mellor yamada モデルの拡張 - jamstec › hpci-sp › info › meso_sg2014 › pdf › ... ·...

16
1/14 Terra Incognita となる解像度のための MellorYamada モデルの拡張 An extension of MellorYamada model to apply for the resolution of Terra Incognita 伊藤純至(東大AORI/気象研) 新野宏 (東大AORI) 中西幹郎 (防衛大) ChinHoh Moeng (NCAR)

Upload: others

Post on 06-Jul-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

1/14

Terra Incognita となる解像度のためのMellor‐Yamada モデルの拡張

An extension of Mellor‐Yamada model to apply for the resolution ofTerra Incognita

伊藤純至(東大AORI/気象研)新野宏 (東大AORI)中西幹郎 (防衛大)Chin‐Hoh Moeng (NCAR)

Page 2: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

2/14

1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

LES’s subgrid‐scale

1‐D PBL model

f

×

Spectrum density

Frequency 100 s1 Day4 Days

Resolution of conventional numerical model (≧5km) Horizontal resolution near

L (〜1km‐100m): Terra Incognita problem 

Spectrum of wind near surface

Size of energy containing eddy L

※ Terra Incognita: 未知の領域PBL: Planetaty Boundary Layer 大気境界層

Page 3: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

3/14

ObjectivesDevelop a new PBL model for resolution of Terra Incognita(TI)•Design based on LES reference

e.g. Honnert et al. (2012); Shin and Hong (2013); Kitamura@MRI (Submitted)

•Test run of a new model with horizontal resolution of TIe.g. Ramachandran et al. (2011); Boutle et al. (2014)

→  extension of Mellor‐Yamada’s PBL scheme used in JMANHM etc.

An extension of UK model’s non‐local PBL scheme

Page 4: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

4/14

2. LES Reference• Basic equation: Boussinesq approximation, Dry, Smagorinsky‐

Lilly’s SGS• Domain:  18km×18km×4.5km• Resolution:  25m• Lateral Bound.: Doubly periodic• Bottom Bound.:    Bulk method (Momentum) / Heat flux 

Q=0.2K×m/s• Geostropic winds: V=5 m/s• Δt:  0.2 S

Vertical velocity in middle of CBL18km

0 km

CBL (Convective Boundary Layer) grows

Sensitivity on Q and V (not shown)

Page 5: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

5/14

Horizontal filter 

Grid valuesHorizontal filter of size ΔH=3×25=75m

Obtain

Larger ΔH

f (xi ) f (xi )G(xi xi ) d xi G(xi xi ) 1/h ( xi xi h )

0 ( xi xi h )

ΔH=5×25=125m

Top‐hat filter:

Only wAverage  wVariation w2

Triplet  w3

Grid scale motionsSub‐filter TKETurbulent transport

Associated with

should be modeled for resolution ΔH

Page 6: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

6/14

Scaling and TKE budget 

etui

exi

ijSij g0

f3 xi

ujeuj e 10

uj puj p Production

Free convection scaling (Deardroff 1970) is operated on results with “*” • Length: height of convective 

boundary layer h• Velocity : Convective velocity w*

• Heat flux : Surface flux Q

MY model solves prognostic equation of TKE(乱流運動エネルギー): 

Advection Turbulent transport Dissipation

Vertical profiles ofover whole x‐y cross section

ΔH*→∞: Meso‐scale 

limit

Range 0.5 < ΔH* (≡ΔH/h)  <2 correspond to TI scale 

Page 7: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

7/14

3. Design of model: Estimated terms in TKE budget for various ΔH

*

0

0.2

0.4

0.6

0.8

1

1.2

-0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

Hei

ght z

*

e* budget

H*=3.16

1.270.800.33

0

0.2

0.4

0.6

0.8

1

1.2

-0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1H

eigh

t z*

e* budget

H*=3.16

1.270.800.33

Vertical Profiles of terms in TKE budget (horizontal average)

q3

L

① Dissipation:

[Solid] Turbulenttransportsmaller for small ΔH

*

Dissipation(almost constant)

[dotted] Advection(resolved)

Production(resolved)

② Turbulent transport (T.T.): T.T. qLh

exi

Assumed down‐gradient form

→ q: Turbulent velocity ∝ TKE

Page 8: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

8/14

① Horizontally averaged Dissipation length <Lε>

Estimated <Lε> for various ΔH*

Lε of Meso‐scale limit (e.g. Nakanishi and Niino,2009)

“Partition function” of TKE

(Honnert et al. 2012)can prescribe <Lε> for TI scale asLε×F(ΔH

*)

Page 9: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

9/14

② Diffusion lengthTurbulent transport=‐Ld q∂TKE/∂xi

Horizontal diffusion Vertical diffusion

Slope correspond to Ld

!Neglect counter‐gradient diffusion plotted in 2nd and 4thquadrants

ΔH*=1.27

ΔH*=0.33

To reduce Ld(ΔH*) as

Ld(∞)×F(ΔH*) seem 

reasonable To rationalize the neglect of counter‐gradient diffusion  → A posteriori test

Page 10: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

10/14

4. A posteriori test

• Replace Smagorinsky’s model in LES to“TI model” | MY model (Nakanishi and Niino 2009)

• Grid number: 12×12×150• Horizontal resolution dx=1.5km | 1km | 500m 

Typical “TI scale” ↑• Vertical resolution: dz=25m• Environments are the same for LES reference to compare

Page 11: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

11/14

Non‐dimensional horizontal resolution dx*(≡ dx/h)

0

0.5

1

1.5

2

2.5

3

1 2 3 4 5 6 7 8 9 10

dx*

Time (h)

TI dx=1500mTI dx=1000m

TI dx=500m

TI scale

LES scale

1.5 km 

1 km 

500 m 

Page 12: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

12/14

0

0.1

0.2

0.3

0.4

0.5

0 2 4 6 8 10 12

<w2 >/

w*2

Time (h)

MYNN =500mTIB =500m

@z=0.5h

0

0.1

0.2

0.3

0.4

0.5

0 2 4 6 8 10 12

<w2 >/

w*2

Time (h)

MYNN =1000mTIB =1000m

Strength of resolved convection 

“TI Model” reproduce resolved convection in accord with LES reference

LES, Observation<w*2> 〜0.4

MY

TI model

MY

TI model

w*2

w*2

dx=1km dx=500m

+ : Suggested by LES reference 

Page 13: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

13/14

Conclusion 1/2

• Design of “TI model” based on LES→ Almost completed (To be sumitted to BLM?)

• Implement in a numerical weather prediction (e.g. JMANHM, WRF) would be very easy↑ only reduce length scale in MY model

• We will not peruse further refinement (3rd order closure? stochastic forcing?) 

Page 14: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

14/14

Conclusion 2/2

• Current JMANHM with MY model maybehave qualitatively similar manner to “TI model” (伊藤ら,2013年春気象学会), but it is unjustified

• Seeking verification by JMANHM with “TI model” ← Karman vortex shading was not good…

Simulated cloud

by JMANHM with a LFM configuration (dx=2km)

Page 15: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

15/14

Local LεLε without horizontal average  @ ΔH

*=0.8, z*=0.5

Page 16: Mellor Yamada モデルの拡張 - JAMSTEC › hpci-sp › info › meso_sg2014 › pdf › ... · 2014-06-27 · 1. “Terra Incognita” (Grey Zone) problem in PBL turbulence modeling

16/14

鉛直速度w@5h,z=0.5h

dx=1km

dx=500m

x (km) x (km)

y (km)

y (km)

鉛直速度w(m/s)