Dynamics of West African Weather Systems

1
Dynamics of West African Weather Systems

• Luis Garcia-Carreras
• Supervisor Doug Parker

2
Talk Outline

• Introduction
• Diurnal cycle of the boundary layer
• Impacts of vegetation
• Objectives
• Observational Case Studies
• Results
• Advection by night
• Vertical mixing by day
• Conserved-variable diagrams
• Conclusions and Future Work

3
Diurnal Cycle of the Boundary Layer
Stull (1988)
Vertical mixing by day
Horizontal advection by night
4
Impacts of Vegetation?

• Strong land surface coupling in West Africa
• Radiative
• Albedo
• Mechanical
• Roughness length
• Thermodynamic
• Latent and sensible heat fluxes
• Evapotranspiration
• Heterogeneity of the surface

5
Objectives

• What is the balance between mixing and advection
  in the boundary layer over the West African
  region?
• Diurnal cycle.
• The vertical mixing below and into the cloud
  layer, and how it is forced in the mesoscale
• Isoprene (a biogenic volatile organic compound)
  is used as a tracer of the dynamics.
• What is the relationship between vegetation
  patterns and the atmosphere?
• Dynamics and thermodynamics (e.g. water vapour,
  recent rainfall)
• Chemical fluxes, mixing and transport (isoprene,
  O3, etc.)

6
Objectives
OBSERVATIONS - aircraft - remote sensing of
surface - rainfall-radar satellite
UEA flux estimates
How does isoprene relate to vegetation
patterns? Is there evidence of rainfall
constraints on isoprene fluxes? How is it mixed
vertically? How is it transported horizontally?
CEH remote sensing
LARGE EDDY MODEL - simple transport runs -
simple pattern of fluxes - diurnal cycle
UNIFIED MODEL (UM) - operational model
(evaluation) - extension to large/continental/glo
bal scales
Met office
7
Observational Case Studies

• Two test cases B219 (25th July 2006 white
  line) and B235 (17thAugust 2006 white line)

8
Observational Case Studies

• Data collected at varying altitudes (below,
  through, and above the cloud layer)
• Data collected by day and night on each day

9
Advection by Night

• Cross-correlation between day and night-time data
  taken at 1000m
• Isoprene profile is maintained, but has been
  advected 60km east
• Lag is consistent with wind data

10
Vertical Mixing by Day
Correlation 0.57 Number of points 2811

• Isoprene in the cloud layer is characterized by
  sharp peaks profile linked with convective
  activity, not the land surface
• Significant correlations with the vertical wind
  velocity and the liquid water content confirm this

11
Vertical Mixing by Day
B219 1000m
B235 1400m

• Power spectra of the vertical velocities
• Longer length-scale measurements are generally
  unreliable
• Clear presence of turbulent convection (LS
  2-4km) by day, but none by night

12
Day-Time Mesoscale Circulation

• Graph of horizontal winds at 500m and 1400m, and
  isoprene at 1400m.
• Anti-correlation between low level and cloud
  level day-time winds
• Length-scale (30km) and time-scale (30-60
  minutes) inconsistent with updraughts associated
  with turbulent eddies
• Suggests a mesoscale forcing of the updraughts
  relative importance compared to turbulent effects?

13
Conserved-Variable Diagrams

• Need a quantitative analysis of the mixing
• Plot conserved variables for air through the
  cloud vertical profile
• Mixing line shows the source of the air in the
  cloud layer
• Can plot chemical mixing ratios
• Any deviation from the expected mixing line will
  represent non-dynamical losses

Emanuel (1994)
14
Conserved-Variable Diagrams - Examples
15
Conserved-Variable Diagrams - Examples
16
Conclusions

• Convection dominates the day-time dynamics, while
  advection dominates at night
• Air into the cloud layer (altitude 1400m) is
  brought up convectively during the day, and then
  remains largely unvaried as it is advected during
  the night
• By day, there is evidence for mesoscale forcing
  of the updraughts (scales 20-50km), with
  associated day-time horizontal winds, as well as
  turbulent eddies (2-4km)

17
Future Work

• High resolution, cloud-resolving modelling
  covering the same domain as the test-cases (LEM,
  possibly UM in the future)
• Look at the reaction products of isoprene, to
  look at the longer term evolution of plumes
  originating at the land-surface
• Estimate the mixing time-scales of various
  chemical species to infer the rate of chemical
  depletion
• Investigate the impact of the tree-cover on the
  dynamics (e.g. the observed mesoscale
  organization of the updraughts), and see if this
  leads to areas of preferred convection