly rising long wave events. It also gives the port operators a better idea of when wave heights will decrease to safe working levels. Geraldton Port accesses the forecast information through the web-based MetOceanView platform and through their local environmental monitoring software. On the same plot we show the measured long waves at the berth along with the values that are forecast by our standard system plus those predicted by the buoys.” 
 

Comparison between measured and predicted significant long wave heights (Hslpw) in the Port of Geraldton in 2012 based on Jurien Bay real-time wave buoy data.

Comparison between measured and predicted significant long wave heights (Hslpw) in the Port of Geraldton in 2012 based on Jurien Bay real-time wave buoy data.

Contact us for a discussion of what we can do to help your port. Call +64 6 758 5035, email enquiries@metocean.co.nz or visit www.metoceanview.com.
 

SurfZoneView part of Defence Force training exercise

MetOcean Solutions joined the Defence Technology Agency (DTA) in a training exercise in Army Bay, Whangaparoa during May. As part of a larger trial involving the New Zealand Navy and Air Force, DTA were testing a wave buoy and integrating the data into the SurfZoneView beach landing software.

The field phase of Exercise Joint Waka was held at Army Bay and the inner Hauraki Gulf from 14-19 May 2017. It involved New Zealand Army vehicles, amphibious sealift vessel HMNZS CANTERBURY, and Royal New Zealand Air Force NH90 medium utility helicopters. The exercise sought to enhance the New Zealand Defence Force’s ability to deploy offshore to deal with any contingencies including humanitarian crisis, natural disasters and instability within our region.

Figure 1: HMNZS Canterbury

Figure 2: Amphibious vessel used for the landing exercise.

Developed in a collaboration between DTA and MetOcean Solutions, SurfZoneView allows the visualisation of beach landing conditions. Moving people and equipment from sea to land is one of the most complex tasks completed by the New Zealand Defence Force. Such operations are necessary when port facilities are not available, for example when providing natural disaster relief in New Zealand and the Pacific. The safety and success of shore landings are largely dependent on surf zone conditions, and SurfZoneView uses hydrodynamic modelling to provide a rapid and accurate assessment of the waves and currents at any location. Clear, easy-to-use maps of the nearshore conditions are displayed along with tools to assist with operational decision making.

Figure 3: Dr Jamie Halla (left) and Theo Zlatanov (right) from DTA and Dr Rafael Guedes from MetOcean Solutions (centre) getting ready for the trial on the Canterbury ship.

Senior Oceanographer Dr Rafael Guedes was involved with the trial. “It is very hard to assess the beach conditions when you are out at sea. Wave breaking patterns and surf zone currents can change drastically over only a few tens of metres as they are influenced by local bathymetry. By the time a vessel is close enough to the shore to allow operators to judge conditions, it is often already impacted by the waves. SurfZoneView allows operators to visualise the conditions over a stretch of coastline, to help them decide where the best landing place is that day, or sometime in the near future.

Figure 4: It is very hard to assess nearshore conditions from sea.

“The surf zone conditions vary from day to day depending on wind direction, swell characteristics and general circulation. To model the surf zone, the software needs input data describing offshore conditions. These data can either come from forecasts or from real-time measurements such as those from a wave buoy.

Figure 5: Screenshot from MetOceanView showing the forecast site at the location where the New Zealand Defence Force wavebuoy was deployed.

“In this training exercise, we used both. DTA deployed a wave buoy to provide real-time conditions, and we also set up a high resolution forecast site for the location. We can set up a forecast site within an hour, rapidly making available reliable wind, wave and current predictions. The buoy data provides an accurate description of the local conditions, and the forecast allows us to predict how these conditions will change over time.

“During the training exercise, a local storm developed. Waves near the landing site rose to over 1.5 m in just a few hours, with wave periods progressively increasing and directions shifting from north-east to northerly. The development was tracked on SurfZoneView, allowing us to predict conditions around the shore landing site, and how these would change over time as the storm progressed.”

See the example maps of the maximum wave height predicted by SurfZoneView in Figure 7, during the events marked by the black vertical lines in Figure 6. Information from the model is automatically processed to define safety thresholds including whether it is safe to attempt landing on the shore. These thresholds can then form part of the information assessed by operational staff when making a go / no-go decision.

Figure 6: Forecast provided by MetOcean Solutions during the period of the exercise, showing significant wave height  (Hs) and peak wave period  (Tp). Black vertical lines show the events along the development of the storm chosen as input conditions for SurfZoneView. Significant wave height, Hs, is the average height (in metres) of the largest one-third waves. It approximately corresponds to the height of waves as estimated by a trained observer at sea. Peak wave period (Tp) is the wave period (in seconds) of the most energetic waves in a sea state.

Figure 7: As the storm progresses, landing conditions worsen. Example output for the two events highlighted in Figure 6, showing maximum wave height increasing around the landing location. Coloured line along the shoreline displays safety thresholds for the landing vessel. 

“Models are most accurate when we use the best possible input data," adds Rafael. "Wave buoys provide site-specific, accurate data very quickly. Using such buoys alongside SurfZoneView allows the New Zealand Defence Force to go to any location, deploy a buoy and within less than an hour access accurate data to help them land personnel and supplies safely. To aid operation planning, MetOcean Solutions can set up a site-specific forecast for anywhere in the world. This means that we can quickly generate a forecast for wherever SurfZoneView needs to be used, thereby providing the best possible support for the Defence Force and other users.

“We are grateful to be invited to participate in this training exercise. Testing it under real conditions provides important information on where the tool adds value operationally. Coincidentally, we also assisted the Italian Littoral Warfare Unit with similar tests in Sardinia during May, and gained valuable feedback. ”

Figure 8: Based on chart depth, wave height, tide level and wave setup, SurfZoneView displays safe water levels for vessels approaching the shore. Left: map displaying safety thresholds. Right transect (indicated on map) profiles for (top): maximum wave height, including cross-hatched area for breaking waves; and (bottom): water depth and corresponding safety thresholds.

SurfZoneView was developed in close collaboration with DTA, and several overseas navies have indicated interest in purchasing the software.

Sea surges as ex-cyclone Debbie moves through

Heavy rains and surging seas plagued New Zealand as the remnants of cyclone Debbie moved  across the country this week.

“Storm surges occurred in different places across New Zealand as the storm moved through,” explains Senior Oceanographer Dr Rafael Soutelino. “We were lucky that the system moved through fast and did not coincide with large tides, otherwise the coastal flooding could have been a lot worse.”

Storm surges occur when the sea rises as a result of wind and atmospheric pressure changes associated with a storm. The surges build up over time and will worsen when a low pressure system lingers. 

“MetOcean Solutions forecast storm surges nationwide using a complex high definition 3D hydrodynamical model,” continues Rafael. “The model computes the atmospheric effects on coastal water levels. It combines this with baseline water levels generated by open ocean eddies and water column expansions and contractions caused by the spatially-variable vertical density distribution.

“Our modelling shows that as the storm progressed, it caused storm surges in different parts of the country. Although mostly mild, the storm surges were still high enough to cause coastal flooding in sensitive areas. 

The storm progresses: wind and rain (right) and storm surges (left) coincide as ex-cyclone Debbie passes over New Zealand.  

The storm progresses: wind and rain (right) and storm surges (left) coincide as ex-cyclone Debbie passes over New Zealand.
 

Storm surges up to 35 cm high occurred in a number of coastal locations around New Zealand.  

Storm surges up to 35 cm high occurred in a number of coastal locations around New Zealand.
 

“We were very lucky that the storm moved through so fast, and that it coincided with neap tides. Had it happened at spring tides, when the high tide levels are higher, flooding in the area could have been much worse. Our models show that at spring tide, the storm surge would have been high enough to almost match the measured river level after all the rain. In that scenario, the river would not have drained into the ocean as fast, which means the flooding would have been worse and lasted longer.”

Coastal flooding could have been much worse had the storm surge coincided with spring high tide at the Whanganui River mouth.

Coastal flooding could have been much worse had the storm surge coincided with spring high tide at the Whanganui River mouth.

Storm surge forecasts provide valuable information for for low-lying coastal locations, and these solutions are readily available to public authorities. 

“MetOcean Solutions’ New Zealand storm surge model has a resolution of 5 km. It’s the only operational hydrodynamical model for the country and we’ve been running it for more than five years now. During that time we have used it for oil spill and search and rescue modelling, and during the Rena disaster is provided essential guidance for the national response activities. We developed a similar model for the south coast of Australia, which has been operational for more than one year. This model is used by our Alliance partners OMC International in their specialist under keel clearance applications. 

“We run these types of models all over the world,” says Rafael - even in his home country of Brazil where the complex flows along the continental shelf are important to the offshore oil field operations.   

For further information about storm surge warnings and the New Zealand 3D hydrodynamical model, please contact us at enquiries@metocean.co.nz
 

Two million data points a day, and counting

Every day, the MetOceanView service ingests and serves up to our clients more than 2 million unique data points. These are modelled and observed data providing vital marine weather information to users.

The MetOceanView platform displays forecast and historical data for a range of locations. Clients worldwide use the site to access the results from customised wave and hydrodynamic models, helping them make important decisions to maximise safety and improve efficiency.   

“MetOceanView provides a phenomenal amount of information for a wide range of clients,” explains Andre Lobato, who works on data management for MetOceanView. “In order to run such a system, the platform has to process an enormous amount of data.”

Model and observational data are ingested into MetOceanView to provide a complete picture of ocean weather conditions for our clients.

Model and observational data are ingested into MetOceanView to provide a complete picture of ocean weather conditions for our clients.

“Every day, we ingest about 2.25 million discrete data points. More than 2 million of these are unique rows of modelled data from global weather and marine models. In addition to modelled data, we continuously incorporate satellite, lightning, weather station, wave buoy, current meter and tide gauge data as part of the operational infrastructure behind MetOceanView. Some of these data, like METAR stations, NOAA-NDBC buoys, NOAA-MADIS, Himawari 8, GOES and MODIS satellite images are displayed directly on the MetOceanView interface. Others are shown to provide comparisons with our modelled data - e.g. wave buoy data displayed on a graph comparing observed to forecasted wave height.

“Real-time lightning data at times add a huge number of additional observations. Provided through Blitzen (TOA and GPATS), each single lightning strike constitutes a discrete observation. This means that on some days we incorporate millions of lightning data points per day, displaying real-time strikes for Australia, New Zealand and Europe.

Example of one-hour real-time lightning observations for Wednesday 29 March as shown in MetOceanView. Red dots represent clusters of lightning strikes.

Example of one-hour real-time lightning observations for Wednesday 29 March as shown in MetOceanView. Red dots represent clusters of lightning strikes.

We also use observational data to calibrate and validate our meteorological, wave and hydrodynamic forecast models. Observed data can also be used to directly improve our near-real-time forecasts, and can result in significant accuracy gains.

“All this information comes from a variety of sources. Much of the data used in MetOceanView are from our own models and instruments, but some observational data come from external providers. Some of it is private, for example where clients have observations that can help improve the models for their locations.  

“Ingesting such quantities of data requires a range of techniques. Often we have to process the raw information coming in to make it useable for our internal databases. We have designed our systems so that they can handle any data format.

“Ultimately, our clients use MetOceanView as a one-stop-shop for their marine weather information needs. The data we incorporate are valuable to our clients because they help them gain the complete picture of the atmospheric and marine conditions at their site. Good data visualised in an easy-to-understand format allows informed decision-making, which makes for safer operations and increased efficiency, and that is what MetOceanView is all about.”

For more information about MetOceanView, watch our introduction video here, see www.metoceanview.com or email enquiries@metocean.co.nz
 

An operational oceanographic forecast / hindcast model for Shanghai, China

MetOcean Solutions recently completed the development of operational high-resolution wave and hydrodynamic models for the Yangtze River mouth and coastal areas off Shanghai. 

The work combined cutting-edge science within our agile operating system to set up wave and current models for Hangzhou Bay, a region within the East China Sea which is partially enclosed by the Ryukyu chain of islands. 

“It is a very tricky area to model,” notes Senior Oceanographer Dr Rafael Guedes. “The region is characterised by a wide, shallow and highly irregular shelf with many small islands and underwater reefs. Accurate bathymetry for the area is limited. The site is strongly influenced by the phenomenal seasonal discharge from the Yangtze River, which is one of the largest rivers in the world, and is also subject to strong tidal currents.” 

 Bathymetry of the East China Sea. Red dots show the locations of measured data used to validate the models.

 Bathymetry of the East China Sea. Red dots show the locations of measured data used to validate the models.

 
Progressive downscaling of outputs from MetOcean Solutions’  global wave model WAVEWATCH III  using two SWAN nests. 

Progressive downscaling of outputs from MetOcean Solutions’ global wave model WAVEWATCH III using two SWAN nests. 

Snapshot of surface salinity from the ROMS model. Blue denotes low salinities; red high.

Snapshot of surface salinity from the ROMS model. Blue denotes low salinities; red high.

“In order to model the location well, we had to capture the meteorological events occurring within the East China Sea as well as the swell generated beyond the Ryukyu Islands which propagates into the bay. Frequent typhoons ravage the area, and these are always hard to resolve well. All in all, the area displays a challenging combination of highly variable bathymetry, strong temperature and salinity differences and complex mixing processes.”

The SWAN (Simulating WAves Nearshore) model was used to resolve the wave climate and the Regional Ocean Modeling System (ROMS) was applied to simulate the circulation. 

“To model the area we used a technique known as ‘dynamical downscaling’,” explains Rafael. “This process uses information from large scale global models to drive regional models at much higher resolution. The technique allows us to resolve fine-scale features near the coast while still accounting for remote influences to the area from long-generated swell or meso-scale currents.”

Quantile-quantile plot comparing measured and modelled significant wave height (Hs) for wave hindcast using (black) existing  CFSR  wind fields and (red) adjusted wind fields to correct for observed wind bias.

Quantile-quantile plot comparing measured and modelled significant wave height (Hs) for wave hindcast using (black) existing CFSR wind fields and (red) adjusted wind fields to correct for observed wind bias.

“High-quality input data sources are critical to running wave and hydrodynamic models in such complex settings,” continues Rafael. “We found persistent wind speed bias near the bay in the global reanalysis data source that we used to calibrate the high resolution models. Correcting this bias before running the wave model significantly improved model results just offshore of the bay as shown in the comparison of measured and modelled significant wave height.

The area has heavy shipping traffic, and the operational system outputs, including 7-day forecasts of site-specific waves, winds and currents, are now available to marine users. Please contact us and we will connect you with our partner agency in China.

The world's southernmost open ocean moored wave buoy deployed

The buoy will provide essential data about waves in the rarely studied Southern Ocean. Plot shows wave height in metres; the red dot marks the wave buoy location.

The buoy will provide essential data about waves in the rarely studied Southern Ocean. Plot shows wave height in metres; the red dot marks the wave buoy location.

In collaboration with MetOcean Solutions, the New Zealand Defence Force yesterday launched a moored wave buoy about 11 km south of Campbell Island. The site is the southernmost location that a wave buoy has ever been moored in the world.

Deployed from the HMNZS OTAGO, the buoy is part of a collaborative project between the Defence Technology Agency and MetOcean Solutions. The buoy is planned to remain in location for the next six months, where it will be used to gather precise wave spectral data as well as
wave height and wave direction.

"We are very pleased about our research partnership with the Defence," says oceanographer Dr Peter McComb who led the deployment on OTAGO. "The Southern Ocean is an incredible engine for wave energy generation due to the persistent westerly winds and the expansive ocean fetch. This makes it a difficult region to work in, but we were fortunate with a period of relatively good weather to launch the buoy. The data will be of international significance and the wave research community will benefit from open access to the measurements."

Dr Tom Durrant, the manager of MetOcean Solutions' wave modelling, says that the buoy will provide invaluable data for an area which remains poorly studied. 

"Due to the harsh ocean environment and remote location, the Southern Ocean is the least observed of any ocean body," he explains. "The wave buoy data will aid our understanding of waves in extreme conditions, and provide measurements against which we can validate and improve our global wave models. To help the deployment we provided detailed forecasts, and we are relieved that the conditions were calm enough to launch the buoy."

For more about the deployment, see the DTA website

 

High resolution wave forecasts for Chile now available

Good forecasts improve port safety and efficiency.

Good forecasts improve port safety and efficiency.

MetOcean Solutions have set up a high resolution wave forecast model for the coastline of Chile in South America. 
 
"We are delighted to now provide a high quality wave model for Chile," says Senior Oceanographer Dr Rafael Guedes. "We've set up a regional domain covering the central and northern Chilean coast and can now provide nearshore wave forecasts for the area north of 41°S. Accurate wave forecasting is important for ports located along this dynamic, exposed coastline."

The Chile model domain, showing depth (left) and sample wave height (right).

The Chile model domain, showing depth (left) and sample wave height (right).

The work was initiated following a visit by MetOcean Solutions to Chile in October, where the need for high resolution port scale wave forecasts was made apparent.  
 
"We've used the state-of-the-art SWAN (Simulating WAves Nearshore) model," continues Dr Guedes. "Like many New Zealand ports, Chilean ports suffer from wave exposure. Accurate modelling can help ports save money and time, and increase safety. MetOcean Solutions specialise in forecasting wave conditions for weather-exposed ports, and we provide expert forecasts for a number of ports internationally already. We are very happy to potentially extend the service to Chile. Of course, very high accuracy forecasts require accurate bathymetry." 
 
The model domain was set up to cover the coast between 41°S and 17°S at 5 km resolution and was set up using full spectral boundaries from MetOcean Solutions' new, upgraded global WAVEWATCH III wave model. The new model can be accessed via the MetOceanView platform.