1. RENUMERATE: Reducing numerical mixing resulting from applying tides explicitly in a global ocean model Alex Megann1 and Maria Luneva2 Project kickoff meeting, 5 April National Oceanography Centre, Southampton, UK 2National Oceanography Centre, Liverpool, UK

2. Context
Objectives
The team
Impact and relevance
Interactions with TACs and MFCs

3. Context
Most ocean-only, climate and operational forecasting models do not represent tides explicitly, instead using parameterisations of tidal mixing.
As a consequence, these models omit some key dynamic processes present in the ocean, including:
- high-frequency tidal currents, contributing to on- and off-shelf transport of physical and biogeochemical tracers
- mixing processes resulting from internal tides generated at the shelf edge;
- mixing over steep topography.

4. Ratio of vertical diffusivity with and without tides in Arctic NEMO
Context
Tides already implemented in an Arctic NEMO domain, and are currently being included in the ORCHESTRA high-resolution Southern Ocean model.
Ratio of vertical diffusivity with and without tides in Arctic NEMO

5. Context
Vertical tidal velocities can reach over 100 m/day in internal tides, and similar motions have been shown to result in spurious numerical mixing in ocean models of km resolution typical of CMIP6 climate models and many shelf sea models.
Megann (2018) showed, using a method based on watermass transformation rates, that these are associated with effective diapycnal diffusivities up to ten times those prescribed by model mixing scheme, and lead to long-term drifts in ocean interior.

6. Context Temperature drift (°C) in GO6
Diagnosed diffusivity (m2s-1) in GO6
Temperature drift (°C) in EN4
Ratio of diagnosed to explicit diffusivity
(from Megann, 2018)

7. Context
Enabling tides in current generation of ocean models is therefore likely to seriously over-represent their mixing effect in the ocean, unless a means can be found to mitigate this spurious mixing.
Potential solutions include:
Modification to vertical coordinate (z~) to enable elastic deformation of depth levels on short timescales;
Higher-order (less diffusive) vertical advection schemes.

8. Objectives
Overall objective: to improve representation of tides and tidal mixing in global and regional ocean models
Objective 1: To add explicit tidal forcing to latest global ¼° NEMO configuration and evaluate its effects on water mass characteristics and mixing.
Objective 2: To add z~ semi-lagrangian coordinate to this configuration and assess its effect in ameliorating numerical mixing from tidal and other high-frequency motions. Also implement, evaluate and optimise higher-order advection schemes.
Objective 3: To investigate effects of tides on large scale ocean circulation including global stratification; overturning and gyre circulations; shelf slope exchange, seasonal mixed layers; deep and bottom waters; and sea ice.

9. The Team
Dr Alex Megann, Marine Systems Modelling, National Oceanography Centre, Southampton
Over 25 years experience with ocean models
Particular interest in numerical mixing and vertical coordinate issues.
A core member of UK JMMP project, leading efforts to improve model physics towards and beyond CMIP6
Alex leads the project overall, as well as leading Work Packages 1, 3 and 4. He will set up the z~ configuration and carry out the integrations including this modification.

10. The Team
Dr Maria Luneva, Marine Systems Modelling, National Oceanography Centre, Liverpool
20 years of experience in coastal, regional sea and ocean modelling
A leading contributor to the ORCHESTRA Southern Ocean modelling project
Particular focus on the effects of convection, eddies and tidal mixing. 
Maria leads WP1, and is responsible for the creation, optimisation and evaluation of the tidally-forced configuration.

11. Project work packages and timeline
WP1: Set up GO8.0 model configuration and implement tides
Apr-Dec 2018 (Luneva)
WP2: Implement z~ scheme in GO8.0 configuration
Apr-Dec 2018 (Megann)
WP3: Construct and run configuration with tides and z~ combined
Jan-Dec 2019 (Megann, Luneva) 
WP4: Analysis of model integrations
Sep 2019-Mar 2020 (Megann, Luneva)

12. Impact on CMEMS SES
The objectives of the project address most directly Section 4.5 of the CMEMS SES R&D Priorities: Circulation Models for Global Ocean, Regional and Shelf Seas, in particular the short- to mid-term objectives.
CMEMS operational systems use NEMO ocean model, so improvements delivered by RENUMERATE will have direct pull-through.
Specifically, addition of tides will give much more realistic surface currents on hourly-daily timescales, leading to improved simulation of transport of biogeochemical tracers across the shelf edge.

13. Impact on CMEMS SES
The more realistic surface currents over the diurnal cycle and better representation of tides and mixing that the project will deliver are likely to lead to substantial improvements in the representation of sea ice processes, so will address
New Generation of Sea-Ice Modelling.

14. Impact on CMEMS SES
The use of the z~ scheme leads to mitigation of spurious mixing intrinsic to the NEMO model, as used by CMEMS operational systems, particularly in the presence of tides, so will benefit:
Development of advanced numerical schemes and parameterizations required for numerical codes implemented with a target effective resolution in the kilometric range
Development of advanced physical parameterizations, numerical schemes and alternative grids to improve the performance of the medium-resolution ocean models).

15. Impact on CMEMS SES
Seamless forecasting strategy will ensure that model improvements demonstrated with CMIP-class global model will feed through into coupled prediction products for seasonal, interannual and decadal forecasts using both physical and Earth System (physics plus biogeochemistry) models, and will therefore contribute to
Data Assimilative Modelling of Marine Ecosystems and Biogeochemistry
Coupled Ocean-Atmosphere Models with Assimilative Capability,
Ocean Climate Variability: Reanalysis, Forecasts and Scenarios for Future Changes).

16. Propagation pathways for results
Direct links to Mercator Ocean
PI Alex Megann is already working with Jerome Chanut at Mercator Ocean on development and testing of the z~ coordinate in NEMO.
Validation of this scheme in global NEMO implementation resulting from RENUMERATE will therefore directly feed back into CMEMS research and development, and is likely to lead to further code developments.

17. Propagation pathways for results
UK Met Office
Close collaboration already exists between NOC and Met Office through JMMP and CMIP6
Productive cross-fertilisation between RENUMERATE and JMMP in development of new GO8 global ocean configuration
Will enable pull-through of the model improvements to North-West Shelf Monitoring and Forecasting Centre, a CMEMS Co-Contractor.

18. Propagation pathways for results
DRAKKAR workshops
These annual meetings in Grenoble provide established international forum for exchange of technical and scientific modelling results with NEMO code.
Well attended by representatives from Mercator TACs and MFCs, including IFREMER, Meteo France, Nansen Centre and Danish Meteorological Institute, as well as UK Met Office.
The PI will propose a side meeting at a DRAKKAR workshop on numerical mixing, including effects of tides.