Shear-enhancing mesoscale momentum fmux? Using huge explicit - - PowerPoint PPT Presentation

Shear-enhancing mesoscale momentum fmux? Using huge explicit convectjon simulatjons as artjfjcal imaginatjon

Brian Mapes, Suvarchal Cheedela, Matuhew Niznik

University of Miami

Outline

• GEOS-5 Nature Run: an opportunity
• A primer on convectjve momentum

transport (CMT)

• A global perspectjve of CMT.
• Tools for sampling interestjng cases.
• A monsoonal case study.

G5NR: An opportunity

G5NR: An opportunity.

Beyond pretuy pictures: Hourly for 2 years: >4 PB

0.5-deg coarsened grids including nonlinear product terms

What to do with it all?

• Unique value: resolutjon (7km)

–Permits mesoscale organized deep convectjon

• Strategy

– Sample it for interestjng cases – Examine those cases in detail, understand

interestjng cases

• Eddy momentum fmux, a showcase metric of

mesoscale organized convectjon

• How does it efgect mean fmow?

Downshear case

• Isolated cumulus tend to consume shear .

• Can be upgradient
• That is, can increase

shear

• If the eddy is a tjlted

updraf

Does it need to be parameterized? How common is it? Use GEOS-5 Nature Run as -- Artjfjcial Imaginatjon

Can we construct a simple scalar score summarizing its contribution? → Then we can find how common the events are and analyze each event. → Can be used to design a simple parametrization, at-least for simple models

"Eddy" fmux

[u'w'] = [uw]-[u][w]

holds for any coarser rebinnings [] (here, 4 degree at instantaneous tjme scale)

• Eddy Flux has contribution from fluxes due to 7km resolved scale

convection (moist and dry) ,gravity waves from topography and stratosphere.

• NOT from any 7km subgrid scale parametrizations, for example

surface drag...

"Eddy"Momentum Tendency fmux

∂U ∂t =−1 ρ ∂ρ0U ' w ' ∂ z +ADV +PGF+.....[U=u, v] Eddy Flux Convergence=Eddy MomentumTendency=−1 ρ ∂ρ0U ' w ' ∂ z =g ∂ρ0U ' w' ∂ p

• EMT has contribution from fluxes due to convection (moist

and dry) resolved scale waves from topography and stratosphere.

• Vertical mass weighted average of <EMT> ~ 0 because

EMT only transports momentum between levels.

Shear Kinetjc Energy fmux

• EMT can change mean shear
• What we are interested in is where EMT

Increases and decreases mean shear.

• EMT. U where u (ushear) is u-<u>(baroclinic

component )

• LHS is SKE tendency or SKEDOT

~ ~ ~ ~

"KEDOT" fmux

→ When KEDOT < 0 EMT consumes shear kinetjc energy or eddies take this energy from the mean fmow. → When KEDOT > 0 EMT feed shear kinetjc energy on to mean fmow.

KEDOT and precipitatjon

Units: W m-2 Annual mean

JJA KEDOT

Units: W m-2 Annual mean

STD deviatjon of KEDOT

Units: W m-2 Annual mean

All these are means, but where are the cases Mitch and Lemone wrote about?

Tools for picking interestjng cases

→ Make course resolutjon statjstjcs → Select a case (python , Jupyter-notebooks) → Grab slices of data for selected case and make predetermined interactjve

• Visualizatjon. (Unidata-IDV, Jupyter notebook)

4-degree hourly grid cells Pick big SKEdot>0, and big Precip

Notebook includes quicklook imagery

Let's focus on Asian monsoon

• a locus of deep convectjon in strong shear
• does the convectjon damp or enhance the

shear?

Summary

• G5NR a nice opportunity to study mesoscale circulations.
• Time mean of SKEDOT, is overwhelmingly negative, implying sub-4deg

resolved motions act as positive viscosity on large scale shear.

• A global mean of -0.08W.m-2 indicates kinetic energy is extracted from

sheared (baroclinic) component of mean (4deg) flow into resolved sub 4-deg (meso) scales.

• Harnessing the power of “Big Data”sets needs efficient (modern?) tools.

https://github.com/matthewniznik/CHAD_G5NR http://bit.ly/Mapes_IDV https://github.com/suvarchal/JyIDV

Questjons to ponder -concurrent with the theme of workshop

Role of CMT in northward propagation, Kang et al 2010 Balance of forces and role of CMT in largescale budgets of Momentum in monsoon region Lin et al 2007