1. The effect of tides on the hydrophysical fields in the NEMO-shelf Arctic Ocean model. Maria Luneva National Oceanography Centre, Liverpool 2011 AOMIP meeting Woods Hole

2. NEMO-shelf Arctic Ocean model NEMO with focus on shelf seas = NEMO-SHELF Developed in MYOCEANS project by: Met. Office, MERCATOR, L’OCEAN and NOC(Liverpool) Horizontal resolution : 1/32 o in the rotated coordinates system and 1/6 as a test domain Vertical grid: new generalised terrain following coordinates- s-coordinates= (s-z with partial step) high precision Horizontal Pressure Gradient algorithm ( the Pressure Jacobian Method ), Vertical advection (Piecewise Parabolic Method), (Hedong Liu, NOCL) variable volume Explicitly resolved tides. Generic tidal boundary conditions defined in NEMO code directly from global Oregon State University Inverse tidal model (1/4 o resolution ) + tidal geopotential. Vertical mixing : 2.5 turbulence closure model (GLS). Boundary conditions on the open boundary for tracers. Now : generic procedure – climatic Levitus. Coming soon: Global NEMO ¼ (NOCS). Generic procedure for generating the boundary conditions for regional model from tri-polar grid (James Harle, NOCL)

3. Local domain for test problem Locations of shelf edge cascades from observations (Shapiro et al, 2003, Ivanov et al, 2004). Illustration of cascading event

4. Problem 1: vertical coordinates We developed and tested hybrid coordinates with option of change of the depth of terrain –following s- coordinates and z-partial steps in the deep layers Hs=50m Hs=300m

5. Z-coordinates or too shallow hybrid coordinates (even with advective BBL) do not allow cascading on the weak slope: The concentration of passive tracer indicating the presence of cascading water at time t=100 days for different depth of s-layer. Hs=50mHs=300mHs=600m

6. And Errors : pure effect of s-coordinates depth on the strength of spurious currents with initial horizontally homogeneous stratification SSH for Hs=50m at day 60 Max value <0.003m SSH for Hs=600m at day 60 max value 0.02m Log 10 (KE)/2 for Hs=600m at day 60 max value -0.4 (TKE~ 0.1 m 2 s -2 )

7. Forcings and runs 1. Surface fluxes: U10, V10 CORE fluxes +ERA40, ClIO formulation 2. New river runoff forcing from Dai and Trenberth (Feb2011) 1200 gages 3. Tidal boundary conditions (8 constituents) + tidal geopotential (15 components) 4. Tides+ Flather condition Correction of Bering Strait inflow 0.8 SV mean + seasonal variability +. 5. Initial 10 years runs for 18km resolution, starting from Levitus cimatology and 1958 forcing 6. Short 1 year run for 3km resolution.

8. Tides are reproduced relatively well The amplitude of M2 tide from inverse model AOTIM (5 km resolution) The same from pan-Arctic NEMO-shelf and OSU boundary conditions (full baroclinic problem). No tides in the open boundary – amplitude does not exceed 0.3 m

9. Cascades are in the right place. Using of tracers to identify the brine rejection and locations of cascades Brine tracers at the lower terrain following layer (bottom layer of the shelf) Brine tracers below the last terrain following layer (400m) identify the intensive cascading at the Sviataya Anna Trough (layer below the f the shelf is shown)

10. Ice cover and volume: not very bad.. but have strong cooling/ice grow trends

11. Ice cover: not bad in winter, too large in summers Problem: Ice volume increases, even during 2000-2007… Too deep mixed layer Wrong heat balance and inflow? Too strong mixing?

12. What additional terms in momentum equations we get with tides Momentum equation:

13. Evaluation of the effects of tides on the mean currents: Save moments every 10 days, calculate yearly mean Yearly mean SSH differences for simulations with and without tides

14. Amplitude of bottom shear stresses Tide induce a shear stresses near bottom amplification up to 3 times, change of angle

15. Associated bottom Ekman pumping:

16. Lateral temperature fluxes due to small-scale/ high frequency processes In the simulation with tides the amplitude Is much higher, with large differences In the deep part of basin. tide No tide difference

17. Basin mean vertical salt and heat fluxes in comparison with simulation with turned off tides -, yearly and monthly mean, Effect of tides ~ increases heat flux in 20% -, yearly and monthly mean, Effect of tides decreases salt flux in 20%

18. Conclusions The effects of tidal Reynolds stresses on the slow varying components of currents are relatively small. The effects of tidal bottom shear stresses on the slow varying component of currents are strong. Tidal bottom (and sea-ice) shear stresses generates the Ekman pumping O(0.01-1 mm/s) Surprisingly, tides change the vertical transports of heat and salt at the depths of 1000km by 20% Horizontal small scale (high frequency) heat and salt fluxes due to tides an order higher then caused by other processes