NERC Centre for Global Atmospheric Modelling

1

NERC Centre for Global Atmospheric
Modelling Department of Meteorology, University
of Reading
Scale Interactions on Diurnal to Seasonal
Timescales Their Relevance to Seasonal Model
Systematic Error Julia Slingo, Peter Inness,
Richard Neale, Steve Woolnough and Gui-Ying
Yang
2
CGAM Tropical GroupTOWARDS TROPICAL CLIMATE
PREDICTION
SST Variability e.g. El Nino
Diabatic Heating Response
Global Circulation Anomalies
Teleconnection
Translation
Atmospheric Bridge
Final Impact on Statistics of Local Weather
Lagged Ocean/Land Response
Scale Interactions e.g. MJO
Applications e.g. Crop Models
Primary route
Secondary route
3

• THE TALE OF TWO ERRORS!
• The Maritime Continent
• The Madden-Julian Oscillation

4
TOOLS

• Integrations of the Met Office Unified Model
• -HadAM3 AMIP II (observed SST, 1979-95)
• -Aquaplanet version of HadAM3
• -HadCM3 Control
• CMAP Precipitation data
• High resolution (0.50, 3 hourly) window
  brightness temperature data from the EU Cloud
  Archive User Service (CLAUS)

5
Typical window brightness (K) image from the
CLAUS dataset 12z 1 January 1992
6
Annual Mean Precipitation (CMAP) and AMIP II Mean
Model Errors
Climatology
NCAR
ECMWF
NCEP
Met Office
JMA
7
Topography of the Maritime Continent
8
Resolving the Maritime Continent in GCMs
9
Annual Mean Precipitation Errors in HadAM3
Sensitivity to Horizontal Resolution
10
HadAM3 Sensitivity Experiments Impact of
removing the islands of the Maritime
Continent(Neale and Slingo, 2001 Submitted to
J. Clim.)

• Land grid-points removed and replaced by ocean
  grid-points.
• Increased moisture availability from the sea
  surface leads to enhanced convection and partial
  correction of the model dry bias.
• Note also corrections to models wet bias in
  adjacent areas.

11
Global Impacts of Improved Maritime Continent
Heat SourceDJF 500hPa height (m) and Surface
Temperature (K)

• Potential improvements in the Maritime Continent
  heat source can have significant remote effects.
• Related to the generation of Rossby waves by the
  enhanced divergent outflow from the Maritime
  Continent heat source.
• Substantially reduces model systematic error over
  the extra-tropics of the winter hemisphere.
• Emphasizes the importance of considering the
  global context of model systematic error in which
  biases in the tropics may be a key factor.

12
The Diurnal Cycle in the Tropics(Yang and
Slingo, 2001 MWR, 129, 784-801)
Amplitude (K) of the diurnal harmonic
DJF
JJA
13
Phase of diurnal harmonic Local time of maximum
brightness temperature
DJF
JJA
14
Phase of diurnal cycle showing systematic
propagation of convective signal away from the
coast
Bay of Bengal, JJA Implied propagation speed
15-20 ms-1 ?Deep gravity wave
Mexico, JJA Implied propagation speed 10
ms-1 ?Shallower gravity wave associated with
land/sea breeze
15
Rapidly propagating squall lines down Bay of
Bengal as observed in JASMINE (Webster et al.
2001)
16
Maritime Continent, DJF Evidence of complex
land/sea breezes which organize convection for
several 100 km
Are sub-gridscale land/sea breezes a crucial
component of the energy and hydrological budgets
of the Maritime Continent?
17
Schematic of a sea breeze

• Sea breeze has two major impacts
• Convergence along sea breeze front provides
  additional convective mass flux
• Winds associated with land and sea breezes
  enhance surface fluxes leading to increased
  moisture supply

18
Investigating Maritime Continent sea breezes
using a mesoscale model (MM5)
Embedded MM5 simulations with Kain-Fritsch
convection scheme. Morning versus evening
precipitation differences show signal over ocean,
indicative of land-sea breezes.
System of propagating land-sea breezes evident in
model. Precipitation is generated over the ocean
during the early morning by the convergence
initiated by the land breeze. Evidence that
orographic effects enhance the land breeze.
19
CONCLUSION I

• Maritime Continent heat source is a key component
  of the global climate. Improvements in its
  simulation may have significant impacts on remote
  systematic errors.
• Specifically, land/sea breezes may be a crucial
  part of the energy and hydrological budgets of
  coastal regions and especially around large
  island complexes.
• In general, horizontally propagating gravity
  waves, generated by convection, may be important
  for organizing convection on larger scales?

20
Typical window brightness (K) images showing
scales of convective organization
Note tendency for cloud clusters to congregate
together to form super-clusters with multi-day
life cycles e.g. Madden Julian Oscillation
Self organization
21
Temporal behaviour of convection around the
equator from window brightness temperature for
Jan.-Feb. 1992
Note evidence of coherent propagation.
22
Space-time spectra showing the organization of
convection in association with theoretical
equatorial waves.
Anti-symmetric
Symmetric
Inertio-gravity
Mixed Rossby-gravity
Inertio-gravity
Kelvin
Rossby
MJO
From Wheeler and Kiladis 1999 J. Atmos. Sci.
23
Space-time spectra from R30 version of GFDL model
Note lack of organization, an error common to
many GCMs. Lack of self-organization mechanism?
24
Spectra of the zonal and meridional wind in the
upper troposphere. Data points show actual
observations from commercial aircraft flights.
Solid curve is for the N270L40 SKYHI model along
the 45N latitude circle and at 211hPa, monthly
averaged for a single July. For clarity the
results for the meridional wind have been shifted
one decade to the right.
From Koshyk and Hamilton 2001 J. Atmos. Sci.
25
July mean spectra as a function of total
horizontal wave-number of (a) the total
KE spectrum, (b) the rotational part of the total
KE spectrum, (c) the divergent part of the total
KE spectrum.
Presence of strong divergent component at
meso-scales consistent with presence of resolved
gravity waves?
From Koshyk and Hamilton 2001 J. Atmos. Sci.
26
GOES Visible Images for 30 September 2001
2245 UTC
2000 UTC
2345 UTC
2130 UTC
27
Florida sea breezes and mesoscale organised
convection
28
Super-parametrizations Results from T21
simulation with an embedded 2-D CRM (1km
resolution) in place of convective
parametrization. Note that the cloud-resolving
models from neighbouring columns interact only
through the large-scale dynamics. Therefore
limits the propagation of gravity waves from one
GCM column to another but does allow gravity
waves to organise convection within the GCM
column. Note dramatic improvement in MJO (upper
panels) and synoptic waves (lower
panels). Courtesy David Randall, CSU
Standard AGCM
AGCM CRM
29
(No Transcript)
30
Why the MJO is important

• Intimately related to active/break cycles of the
  Australian and Asian Monsoons
• Offers potential to provide extended
  predictability up to 15-20 days in tropics
• Affects weather over the western US and possibly
  elsewhere
• Associated westerly wind events generate ocean
  Kelvin waves which may significantly modify the
  evolution and amplitude of El Nino (e.g. 1997)
• Large interannual variability in the activity of
  the MJO has implications for the predictability
  of the coupled ocean-atmosphere system

31
Sensitivity of the MJO to AGCM vertical
resolution(Inness et al., 2001 Clim. Dyn., 17,
777-793.)
32
Model Levels L19 vs. L30
Note additional levels in free troposphere
33
Exploring sensitivity of convective organization
to vertical resolution

• A water-covered or aqua-planet version of the
  UM is used to investigate the behaviour of
  tropical convection when the vertical resolution
  is doubled.
• Aqua-planet version of the UM chosen because
• Homogeneity of the model allows us to obtain a
  large sample of convective events over warm SSTs
• Removal of the land areas excludes circulations
  forced by land-sea contrasts. Convective events
  in different geographical locations are subject
  to the same large scale forcings, giving a
  cleaner comparison.
• Aqua-planet provides a more realistic test of the
  convection scheme than using a single-column
  model with idealized boundary and large scale
  forcing functions.
• Aqua-planet setup
• Zonally symmetric SST distribution, typical of
  equatorial Indian Ocean/West Pacific warm pool
  values.
• Incoming solar radiation fixed at zonally
  symmetric, equinoctial (March) values.
• Aqua-planet model integrated for 15 months with
  both 19 and 30 levels in the vertical. First 3
  months of each integration were discarded.

34
Time-height evolution of convective cloud over
3x3 model grid boxes (7.50 lat. x11.250 long.)
centred on the equator
L30
L19
35
Time-height sections of specific humidity
increment (g/kg/day) by the convection scheme
L30
L19
Note periods of moistening in the L30 case
convection is generally a moisture sink.
36
TOGA-COARE IFA apparent heat source (Q1) and
moisture sink (Q2) for suppressed (A) and active
(B) periods(Lin and Johnson, 1996 J. Atmos.
Sci., 53, 3367-3383)
A Suppressed
B Active
Q1
Q2
Note periods of moistening (negative Q2) during
suppressed period
37
Time-height sections of potential temperature
lapse rate
L30
Observations from TOGA-COARE (Johnson et al.
1999, J. Clim.)
L19
Note presence of stronger stable layer between
600 and 400hPa in L30, and similarity with
observations.
38
Inferences from aqua-planet model results

• When the vertical resolution is increased, the
  spectrum of tropical cloud types changes from a
  bimodal to a tri-modal distribution with a third
  peak in mid-troposphere near the melting level.
  Associated with periods when these mid-level
  congestus clouds are dominant, the detrainment
  from these clouds significantly moistens the
  mid-troposphere.
• The appearance of these congestus clouds is shown
  to be partly due to improved resolution of the
  freezing level and the convective processes
  occurring at this level.
• The resulting cloud distribution more closely
  resembles observations, particularly during the
  suppressed phase of the MJO when cumulus
  congestus is the dominant cloud type.
• The moistening of the free troposphere by cumulus
  congestus clouds acts to precondition the
  atmosphere for deep convection. This
  preconditioning may set the timescale for the
  next active phase of the MJO and thus influence
  the intraseasonal organization of convection.

39
Trimodal distribution of convection and cumulus
congestus
TOGA COARE results emphasize the dominance of
cumulus congestus and point to a TRIMODAL cloud
distribution in which the freezing level
inversion is the key
Many conceptual models of tropical convection are
based on a BIMODALcloud distribution, emphasizing
shallow trade-wind or boundary layer cumuli and
deep cumulonimbi.
From Johnson et al. 1999, J. Clim.
40
CONCLUSION II

• Vertical resolution in the free troposphere must
  be adequate to resolve the formation of the
  freezing level inversion and the cooling
  associated with melting precipitation
• Convective parametrizations need to represent a
  TRIMODAL rather than bimodal cloud distribution.

41
Coupling with the upper ocean Bringing together
the diurnal cycle and the MJO
TOGA-COARE buoy data showed pronounced diurnal
variations in skin temperature in excess of 1K
are evident, as well as slower variations
related to the MJO. Note that the diurnal
variations occur only during break (B) periods.
Active (A) periods are preceded by a warming on
sub-seasonal timescales. (From Anderson et al.
1996, J. Clim. )
B
A
B
Nov.
March
42
Cumulus congestus and the diurnal cycle
TOGA-COARE observations also suggest that cumulus
congestus clouds are most prevalent during light
wind conditions in the presence of a strong
diurnal cycle in SST. Further, these cloud occur
most frequently in the late afternoon suggesting
that they are triggered by the diurnal cycle in
SST.
Coupling with the upper ocean is important on
diurnal timescales
43
The MJO and coupling with the ocean
Observations(Woolnough et al., 2000 J. Clim.,
13, 2086-2104)
Observations show a coherent relationship between
convection and SST. Warm SSTs precede convection
by 5-10 days and are the result of weaker winds,
reduced LH flux and increased SW flux during
suppressed phases of the MJO.
44
The MJO and coupling with the ocean
Modelling(Inness, personal communication)
CGCM has a propagating convective signal
compared with standing oscillation in AGCM.
Coherent variations in SST in CGCM
Coupling with the upper ocean is important for
the MJO
45
BUT intraseasonal SST variations in CGCMs are too
small and the MJO signal is still weak
Is the representation of the upper ocean adequate?
Schematic showing formation of salt barrier layer
Large freshwater flux sets up a salt stratified
barrier layer so that a shallow mixed layer forms
which can respond rapidly to flux variations,
such as the diurnal cycle in solar radiation. The
presence of this barrier layer can potentially
provide much stronger local coupling in the warm
pool region than is currently found in coupled
models which do not resolve the detailed
structure of the warm pool upper ocean.
(From Anderson et al., 1996 J. Clim)
46
Temperature cross-section from the TOGA-COARE
WHOI mooring Note complex temperature structure
in top 40 meters during periods of light winds,
associated with suppressed phase of the MJO and a
strong diurnal cycle.
47
CONCLUSION III

• Good evidence that MJO and diurnal cycle of
  cumulus congestus involve coupling with the upper
  ocean
• To simulate diurnal and intraseasonal variations
  in SST requires detailed representation of
  salinity and temperature gradients in the mixed
  layer
• Need to consider an upper ocean/atmosphere system
  in which the structure of the upper ocean is
  adequately resolved

48
Probability distribution functions (PDF) of
monthly mean SST and precipitation over the
tropical Pacific DJF (upper panels), MAM (lower
panels)
CMAP
HadAM3
HadAM3-CMAP
Note tendency for HadAM3 to overestimate
precipitation over warm SSTs. PDF is also too
tight, following closely the exponential
relationship implied by the Clausius-Clapeyron
equation for saturated vapour pressure.
49
Summary of Key Issues (1)

• Maritime Continent heat source is a major driver
  of the global circulation yet is poorly
  represented in GCMs. This complex system of
  islands gives rise to extensive sea/land breeze
  circulations which may critically influence the
  energy budget and hydrological cycle of the
  region.
• Moistening of the free troposphere by cumulus
  congestus clouds, which form during the
  suppressed phase of the MJO, may be crucial for
  convective preconditioning. This dominant cloud
  type is not represented in models which generally
  fail to capture the observed tri-modal
  distribution of convection.

50
Summary of Key Issues (2)

• Diurnal cycle in SST is large during suppressed
  or light wind conditions in the tropics and may
  be a trigger for cumulus congestus. It may
  therefore be a crucial part of the energy budget
  and hydrological cycle of the Warm Pool
• SSTs vary coherently with the MJO in such a
  manner as to suggest that they are an important
  component of the eastward propagation and
  timescale of the MJO.
• Both the diurnal and intraseasonal variations in
  SST involve detailed changes in the salinity and
  temperature structure of the mixed layer which
  cannot be adequately represented in current
  coupled models.