Coupled vs. Decoupled Boundary Layers in VOCALS-REx ACP article describing these results:

http://www.atmos-chem-phys-discuss.net/11/8431/2011/

Chris Jones Department of Applied Math

Chris Bretherton

Department of Atmospheric Sciences

Dave Leon Department of Atmospheric Sciences

University of Wyoming

VOCALS RF05, 72W, 20S

Importance of Decoupling • Cloud-topped boundary layers (CTBLs) can vary from simple, vertically

well-mixed structure to structure w/vertical gradients to cumulus regions • Vertical structure of CTBL important to …

– Cloud cover – Vertical mixing processes – Precipitation

• CTBL decoupling is the separation of BL turbulence into separate surface-driven and cloud-driven layers

(Albrecht et. al. 1995)

October 2008-November 2008

SE Pacific

(http://www.atmos.washington.edu/~robwood/VOCALS/vocals_uw.html)

C-130 data available at http://www.eol.ucar.edu/projects/vocals/

C-130 flight path (grey) Cloud base (lidar-derived) LCL (“well-mixed cloud base”)

Radar reflectivity (drizzle proxy)

(courtesy of Rob Wood)

We use vertical profiles and subcloud level legs

Decoupling in VOCALS-REx Overview

• Use C-130 flight legs to measure extent of decoupling

– Profile-based decoupling index

– Subcloud leg decoupling index

• Dominant mechanism(s) for decoupling?

• Investigate relationship between inversion jumps, decoupling, and Sc breakup

Profile data and selection criteria

• 110 profiles included: – North of 25 S (ignore 2 coastal aerosol flights) – Extend from ~150 m to above inversion

• Used 1 Hz C-130 in-situ atmosphere state measurements, averaged into 10 meter vertical bins (courtesy of Chris Terai)

• Inversion base (𝑧𝑖) defined as height of minimum T provided RH > 45%

• Decoupling classified using: – Total water mixing ratio (𝑞𝑇 = 𝑞ℓ + 𝑞𝑣)

– Liquid potential temperature (𝜃ℓ = 𝜃 −𝐿

𝑐𝑝𝑞ℓ )

Profiles Decoupling metric(s)

Subcloud layer

Cloud layer

Δ𝑞 Δ𝜃ℓ Well-mixed Decoupled

Profile-based decoupling index: Δ𝑞 = 𝑞𝑡 𝑧∈[0,0.25𝑧𝑖] − 𝑞𝑡 𝑧∈ 0.75𝑧𝑖,𝑧𝑖

(bottom 25% - top 25%)

Δ𝜃ℓ = 𝜃ℓ 𝑧∈[0.75𝑧𝑖,𝑧𝑖] − 𝜃ℓ 𝑧∈[0,0.25𝑧𝑖]

Subcloud data and selection criteria

• 89 level legs flown at 150 meters – 10 min segments (~60 km horizontal extent)

• In-situ T and q measurements used for LCL

• Wyoming Cloud Radar (WCR) – Cloud top

– Column-max radar reflectivity (drizzle proxy)

• Wyoming Cloud Lidar (WCL)-derived cloud base

• Decoupling classified by leg mean Δ𝑧𝑏 = 𝑧𝑏 − 𝐿𝐶𝐿

Subcloud legs Decoupling metric: Δzb = 𝑧𝑏 − 𝐿𝐶𝐿

(actual cloud base – “well-mixed” cloud base)

drizzle

Profiles Decoupling metric(s)

Surface layer

Cloud layer

Δ𝑞 Δ𝜃ℓ Well-mixed Decoupled

Decoupling Distribution

Δ𝑞 < 0.5g/kg Δ𝜃ℓ < 0.5 K

Δ𝑧𝑏 < 125m

Well-mixed criteria

Well-mixed (28%)

Profiles Subcloud legs

Consistency of decoupling metrics

Least-squares fit

Thermodynamicargument

Composite profiles

𝜃ℓ 𝑞𝑇

10 𝑞ℓ 𝑧𝑖

𝑧𝐿𝐶𝐿

Organizing the results: can we deduce dominant physical mechanism(s)?

• Diurnal decoupling (solar absorption): warms the cloud layer, inhibits mixing b/w cloud and SC layer. – Not enough midday measurements

• Drizzle-induced decoupling: Latent heating of cloud layer and evaporative cooling of SC layer inhibits mixing.

• Wind speed / latent heat flux: Increased LHF => increased in-cloud buoyancy production of turbulence => more entrainment => more decoupling.

• Boundary-layer deepening: Deeper well-mixed cloud-layer => more buoyancy flux => more turbulence => more entrainment => more decoupling.

POCs: Pockets of Open Cells

• Several VOCALS flights sampled POCs • POCs characterized by …

• Low droplet concentration • Enhanced drizzle • Broken clouds • Pronounced decoupling

Are decoupling processes in POCs statistically different than in other Sc regions?

Decoupling somewhat correlated to drizzle, but drizzle not necessary for decoupling

Non-drizzling

Drizzling

Decoupling uncorrelated to wind speed

Bretherton and Wyant (1997) suggested stronger latent heat fluxes should promote decoupling – not seen in our results.

𝐿𝐻𝐹 = 𝜌0𝐿𝑣 𝑤′𝑞𝑡′ = 𝜌0𝐿𝑣𝐶𝑇𝑉 𝑞𝑠𝑢𝑟𝑓

∗ − 𝑞𝑡𝑀

𝐶𝑇 ≈ 0.001 (transfer coefficient) 𝑉 = leg-mean 150 m horizontal wind speed 𝑞𝑠𝑢𝑟𝑓

∗ = saturation mixing ratio at

sea surface 𝑞𝑡𝑀 = subcloud leg-mean 𝑞𝑡

“Well-mixed Cloud Thickness” Δ𝑧𝑀 best predicts decoupling

(this plot is the centerpiece of our results)

Δ𝑧𝑀 = 𝑧𝑖 − 𝐿𝐶𝐿: thickness the cloud would have if it was well-mixed

Subcloud Legs Profiles

Hollow marker = POC

Mixed-layer model flux ratio condition for decoupling (Bretherton and Wyant, 1997)

• Decoupling occurs when Δ𝐹𝑅

𝐿𝐻𝐹< 𝑄 ≡ 𝐴𝜂

Δ𝑧𝑀

𝑧𝑖

• Too much uncertainty for meaningful quantitative test of this relationship, but qualitative agreement with observations

• In VOCALS-REx, 𝑄 varies more than Δ𝐹𝑅

𝐿𝐻𝐹. From east to west

–Δ𝐹𝑅

𝐿𝐻𝐹 varies from approximately 1.0 to 0.7

– 𝑄 / Δ𝐹𝑅

𝐿𝐻𝐹 increases from 0.3 to 0.9 (decoupling threshold at

approximately 0.4)

𝐴 ≈ 1.1 (Caldwell et al.,2005) 𝜂 ≈ 0.9 (thermodynamic variable)

Δ𝐹𝑅 (radiative flux divergence) ranges from 71 to 95 𝑊𝑚−2 𝐿𝐻𝐹 ranges from 70 to 135 𝑊𝑚−2

(restricted to morning, non-drizzling legs)

Organizing the results: dominant decoupling mechanisms from VOCALS data

• Diurnal decoupling (solar absorption) – Not enough midday measurements

• Drizzle decoupling – No heavily drizzling well-mixed profiles, – Drizzle promotes decoupling, but is not primary cause

• Wind speed / latent heat flux – Not at any given longitude, but LHF does increase to west

along with decoupling

• Boundary-layer deepening – Well-mixed cloud thickness Δ𝑧𝑀 = 𝑧𝑖 − 𝑧𝐿𝐶𝐿 is the best

predictor of decoupling in VOCALS-REx data – Δ𝑧𝑀 > 500 𝑚 ⇒ decoupled

Inversion Jumps

• Lock (2009) and others have suggested high values of

𝜅 = 1 +𝑐𝑝𝛿𝜃ℓ

𝐿 𝛿𝑞𝑡

induce strong entrainment and Sc cloud breakup.

• Strong entrainment might also favor decoupling.

𝛿𝜃ℓ

𝛿𝑞𝑡 Inversion base

Inversion “top”

Calculating inversion jumps – Inversion base is already determined

– Inversion top: objective criteria based on RH and 𝜃ℓ profiles

– Complex POC inversion structure => identify jumps visually

POC Profile

𝛿𝜃ℓ

𝛿𝑞𝑡

Non-POC Profile

Inversion base

Inversion “top”

Decoupling not correlated with inversion jump parameter

𝜅 = 1 +𝑐𝑝𝛿𝜃ℓ

𝐿 𝛿𝑞𝑡

• Use REx C-130 profiles to calculate jumps/decoupling, adjacent subcloud

legs to calculate cloud fraction. Restrict to flights before 10:00 LT in left panel.

• κ > 0.4 often (but not always) goes with broken cloud.

• For κ < 0.5 there is no obvious correlation of κ and decoupling.

• POC and non-POC distributions overlap

Blue = well-mixed Red = decoupled Hollow = POC Dash = Lock (2009) LES results

Summary

• Well-mixed cloud thickness Δ𝑧𝑀 was best predictor of decoupling. – Δ𝑧𝑀 > 500 m: decoupled – Δ𝑧𝑀 < 500 m: well-mixed

• LHF and drizzle increase to the west. Both are likely contributing mechanisms in VOCALS decoupling, but no single parameter predicts decoupling as well as Δ𝑧𝑀.

• Inversion jump parameter is 𝜅 not a good predictor of decoupling, but qualitatively agrees with cloud cover predictions.

Extra slides

EPIC 2001 (Bretherton, et al.)

Some important mechanisms that come into play

Consistency of decoupling metrics (thermodynamic argument)

0.5 g/kg

0.0048 g/(kg m)

Δ𝑧𝑏 ≈ 104 m

(+ a bit for the term we dropped)

Δ𝑞 = 𝑞𝑡1 − 𝑞𝑡2 = 𝑞𝑣 𝑧𝑆𝐶 − 𝑞𝑣 𝑧𝑏 = 𝑞∗ 𝑝𝐿𝐶𝐿 , 𝑇𝐿𝐶𝐿 − 𝑞∗ 𝑝𝑏, 𝑇 𝑧𝑏 = 𝑞∗ 𝑝𝐿𝐶𝐿 , 𝑇𝐿𝐶𝐿 − 𝑞∗ 𝑝𝑏, 𝑇𝑑𝑎 𝑧𝑏 + 𝑞∗ 𝑝𝑏, 𝑇𝑑𝑎 𝑧𝑏 − 𝑞∗ 𝑝𝑏, 𝑇 𝑧𝑏

𝑛𝑒𝑔𝑙𝑒𝑐𝑡

Δ𝑞 ≈ −𝑑𝑞∗

𝑑𝑧𝑑𝑎

𝑧𝑏 − 𝑧𝐿𝐶𝐿

𝑧𝑆𝐶

𝑧𝐿𝐶𝐿

𝑧𝑏

𝑞𝑡1 = 𝑞∗(𝑝𝐿𝐶𝐿, 𝑇𝐿𝐶𝐿)

𝑞𝑡2 = 𝑞𝑣 𝑧𝑏 = 𝑞∗ 𝑝𝑏, 𝑇 𝑧𝑏

𝑞𝑡1 = 𝑞𝑣 𝑧𝑆𝐶

z 𝑞𝑡

• Neglect weak p-dependence in 𝑞∗ • Use characteristic BL reference T,p

• 𝑝0 ~ 950 hPa • 𝑇 ~ 285 K

Consistency of the two metrics (rough thermodynamic argument)

0.05 g/kg 0.0048 g/(kg m)

104 m (+ a bit for the term we dropped)

Assumed to follow approximately dry adiabat, s is conserved

Calculating inversion jumps

• Want systematic approach to identify inversion jumps. – Inversion base is already determined – Inversion top:

• Relative humidity (RH) gradient below some threshold for at least 100 meters (to eliminate “jitters”):

𝑑 𝑅𝐻

𝑑𝑧< 0.3% per meter

• RH close enough to min value in free troposphere (to make sure it’s really the full inversion jump):

𝑅𝐻 < min 𝑅𝐻 + 10% • 𝜃ℓ gradient also below threshold (moisture and temp jumps

should be the same, but not always the case here): 𝑑𝜃ℓ

𝑑𝑧< 0.1 K per meter

– Identify inversion top visually for POC flights

Decoupling when Δ𝐹𝑅

𝐿𝐻𝐹< 𝑄 ≡ 𝐴𝜂

Δ𝑧𝑀

𝑧𝑖

70∘ − 75∘W

Flux ratio Q Decoupled

Coupled

Decoupling due to increased Q more than change in flux ratio (consistent with “deepening/warming” mechanism)

75∘ − 80∘ 80∘ − 86∘