The mesoscale convection life cycle: building block or prototype for larger-scale tropical waves? Brian Mapes @ U. of Miami Stefan Tulich, Jialin Lin.

The mesoscale convection life cycle:building block or prototype for larger-scale tropical waves?

Brian Mapes @ U. of Miami

Stefan Tulich, Jialin Lin @ NOAA ESRL/PSD (formerly CDC)

Paquita Zuidema @ U. of Miami

paper in press at Dynamics of Atmospheres and Oceans

contact [email protected]

Resemblancesa) MCS: Zipser 1969b) MCS: Zipser et al. 1981c) 2-day: Takayabu et al. 1996d) Kelvin: Straub & Kiladis 2004e) MJO: Lin and Johnson 1996

Outline• Multiscale resemblance illustrated from obs

– Samples and regression composites– Smallest scale (building block): mesoscale (10’s km, h’s)

– but with convective cell substructure

– On up to intraseasonal

• Null hyp: LS structure inherited from MCS ‘block’ – rejected

• Instead: zones of (cu, cg, cb, ns) cloud type enhancements• Why and how?

– How to interpret humidity, divergence: proxies or causes of ‘convection’?

TOGA-COARE data sources at various scales

Sounding array data: CSU grids VAD: ship radar (48km radius)

Doppler radar VAD results: 48km radiuslag-p regressions of divergence vs. Z-R rainrate

mean of 7 field programs (~104 h data) (pattern similar in each of the 7)

Mass flux (-

Mapes and Lin 2005 MWR

Multi-hour time scale: artifact of spatial scale?

10-6 s-1

per mm/h

Only slight mass-balance adjustment neededOnly slight mass-balance adjustment needed

Point data (5’ vertically pointing cloud radar coverage vs. gauge rain, EPIC 2001):

irreducible multi-hour time scales?

cudynamics

(multi-cellular)

anvil micro-physics…

…includingstratiform

rain

Evolution (tilt) similar

across timescales with fixed

spatial scale

(OSA divergence)

30d

total

LP

+BP

+HP

=

dawn visible pix at those 4 timesb)

d)

a)

c)

Building block null hypothesis:

that multiscale tilt is reproducible by superposition of fixed-structure (VAD derived)

divergence building blocks

hmmm… No.

Rather, the MCS life cycle appears to be a prototype for larger-scale waves.

Zipser et al. 1981

zones of different cloud types

LS zone favors cu… … or cb … or ns anvils

How is multiscale structure resemblance achieved?

What variables modulate cloud types in these

LS zones?

Hypothesis: T, q (via parcel

buoyancy): emphasis on leading edge

Rain autocorrelation

Entraining plume buoyancy, and its partsPure parcel effect

Moistness of entrained air

“Explains” “convection”?

capping effect

Wrapup• Multiscale resemblance illustrated from obs

– Smallest scale (building block): mesoscale • with convective substructure

– COARE sounding array: ~2 day, ~2 week, ~40d periods

• Slow evolution is not just a convective cloud life cycle, aliased upscale: rather, there are zones of cloud type (cu, cg, cb, ns) enhancements– On both MCS and larger scales

• Why and how? – Leading edge: perhaps governed via plume buoyancy?

• Circularity: is q proxy or cause? “moisture-convection feedback”

– Trailing edge: more subtle? - Shear? Wave ascent?– Careful! Filtering can introduce spurious congestus stage - stratiform stage

antisymmetry! Or is it a “vertical mode?”

Extra slides

OSA specific humidity

6d 24d 40d

6 hourly data daily means 4d means

COARE divergence: OSA, IFA, VAD

OSA

IFA

VAD

conv

div

48hrain

Smallest scale: mesoscale (hours) even in “point” data

Gauge rain

Composites wrt +1 stdev excursions

Regressions vs. gauge rain

5 minute data