features and performance of the ncar community land model (clm): permafrost, snow, and hydrology...
TRANSCRIPT
Features and performance of the
NCAR Community Land Model (CLM):
Permafrost, snow, and hydrology
David Lawrence
NCAR / CGD Boulder, CO
Components of CLM
T2, T1,
T3,
T10,
Photosynthesis model
Soil thermodynami
c and hydrology
model…
Snow Model
Calculates the surface energy, water, momentum, (and carbon) fluxes
RiverTransportModel
Community Land Model Subgrid Structure
Gridcell
Glacier Wetland Lake
Landunits
Columns
Plant Functional Types
UrbanVegetated
Soil Type 1
15 total PFTs
Snow model
• 5-layer model
• Treats processes such as accumulation, melt, compaction, water transfer across layers, snow aging
• State variables are mass of liquid and ice water, snow temperature, and layer thickness
• Near-term improvements:
– Vertical distribution of solar heating
– Improved snow cover fraction and snow burial fraction
– Improved snow aging
– Aerosol deposition
Soil hydrology
No perpetually frozen layer
At least one perpetually frozen soil layer
• Vertical redistribution -Richard’s equation
• TOPMODEL-based surface runoff
• Coexistence of liquid and ice water at T < 0oC
• Dynamic water table position controls sub-surface runoff
• Groundwater model (unconfined aquifer)
so
il d
ep
th (
m)
Soil thermodynamics
• Solution of heat diffusion equation
• Thermal properties (conductivity, heat capacity) function of soil texture (sand, silt, clay) and soil liquid and ice water contents
• Recent improvements
– Organic soil
– Deeper soil column
CLM soil carbon density datasetSource data from Global Soil Data Task
Kg m-2
Lawrence and Slater, 2007
Calculate thermal and hydraulic soil parameters as a weighted combination of organic and mineral soil values
) + fsc,i Θsat,sc
(1 - fsc,i) (
Θsat,i = 0.489 − 0.00126 %sandi
Thermal and hydraulic parameters for organic soil
Soil type λsat λdry Θsat ksat
Sand 3.12 0.27 0.37 0.023
Clay 1.78 0.20 0.46 0.002
Peat 0.55 0.05a 0.9a,b 0.100b
a Farouki (1981), b Letts et al. (2000)
λsat sat. thermal conductivity
λdry dry thermal conductivity
Θsat volumetric water at saturation
ksat sat. hydraulic conductivity
fsc,I = ρsc,i / ρpeat fraction of layer i that is organic matter
Impact on soil temperature (SOILCARB – CONTROL)
Deeper soil column
Motivation:
account for thermal inertia of deep soil layers
Solution:
add additional layers, hydrologically inactive
3.5m10 levs
~50-125m15-17 levs
Ignore geothermal heat flux (up to ~0.1 W m-2)
Annual cycle-depth soil temperature plotsSiberia
SOILCARB + 50m Soil Column
Near-surface permafrost extent and active layer thickness (1970-1990)
8.5 million km2 10.7 million km2
11.2 to13.5 million km2 observedContinuous + discontinuous
SOIL CARBON + DEEP SOIL
Other features of CLM
• Other Features
– Terrestrial carbon-nitrogen cycle model
– Dynamic global vegetation model
– Can be run at resolution independent of atmos resolution
– (Greenland ice sheet model)
– Broad community participation in model development (contributions from CCSM Land and Biogeochemistry Working Groups)
Current limitations of CLM specific to Arctic processes
• Limitations
– No excess soil ice
– No thermokarst, subsidence process
– Static wetlands, lakes
– Dynamic vegetation does not include shrubs
– No methane emission model
– Sporadic, isolated permafrost not explicitly represented
Summary: Towards simulating Arctic terrestrial high-latitude feedbacks in CCSM
Microbial activity
increases
Enhanced[nitrogen]
Shrubgrowth
CO2
efflux
Lakes drain, soil dries
Globalwarming
CH4
efflux
Expandedwetlands
Carbonsequester
Permafrost warms and
thaws
ArcticArcticwarmingwarming
Summary: Towards simulating Arctic terrestrial feedbacks in an Earth System Model
Microbial activity
increases
Enhanced[nitrogen]
Shrubgrowth
CO2
efflux
Lakes drain, soil dries
Globalwarming
CH4
efflux
Expandedwetlands
Carbonsequester
Permafrost warms and
thaws
ArcticArcticwarmingwarming
TOPMODEL basedapproach permits dynamic wetlands?Thermokarst?
CLM-CN: simulates large Arctic soil carbon store
CLM-DGVM: currently no shrub type
Organic soil and deep soil column improve permafrost simulation
Large-scale degradation of near-surface permafrost likely
Wetland-CH4 emission model
Peatland soil type and/or moss vegetation type?
Near-surface permafrost degradation
Lawrence and Slater, 2007b
Observed estimatecontinuous + discontinuous
Deep permafrost (10-30m)
Most deep
permafrost still
exists at the end of
the 21st century
Soil carbon in CLM-Carbon/nitrogen (CLM-CN)
Soil carbon content: CLM-CNSoil carbon content: Obs
Thermal and hydraulic parameters for organic soil
Soil type λs
W m-1 K-1
λsat
W m-1 K-1
λdry
W m-1 K-1
cs
J m-3 K-1x106
Θsat ksat
m s-1x10-3
Ψsat
mm
b
Sand 8.61 3.12 0.27 2.14 0.37 0.023 -47.3 3.4
Clay 4.54 1.78 0.20 2.31 0.46 0.002 -633.0 12.1
Peat 0.25a 0.55 0.05a 2.5a 0.9a,b 0.100b -10.3b 2.7b
a Farouki (1981), b Letts et al. (2000)
λs soil solid thermal conductivityλsat saturated thermal conductivityλdry dry thermal conducitivitycs heat capacity
Θsat volumetric water at saturationksat saturated hydraulic conductivityΨsat saturated matric potentialb Clapp and Hornberger parameter
Impacts on climate
CAM-CLM
CLM
CCSM3 projections of degradation of near-surface permafrost
We define near-surface permafrost as perpetually frozen soil in upper 3.5m of soil
Lawrence and Slater, 2005