Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

Mixing for the ocean surface boundary layer and 

WAVEWATCH III model

B. Reichl

1Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

Focus: Improving representation of air‐sea interface physics in GFDL’s climate models

• Developed an energetic Planetary Boundary Layer (ePBL)ocean surface mixing framework for climate simulation

• Used a process level approach to introduce Langmuir turbulence to ePBL, reducing bias in ocean vertical mixing

• Examining wave coupled models at GFDL and NCEP to enhance NOAA weather and climate simulation capabilities

Introduction

2Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

Upper Ocean Mixing For Climate Models‐ Parameterized mixing schemes should be efficient & robust‐ ePBL uses 1st order approach trained with 2nd moment & LES results

Simplicity:(Physics)

Cost:(Numerics)

LES2nd Moment

ClosureTKE

Closure1st Order Schemes

Bulk Boundary layer

Constant mixing coefficients

Cheap Expensive

Simple Complex

3Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

ePBL uses an implicit, non‐local energeticmixing constraintWhy? Robust to numeric constraints (grid & time step)

ePBL: Implicit Numerics

Reichl and Hallberg, 2018, OMLarge et al., 1994Rodi, 1987

ΔZ = 1 m, ΔT = 30 s ΔZ = 20 m, ΔT = 30 s ΔZ = 1 m, ΔT = 7200 s ΔZ = 20 m, ΔT = 7200 s

4Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

ePBL: Accurate Physics‐ Mixing constraints for ePBL validated from Large Eddy Simulations‐ LES also used to add wave effects on mixing (ePBL ePBL‐LT)

Reichl and Hallberg, 2018 OM; Reichl and Li, 2019, JPO

Turbulent vertical buoyancy flux

Time 2π/f0

Dep

th [m

]

0

-60

Includes Waves

Langmuir Turbulence

Thorpe, 2004

Large Eddy Simulation, with waves

Dep

th [m

]

0

-60

Large Eddy Simulation, no waves

Increasing mixing

Para

met

eriz

ed M

ixin

g

Mixing in LES

ePBL

ePBL

Para

met

eriz

ed M

ixin

g

Mixing in LES

WithWaves

[m3 s-3]

[m3 s-3]

[m3 s-3]

[m3 s-3]

ePBL-LT

Para

met

eriz

ed M

ixin

g

Mixing in LES

WithWaves

[m3 s-3]

[m3 s-3]

5Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

Ocean Mixing in GFDL Climate ModelsePBL helps improve simulated ocean mixed layer depth

Adcroft et al., 2019, JAMES Held et al., 2019, JAMES Dunne et al. (in prep)

Obs: Obs: Obs:

Obs: Obs: Obs:

10 50 7030

Summer MLDObs

OM2RMS: 5.48 m r2 = 0.70

OM4RMS: 3.06 m r2 = 0.73

CM3RMS: 4.98 m r2 = 0.64

ESM2MRMS: 5.21 m r2 = 0.48

CM4RMS: 3.21 m r2 = 0.70

ESM4RMS: 3.46 m r2 = 0.68

6Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

ePBL helps improve simulated ocean mixed layer depth

Ocean Mixing in GFDL Climate Models

Adcroft et al., 2019, JAMES Held et al., 2019, JAMES Dunne et al. (in prep)

Obs: Obs: Obs:

Obs: Obs: Obs:

10 100 1000

ObsWinter MLD

OM2RMS: 68.60 m r2 = 0.34

OM4RMS: 44.34 m r2 = 0.41

CM3RMS: 91.37 m r2 = 0.24

ESM2MRMS: 70.53 m r2 = 0.24

CM4RMS: 45.34 m r2 = 0.54

ESM4RMS: 43.78 m r2 = 0.51

7Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

GFDL‐NCEP Collaboration & WAVEWATCH III

Atmosphere

Land

Ocean

Imag

es: 1

-usr

a.ed

u, 2

-yal

e.ed

u, 3

-mit.

edu,

4-n

sidc

.org

, 5-n

oaa.

gov

Wind & Stokes Drift Correlation CoefficientJRA-55 + WAVEWATCH III, 2004‐ Ocean waves are critical 

component of air‐sea physics‐ Yet, waves are not explicitly represented routinely in models

‐ Wave coupled models are critical to understand their impacts in Earth System Models

1.00.5

Increase of Summer MLD w/ wave mixing (ePBL-LT) 10m

-10m

8Geophysical Fluid Dynamics Laboratory ReviewOctober 29‐31, 2019

SummaryImproved representation of upper ocean physics

• ePBL in GFDL “4th generation” models (OM4, CM4, ESM4)• Implicit numerics w/ realistic physics• Wave‐driven mixing for realistic Southern Ocean

• WAVEWATCH III coupling for surface wave simulationFuture Work

• Apply ePBL principles continuously through water column• Bottom boundary layers• Breaking internal gravity waves• Internal tides

• Wave coupling• Improve air‐sea flux parameterizations (e.g. momentum, gases [CO2, Reichl and Deike, in revision], heat, mass)

• Represent wave/cryosphere interactions (sea ice, icebergs)• Sea‐state dependent sea‐salt aerosols• Weather vs climate scale: Wave variability and extremes