Transcript Slide 1

Bimodality in the Vertical Structure of
Tropical Diabatic Heating
Chidong Zhang
RSMAS, University of Miami
In collaboration with
Samson Hagos, Wei-Kuo Tao, Steve Lang , Yukari Takayabu,
Shoichi Shige, and Masaki Katsumata
BIRS Conference on Multiscale Processes in the Tropics
April, 27- May 1, 2009
Banff, Alberta
Canada
Derive Q1 from in situ Observations
From sounding data (Yanai et al. 1973):
From radar data (Houze 1983):
Idealized Tropical Meso-Scale Latent Heating Profiles
Schumacker et al. (2008)
Premise:
Diabatic
heating
directly
relevant
to
Hartmann et
al (1984); Mapes
and profiles
Houze (1995);
Schumacher
et al. (2004)
the tropical large-scale circulation are aggregates of
heating profiles associated with different types of cloud
systems.
Questions:
• What are the structure and evolution of prevailing
diabatic heating profiles directly relevant to the tropical
large-scale circulation?
• How do the tropical large-scale atmospheric
circulations respond to the prevailing heating profiles?
• What are the roles of heating profiles in the MJO?
Sounding Data Sources
SCSMEX
NAME
KWAJEX
TWP-ICE
LBA
GATE
MISMO
TOGA
COARE
with help from
Paul Ciesielski, Steve Esbensen, Richard Johnson,
Masaki Katsumata, Yasu-Masa Kodama, Steve
Kruger, Wei-kuo Tao, Wen-wen Tung, Xiaoqing Wu,
Michel Yanai, Xiping Zeng, and Minghua Zhang
Marshal Islands (Yanai et al 1973)
Winter MONEX (Johnson and Young 1983)
AMEX (Frank and McBride 1989)
BOMEX (Nitta and Esbensen 1974)
TAMEX (Johnson and Bresch 1991)
TOGA COARE Q1
mean
mean+std dev
EOF2
REOF1
REOF2
EOF1
TOGA COARE Q1 Time Series
Original
Reconstructed from the first two REOF modes
4
5
3
2
6
7
1
8
original
REOF1&2
REOF1
REOF2
SCSMEX
NAME
KWAJEX
TWP-ICE
LBA
GATE
MISMO
TOGA
COARE
Combined REOF Analysis
REOF
Mode
Field
Experiment
Fractional Variance
1
2
1+2
TOGA CAORE
70.7
12.4
83.1
GATE
52.7
25.7
78.4
LBA
46.7
33.9
80.6
KWAJEX
80.8
17.0
97.8
TWP-ICE
83.1
8.7
91.8
MISMO
62.7
15.6
78.3
SCSMEX
76.4
8.2
84.6
NAME
45.2
19.8
65.0
Total
74.2
10.6
84.8
A
(BH)
B
(MH)
D
(RC)
C
(TH)
6-hourly and daily data
(a)
6-hourly
(b)
daily
A
(shallow)
B
(deep)
D
(clear)
C
(stratiform)
6-hourly and daily data
Dq
Phase A:
(BH)
stratiform deep
15% 45%
shallow/congestus
40%
Phase B:
(MH)
stratiform deep
37% 43%
shallow/congestus
20%
Phase C:
(TH)
stratiform
56%
shallow/congestus
6%
deep
38%
Latent Heating Profiles 15˚S – 15˚N
Latent Heating Profiles (mean ± standard deviation) 15˚S – 15˚N
Probability Distribution Function as a Function of Height
TRMM Storm Height (30˚S-30˚N)
Short and Naramura 2000
Maximum Q1 from the soundings
CSH Max
A linear, steady-state model (nondimensionalized):
-∇·(qV)
mean
A
B
C
Phase A:
(BH)
stratiform
15%
deep
A
45%
shallow/congestus
40%
Phase B:
(MH)
stratiform
37%
deep
43%
shallow/congestus
20%
Phase C:
(TH)
stratiform
56%
deep
38%
shallow/congestus
6%
A
B
B
C
C
Spectra of REOF PCs for TRMM Latent Heating
Composite heating for eight phases defined by the two leading HSVD modes for CSH
based on its two leading REOF modes combined (left) and overlaid (right).
Composite heating for eight phases defined by the two leading HSVD modes for CSH using
(left) its two leading REOF modes plus the time mean and (right) the original data.
90˚E
120˚E
150˚E
Lin et al
(2004)
(Wu 2003)
Zhang and Mu
(2005)
Zhang and Mu
(2005)
A GCM experiment (Li et al. 2009):
R42L9
Radiation scheme: Slingo et al (1996)
Cumulus scheme: Manabe et al. (1965)
Boundary layer: Holtslag and Boville (1993)
Land surface: Xue et al. (1991)
Summary
• Tropical large-scale diabatic (and latent) heat profiles
are ubiquitously dominated by two modes: deep and
shallow, independent of location and data sources
(soundings, TRMM retrievals, global reanalyses);
• These two modes define three prevailing large-scale
diabatic heating profiles that evolve in a sequence of
bottom, middle, and top heavy structures;
• The large-scale vertical overturning circulations
responding to the three prevailing heating profiles are
of multi-cell structures;
• Low-level, bottom-heavy heating appears to be
essential to the MJO.
Questions
• What are the physical/dynamical reasons for the two
dominant heating modes? Or are they simply statistical
artifacts?
• Are the two dominant modes of heating profiles
related in any way to the bimodal distributions of
heating peaks and precipitation echo height?
• Which one is more fundamental to the MJO, low-level
moistening or low-level heating?
• Is the multi-scale interaction in horizontal (synoptic vs.
planetary) or vertical (shallow vs. deep) more
fundamental to the MJO?