Cascade

1
Cascade

• Scale interactions and organized convection in
  the tropical atmosphere
• Project Manager Steve Woolnough

2
Outline

• Motivation
• Examples of scale interactions in organized
  tropical convection
• Scientific and Technical Challenges
• Cascade Project
• The Science of Cascade
• Project Structure

3
Organization of tropical convection

• Tropical convection is organized across a wide
  range of scales associated with
• Individual cloud systems
• Squall lines and mesoscale convective complexes
• African Easterly Waves, Tropical Cyclones,
  Equatorial Waves
• Madden-Julian Oscillation, Monsoon circulations

4
Interactions between space and time scales of
tropical convection
5
Mean Climate over the Maritime Continent

• Precipitation over the Maritime Continent plays a
  dominant role in driving the circulation over the
  Indo-Pacific Warm Pool
• Modelled precipitation shows dry bias around the
  islands of the Maritime Continent
• Large wet biases in Western Indian Ocean and West
  Pacific
• Wet biases over the islands of the Maritime
  Continent

CMAP
HadGAM1
Difference
From Jane Strachan (Walker Institute)
6
Diurnal Cycle in the Maritime Continent

• Observations from TRMM
• Strong diurnal cycle in precipitation
• Little or no precipitation over land during
  morning
• Precipitation develops over land in late
  afternoon and into the late evening
• Precipitation moves out over ocean during early
  hours of the morning
• Convection in the Maritime Continent is strongly
  influenced by diurnal cycle and land-sea breezes

From Jane Strachan (Walker Institute)
7
Diurnal Cycle of Precipitation in HadGAM1

• Maximum precipitation over land during morning
• Precipitation dies out over land in late
  afternoon
• Little or no precipitation over seas around
  Maritime Continent
• Do weaknesses in the representation of key
  physical phenomena in the model impact the mean
  climate of the region?

1230LT
1830LT
0030LT
0630LT
From Jane Strachan (Walker Institute)
8
Organization of convection by equatorial waves

• Observations (e.g. Wheeler and Kiladis 1999, Yang
  et al. 2003) show convection is organized by
  large-scale equatorial waves

from Wheeler and Kiladis (J. Atmos. Sci, 1999)
9
Equatorial Waves in HadGAM
Observed
HadGAM1

• By what processes are the convection and dynamics
  coupled?
• How does the model represent these processes?

10
MJO in Climate Models

• Lin et al. (J. Clim, 2006) (CGCMs)
• MJO variance approaches observed value in only 2
  of 14 models
• Ratio of eastward to westward variance is too
  small, consistent with lack of coherent eastward
  propagation
• Variance in 13 of 14 models not associated with
  pronounced spectral peak.

11
Multiscale Organization in the MJO

• Eastward propagating envelope of convection
• Short lived, westward propagating systems within
  envelope
• With individual mesoscale cloud systems embedded
  within them

from Chen et al. (J. Atmos. Sci, 1996)
from Rickenbach Rutledge (J. Atmos. Sci, 1998)
12
Multiscale Organization in the MJO

• What role do the different scales of organization
  play in the transports of heat, moisture and
  momentum by the convection?
• How are these processes represented within
  conventional parametrization schemes?

13
Organized Convection in AEWs
12Z

• Complex organization on synoptic and meso-scales
• Strong temporal development linked to diurnal
  cycle
• Triggering of convection by orography and
  convergence along cold pools

15Z
18Z
14
Scientific Technical Challenges

• How does convection organise itself across a
  range of time and space scales?
• How are energy and momentum transferred?
• Why do preferred structures emerge?
• How can these be represented in global models?
• But we cannot answer these questions using
  observations - so we must use models as pseudo
  field experiments.
• This requires large domains and very high
  resolution so that the energy spectrum is not
  compromised.
• Needs very significant computational resource,
  only just becoming available.

15
Cascade Project

• Natural Environmental Research Council (NERC)
  Funded Consortium Project
• Walker Institute, University of Reading
• National Centre for Atmospheric Science
• University of Leeds
• University of East Anglia
• Met Office
• 3½ year project which started on 1st October 2007
  

16
Proposal

• Develop modelling framework for large domain
  (100? longitude by 30? latitude), high
  resolution (1-2 km) simulations.
• Perform numerical case studies of organised
  convection over West Africa and the Indo-Pacific
  Warm Pool.
• Evaluate these simulations using advanced
  satellite and in situ observations of cloud
  structures.
• Analyse the simulations in terms of
  scale-dependent energy and momentum budgets.
• Use idealised case studies to explore the links
  between convection and equatorial wave modes.
• Bring this new understanding into the development
  of new representations of tropical convection for
  global weather and climate models.

17
Project Structure

• WP1 Development of Modelling Tools and
  Infrastructure
• WP2 Scale interactions in African Weather
  Systems
• WP3 Organized Convection over the Indian Ocean
  and West Pacific Warm Pool
• WP4 Model Evaluation against Observations
• WP5 Synthesis of Results

18
Development of Modelling Tools and Infrastructure

• Model Development (Met Office)
• Met Office Unified Model
• Build on existing work in the Met Office to
  develop 1km resolution LAM for use as forecasting
  tool and CRM research tool
• Includes options for
• 3D turbulence scheme
• 5 phase microphysics scheme
• Test small domain case studies
• Modelling Infrastructure (Lois Steenman Clark,
  NCAS CMS)
• Optimatization
• Scalability
• Performance of I/O
• Management of TB of data

19
Simulations of the diurnal cycle and land-sea
breezes over the Tiwi Islands in the Timor Sea
Evidence of modelling capability Realistic
physical-dynamical coupling in 1 km version of
the Unified Model
20
Simulations of the diurnal cycle and land-sea
breezes over the Tiwi Islands in the Timor Sea
Evidence of model skill Good agreement with
radar observations
21
Scale Interactions in African Weather Systems

• Diurnal Cycle of Convection (Tony Slingo, Walker
  Institute)
• What mechanisms control the organization of
  convection by the diurnal cycle over varying land
  surface conditions and forcing regimes typical of
  North Africa?
• MCSs and AEWs (Doug Parker, U of Leeds)
• What is the role of MCSs, particular those forced
  by the diurnal cycle, in the dynamics of the
  regional and synoptic waves and vortices over
  Africa?

22
Organized Convection in the Indo-Pacific Warm
Pool

• Scale Interactions in the MJO (Steve Woolnough,
  Walker Institute)
• How do the vertical heating profiles and momentum
  transports associated with convection influence
  the evolution of the MJO?
• The diurnal cycle and the climate of the Maritime
  Continent (Adrian Matthews, UEA)
• What role do the diurnal cycle and land-sea
  breezes play in determining the mean climate of
  the Maritime Continent?

23
Organized Convection in the Warm Pool

• Idealized Simulations of organized convection and
  equatorial waves (Brian Hoskins, Mike Blackburn,
  Walker Institute)
• How is convection organized by equatorial waves
  and how does it modify the structures predicted
  by the dry theory?

24
Case Study Methodology

• Use the high resolution simulations as
  pseudo-observations to investigate
• the energy and momentum budgets associated with
  organized convection
• the interactions between the various scales of
  organization
• Compare with low resolution simulations to
  identify weaknesses and missing processes in
  convection parametrization schemes

25
Model Evaluation against Advanced Observations

• (Robin Hogan,
  Walker Institute)
• Use geostationary Satellites to evaluate temporal
  development of organized cloud clusters
• Use CloudSat, CALIPSO, and surface ARM sites to
  evaluate vertical structure of clouds

26
Synthesis of Results

• 
  (Met Office)
• Review the results from a parametrization
  perspective including
• role of organized convection
• Environmental influences on organization
• Temporal evolution of convection
• Test new parametrization ideas in low resolution
  simulations of case studies

27
Project Timeline