Transcript Land-Atmosphere Interactions and Importance to Sahel Precipitation

Land-Atmosphere Interactions
and Sahel Precipitation
Andrea M. Sealy
ASP/CGD
Advanced Study Program Research Review
March 29th, 2007
Outline
• Background
– Sahel rainfall climatology
– Land-atmosphere interactions
• Review of previous studies
– Land-atmosphere coupling
• Soil moisture-rainfall feedback
• Previous studies in context of current Work
– Land surface impacts on Sahel precipitation and African easterly waves
• Review
• Objectives
• Proposed analyses
– Desert dust impacts on Sahel precipitation
• Review
• Objectives
• Proposed analyses
3/29/2007
2
Background
•
•
•
•
•
Sea surface temperature and its impact on seasonal variability and
predictability of precipitation has been focus of climate studies
Shukla et al. (2002) found many regions have strong response to anomalous
sea surface temperature (SST) such as El Niño/Southern Oscillation
phenomenon
West African precipitation suggested to be linked to Gulf of Guinea/Tropical
Atlantic (Eltahir and Gong, 1996; Vizy and Cook, 2000) and Indian Ocean
SSTs (Giannini et al., 2003)
Other factors such as land state variables (soil moisture, vegetation cover,
albedo, dust) may also contribute to seasonal precipitation variability in the
Sahel
Comprehensive understanding of the feedbacks between land and
atmosphere is yet to be reached
– Observational data of surface and sub-surface properties are often very scarce
(e.g., for soil moisture an observation network over large areas is lacking)
– Numerical results may differ and are model dependent
3/29/2007
3
West Africa and Sahel Rainfall Climatology
8
7
6
5
mm/day 4
West Africa
Sahel
3
2
1
0
Jan Feb Mar
Apr May Jun
Jul
Aug Sep Oct Nov Dec
Source: Legates and Wilmott (1990); 1920-1980 gridded precipitation estimates
Soil moisture-rainfall feedback (Eltahir, 1998; Eltahir and Pal, 2001)
Increase soil moisture
Decrease surface albedo
Decrease ratio of sensible
to latent heat
Increase lower level
water vapor concentration
Increase net surface
shortwave radiation
Decrease ground and
surface temperature
Increase net surface
longwave radiation
Increase total latent and
sensible heat flux
Increase lower level moist static energy
Decrease stability
Increase frequency and magnitude of local convective rainfall
Indirect soil moisture ― precipitation feedback (from Dave Lawrence, CGD)
Theory developed in Betts and Ball (1995), Betts et al. (1996), Eltahir (1997), and Schär et al.
(1999) supported by observations from FIFE, 1-d models, and regional climate models.
Over wet soil:
• enhanced evaporation  lower Bowen ratio  shallower and wetter boundary layer
• darker soil (α ) and cooler surface temperatures  enhanced net surface radiation 
larger total heat flux into boundary layer
• two factors combine to increase Moist Static Energy per unit mass of Boundary Layer air
LH SH
SW LW RNET
dry, warm, bright soil
3/29/2007
MSE /
m3 BL air
LH SH
SW LW RNET
wet , cool, dark soil
MSE /
m3 BL air
7
Land-atmosphere coupling
• Land-atmosphere coupling strength: the degree to which the
atmosphere responds to anomalies in land surface state
• (Koster et al., 2004) Global Land-Atmosphere Coupling
Experiment (GLACE): An inter-comparison study across a
range of atmospheric general circulation models
• Regions with significant land-atmosphere coupling are identified
from multi-model average (including West Africa)
• These hot spots indicate where greater monitoring of soil
moisture could yield the greatest return in seasonal
forecasting
• Results show a broad disparity in the inherent precipitation
responses of the different models
• NCAR’s Community Atmosphere Model (CAM3) showed high
land-atmosphere coupling strength
3/29/2007
8
Koster (2004) shows the land-atmosphere coupling strength diagnostic
for northern hemisphere summer.
Land surface impact on diurnal
cycle and easterly waves
• Taylor and Clark, 2001
– Met Office Hadley Centre Unified Model (HaDAM3)
– SPARSE vegetation (which is more realistic for Sahel
region)
•
•
•
•
warmer and deeper planetary boundary layer
weaker diurnal cycle of precipitation
enhanced daily variability of precipitation
greater easterly wave activity
– Results illustrate close coupling between land surface
and atmosphere
3/29/2007
10
Precip
Obs
V850
Obs
Precip
CAM3
V850
CAM3
Precip
HadAM3
V700
HadAM3
African easterly
waves – Sahel
3 – 5 day period
Does strong SM-P
feedback (strong
dependence of
convection on
surface fluxes) in
CAM3 get in the way
of precipitation
response to AEWs?
Source: David
Lawrence, CGD/CCR
3/29/2007
Taylor and Clark, 2001
11
How does this relate to current work?
• NCAR’s Community Atmosphere Model (CAM3) exhibits greater
land-atmosphere coupling than Hadley Centre model (from Koster et
al., 2004)
• How is easterly wave behavior influenced by land surface
conditions?
• What connection should be investigated?
• Fluxes from land surface into atmosphere and how it affects
boundary layer (smaller evaporation rates, warmer and deeper
boundary layer, weaker diurnal rainfall cycle, greater AEW activity,
more long lived rain events, Taylor and Clark 2001)
• Main parameter to be changed and why?
– Soil moisture (gradient), affects displacement/location, magnitude of
AEJ which creates the environment for AEWs to develop (Cook, 1999)
– Vegetation?
3/29/2007
12
Impact of dust radiative forcing
• Yoshioka et al., submitted to Journal of Climate
• Community Atmosphere Model (CAM3)
• Model of Atmospheric Transport and Chemistry
(MATCH)
• Radiative forcing of dust acts to reduce average
precipitation
• More significant for interactive SST (Slab Ocean
Model) than observed SST (Atmospheric Model
Intercomparison Project) runs
3/29/2007
13
Difference between AMIP with shortwave and longwave dust feedback and no dust feedback
3/29/2007
14
Difference between SOM with shortwave and longwave dust feedback and no dust feedback
3/29/2007
15
How does this relate to current work?
•
•
To investigate and compare the impact of dust radiative forcing, sea surface
temperature forcing and (dynamic) vegetation on Sahel precipitation
Community Atmosphere Model (CAM3) coupled with Dynamic Global Vegetation
Model (DGVM)
–
–
•
•
•
•
•
AMIP with no dust (AMIPndDV)
SOM with no dust (SOMndDV)
AMIP with dust feedback (AMIPDV)
•
SOM with dust feedback (SOMDV)
Analyze and validate the rainfall signal in terms of amount/magnitude, geographical
distribution, seasonal distribution and compare to observations
Analyze dust optical depth, geographical distribution, shortwave and longwave forcing
and net radiative (shortwave + longwave) forcing
–
•
Fifty year simulations, forced either by observed sea surface temperatures (Atmospheric
Model Intercomparison Project/AMIP) or interactive SST using Slab Ocean Model (SOM).
Simulations (with DGVM) that will be used for analysis
compare to previous studies and any differences explained based on model and dust
parameterization used in the respective studies
Examine differences between the dust feedback and no dust simulations’ precipitation
–
–
3/29/2007
differences in shortwave and longwave radiative forcing and near surface temperature
impact of dynamic vegetation we could compare the DGVM runs to (Yoshioka et al,
submitted to J. Climate) runs done without DGVM that use default CAM vegetation
16
3/29/2007
17