UK - China Initiative to Develop Predictive Multi-Scale Ocean Modelling as a Key Aspect of a Joint Environmental Modelling Centre

Lead Research Organisation: Imperial College London
Department Name: Earth Science and Engineering

Abstract

China is increasingly taking the lead in solutions to environmental problems and this will continue as substantial Chinese investment is scheduled for this purpose. The Institute of Atmospheric Physics (IAP) in Beijing is an internationally leading organisation in this area and will substantially benefit from this additional investment. The Applied Modelling and Computational Group at Imperial College London (AMCG-ICL) is an environmental modelling group developing next generation methods. We propose a two year starter project with a combination of training and scientific effort in the UK and China synchronized with a range of supporting activities that will build the foundation for a subsequently self supporting (combination of UK and Chinese funds) 'International Research Centre'. The Centre will combine our world leading technologies and manpower to accelerate research excellence and delivery of numerical modelling insights and solutions to grand challenge environmental problems in the UK and China way beyond the capability of the UK alone. A relatively small investment would leverage China's massive past, current and future investments in IAP and past UK investments in next generation environmental flow models (particularly the multi-scale ocean model Fluidity-ICOM). This collaboration will develop a world leading predictive modelling framework. The starter project funded here will provide the focus for the training and collaboration so as to apply IAP's data assimilation methods to our multi-scale ocean model Fluidity-ICOM.

Next Generation Ocean-Atmosphere Model:

A Grand Challenge in Earth System Science is modelling the global circulation across the full range of relevant spatial and temporal scales. For climate prediction, this means resolving both basin scale and smaller scale features such as boundary currents, mixing; chemical interactions and transport, overflows, and mesoscale eddies. Such simulations will lie well beyond the capability of traditional ocean and atmosphere models. It is now generally recognised that the next generation of ocean models will be based on unstructured mesh technology as currently this is the only feasible way of resolving the important range of scales in coastal regions. As identified by the NERC strategy document 'oceans 2025 WP9', unstructured mesh ocean models are the key ocean modelling technology for the future modelling of multi-scale ocean to estuary and smaller scale modelling. Among existing unstructured mesh models, ICOM-Fluidity is the only model that can be used for simulation of flow on all scales using adaptive mesh resolution and is therefore an ideal platform for the next generation of data assimilation models to be developed on. One result of the training program will be that ICOM-Fluidity will be used to form a forward model of the China Sea. There will be a large amount of data to assimilate into the model e.g. satellite, argo floats and ship tracks. Ensemble Kalman Filter EnKF and gradient or adjoint based data assimilation methods will be used with ICOM-Fluidity to provide forecasts and to interpolate available data.

Planned activities that will support the IAP - AMCG-ICL research:

1) Training courses, workshops and Summer schools.
2) PhD students, PDRAs and senior staff time to apply (and help develop) the model e.g. set up the UK and China sea model and develop uncertainty, reduced order and data assimilation methods.
3) Exchanges of academic staff and PhD students.
4) Development of a new substantial funding grant in China and the UK to fund the centre.
5) Strengthened UK link and development of further initiatives with the Chinese Academy of Sciences.
6) Formalised visiting status for key Imperial College researchers

Planned Impact

The proposed work will build upon an established infrastructure to develop a collaborative relationship between AMCG-ICL and IAP. This project will include extensive training in model use and development with apprentice-piece projects focused on tidal modeling and data assimilation. This initial collaborative work will form the foundation for an 'International Centre' to tackle grand challenge environmental problems. This centre will ultimately focus on establishing a world leading predictive environmental modelling framework. This will enable observations and experimental measurements to be used within models. This predictive framework will allow optimisation of the type and location of expensive observations/measurements and provide optimal control (with feedback from observations) of climate/pollution. One of the best examples of this was performed by IAP to generate the 'blue skies' for the Beijing Olympics. A key advantage of forming a joint centre between AMCG-ICL and IAP is that the combined science and infrastructure puts the Centre in a position to win substantial grants/awards and generate large scientific and economic impacts.

Our partner organizations from across the world would use and develop model technology from the resulting framework and this synergy will further develop the capability of the model and center.

This work will benefit the following: Consulting Companies; Engineers; Observational Data Centers/Companies; Designers in Organizations/Industries; Power Companies; Environmental Agencies/Industries; Government Security Bodies; Government, Regulatory Bodies and Stakeholders.

Planned activities to ensure good engagement and communication with beneficiaries in ocean modelling and wider fields include: exchanges between AMCG and IAP; documentation and tools to improve the user experience; code packaging; post processing tools; improved web presence; training events; publications; collaboration with Universities and industries.

IAP provides a route for exploitation of the proposed technology through its research work and government/industry links. The UK is also committed to developing unstructured mesh ocean models (Oceans 2025) and provides similar routes (e.g. through AMCG-ICL and UK labs NOCS, NOCL and HR Wallingford) for exploitation and application. In addition, ICL has a subsidiary Imperial Innovations with the remit of promoting and commercializing innovation and expertise developed within the University, for the benefit of industry and the wider community. Imperial Innovations supports the transfer of expertise and knowledge into industry and has experience of many software spin-outs. This expertise will be used to negotiate the type of flexible agreements that would be required for research consultancy and collaborative work, or expanding the commercial use of our tools.

Scientists will benefit from this research through improved prediction of the multi-scale coastal flow processes, especially the prediction of extreme weather and climate phenomena such as storms and floods. In particular the ability to determine the sensitivity of these events to key variables will provide reliable information to both the public, and the policy decision-makers at regional, national, and international levels. Individuals in academia would be expected to develop and help deliver an open source, easy-to-use model with which they can perform their own studies. The rapidly growing academic user base of the Imperial College Ocean Model (ICOM) will benefit from the enhanced predictive/analysis ability provided by data assimilation, sensitivity and uncertainty analysis.

We highlight the following academic beneficiaries: Geoscientific computational fields; Meteorology, Oceanography communities; Multi-physics applications; Computer science; and established collaborating Institutes in the UK and globally.

Publications


10 25 50
Ardjmandpour N (2014) Reduced order borehole induction modelling in International Journal of Computational Fluid Dynamics
Buchan A (2013) A POD reduced-order model for eigenvalue problems with application to reactor physics in International Journal for Numerical Methods in Engineering
Che Z (2014) An ensemble method for sensor optimisation applied to falling liquid films in International Journal of Multiphase Flow
Du J (2016) Ensemble data assimilation applied to an adaptive mesh ocean model in International Journal for Numerical Methods in Fluids
Lin Z (2017) Non-intrusive reduced order modelling with least squares fitting on a sparse grid in International Journal for Numerical Methods in Fluids
 
Description The grant supported the development of a collaboration between staff at Imperial College with academics in China. This focussed on training Chinese academics in the use of a computer model that can be used to model fluid flow in the ocean and atmosphere. Part of the training involved several "apprentice-piece" projects, some of which have led to publications and some which will lead to publications in the future.

The model itself is open source, which means it is free for anyone to use. It is however, a model in development (although it is already being used in some industrial applications). The benefit of having skilled Chinese developers is that model development will be faster.

A key development so far that has grown from one of these "apprentice-piece" projects is a new approach to the way that models set up. This new approach will lead to greater accuracy and potentially faster run-times.
Exploitation Route Model development is very much an ongoing process but there are always opportunities to use what has already been developed. The group at Imperial has, for example, been funded by industry to model tides in the ancient past to predict the location of hydrocarbon reservoirs (£400k+) and by companies in the marine renewable business to inform the layout of tidal turbines.

The use of the data-assimilation methods that have been developed so far has not yet been used commercially but it does have the potential to improve the predictive power of the computational flow model that we use.
Sectors Energy,Environment
 
Description The main activity in this project was training with a series of "apprentice-piece"; projects aimed at delivering new methods for data-assimilation for application to computer modelling of environmental flows and real-wrold case studies. The training events resulted in researchers being funded by Chinese organisations to use our model in their work either working in China or studying at Imperial College. The data-assimilation methods have been implemented in the model and can now be used by anyone who wishes to use it in the future. The development of the data-assimilation methods also has some surprising industrial applications and influenced the development of computer software being used by geophysical equipment used to identify the oil-water contact in oil wells. Work focussed on modelling tides in the South China Sea prompted us to model Palaeotides in the region over the last 30 million years. This led to industrial funding by hydrocarbon exploration companies who were interested in the distribution of source rocks and reservoirs.
First Year Of Impact 2014
Sector Education,Energy,Environment
Impact Types Cultural,Economic
 
Description Funding to deliver initial Fluidity Training course at the Institute of Atmospheric Physics in Beijing
Amount £15,000 (GBP)
Organisation Chinese Academy of Sciences 
Department IAP Institute of Atmospheric Physics
Sector Academic/University
Country China, People's Republic of
Start 08/2012 
End 08/2012
 
Description Funding to deliver initial Fluidity Training course at the Institute of Atmospheric Physics in Beijing
Amount £15,000 (GBP)
Organisation Chinese Academy of Sciences 
Department IAP Institute of Atmospheric Physics
Sector Academic/University
Country China, People's Republic of
Start 08/2012 
End 09/2012
 
Description Short course support funds
Amount £3,500 (GBP)
Organisation Chinese Academy of Sciences 
Department IAP Institute of Atmospheric Physics
Sector Academic/University
Country China, People's Republic of
Start 03/2013 
End 03/2013
 
Description Short course support funds
Amount £3,500 (GBP)
Organisation Chinese Academy of Sciences 
Department IAP Institute of Atmospheric Physics
Sector Academic/University
Country China, People's Republic of
Start 03/2013 
End 03/2013
 
Description Student support
Amount £50,000 (GBP)
Organisation Shell Global Solutions International BV 
Department Shell Research Ltd
Sector Private
Country United Kingdom of Great Britain & Northern Ireland (UK)
Start 03/2013 
End 08/2016
 
Description Tidal Modellling of the South China Sea 
Organisation Chinese Academy of Sciences
Department Institute of Atmospheric Physics (IAP)
Country China, People's Republic of 
Sector Academic/University 
PI Contribution The Imperial College team have previously designed and implemented a new computer program for modelling fluid flows that can be used to simulate flow in the atmosphere, ocean and in various industrial applications. Members of the team have visited China and have run detailed short courses so as to train users in China in how to use our model. This training has extended beyond the training event and has also included advice on problem set up and implementation for those course attendees who have decided to use the model in their work.
Collaborator Contribution Our initial collaboration was with IAP but they invited a number of other attendees onto the course and this led to a broadening of the engagement. IAP initially paid for a party of 10 to attend their Institute and present a series of talks outlining the scope and formulation of the model. They also paid for the in China accommodation costs when we next visited to deliver the training. IAP also funded one of their staff members for 2 years to work in China as a collaborator.She used our model and developed sufficient proficiency that she was able to co-teach the course with the Imperial College team. This helped to develop confidence in the capabilities of the model. Some of the students on the short course subsequently used our model for the projects that they were working on and IAP then funded 2 of these students to come to Imperial College to further develop their projects.
Impact The collaboration is in the general area of computational flow dynamics as applied to coastal oceans and atmospheric flows. Outputs so far training courses, engaged researchers and scientific papers. Note that ongoing research includes projects that are only achievable because of the provision of validation data-sets from IAP.
Start Year 2013
 
Description Fluidity Training Course in Imperial College 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Participants in your research and patient groups
Results and Impact This is a training course focussed on our in-house computer model, Fluidity. The course is designed to introduce users to the formulation and operation of the model. Fangxin Fang, who was supported by this grant was part of a team that delivered the course at Imperial.

The course is an essential part of the education for all PhD students who are using or developing this model. At any one point in time there will be 10-20 students working on the model.
Year(s) Of Engagement Activity 2012,2013
URL http://fluidityproject.github.io/
 
Description Training Course on use of Fluidity at Tsighua University 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Participants in your research and patient groups
Results and Impact The course was presented by Fangxin Fang and Chris Pain to a group of faculty and grad students. It led to discussions from grad students about using our model.

This has led to new collaborations and support (data share) for future grant applications
Year(s) Of Engagement Activity 2013