Bild 1

1
Modelling of transient vegetation and soil
related processes
Patrick Samuelsson Swedish Meteorological and
Hydrological Institute patrick.samuelsson_at_smhi.se
2

• The Rossby Centre Regional Climate Model Land
  Surface Scheme (LSS) (Samuelsson and Gollvik)
• The ECMWF TESSEL LSS (Viterbo et al.)

3
Outline

• Introduction
• Net radiation
• Physiography
• Surface fluxes
• Surface resistances
• The forest tile
• Interception of rain
• Soil heat storage
• Soil properties
• Soil water
• Interception of snow

4
The role of the land surface inNWP/climate models

• Act as a lower boundary for the atmosphere.
• Provide diagnostic values of 2m temperature and
  humidity and 10m wind speed.
• Partitioning between sensible heat and latent
  heat determines soil wetness, acting as one of
  the forcings of low frequency variability (e.g.
  extended drought periods).
• At higher latitudes, soil water only becomes
  available for evaporation after the ground melts.
  The soil thermal balance and the timing of snow
  melt (snow insulates the ground) also controls
  the seasonal cycle of evaporation.
• The outgoing surface fluxes depend on the albedo,
  which in turn depends on snow cover, vegetation
  type and season.

Viterbo, 2004
5
The role of the land surface in NWP/climate
models The water balance components
Precipitation
ERA402.2 mm d-1
ERA40-1.4 mm d-1
Evapotranspiration
Storage of water
Runoff
ERA40-0.9 mm d-1
ERA40 from P. Viterbo
6
Definitions of evaporation
Unstressed evaporation or Potential
evapotranspiration (dry vegetation)
Potential evaporation (wet vegetation)
Field capacity
7
The hydrological rosette(Dooge, 1992)
B-C Soil water has decreased to a level where it
starts to limit the rate of evaporation.
A-B After a long episode of rainfall soil
moisture is available in abundance. The
atmosphere controls the rate of evaporation.
C-D Precipitation refills the soil water by
infiltration.
D-A Maximum soil water level is reached. All
precipitation from this point goes to runoff.
8
The role of the land surface in NWP/climate
models model The energy balance components
Incoming shortwave (S?)
ERA40 NetSW134 Wm-2
EvapotranspirationLatent heat (LE)
ERA40-40 Wm-2
Incoming longwave (L?)
ERA40 NetLW-65 Wm-2
Sensible heat (H)
ERA40-27 Wm-2
Phasechanges
Storage of heat
ERA40 from P. Viterbo
9
Surface net radiation
Albedo Emissivity Surface temperature
Arya, 1988
10
Surface net radiationin the forest
Rnforc
Rnforc
The sky view factor divides the radiationbetween
the canopy and the forest floor
Rnforsn
Rnfors
11
Feedback mechanisms involvingland surface
processes

• Surface evaporative fraction1 (EF), impacting on
  low level cloudiness, impacting on surface
  radiation, impacting on
• Bowen ratio2 (Bo), impacting on cloud base,
  impacting on intensity of convection, impacting
  on soil water, impacting on

P. Viterbo (2004)
12
History of land-surface modelling (Viterbo, 2002)

• Richardsson (1922) In his book on numerical
  weather prediction he identified all the
  principles used by most current LSS.
• Manabe (1969) The bucket model for
  evaporation and runoff.
• Deardorff (1978) introduced the importance of
  vegetation in controlling the evaporation. Many
  of todays LSS are build on these principles.
• Jarvis (1976) described how different stress
  functions affect the stomatal conductance.

13
The mixture contra the tile approach (Koster and
Suarez, 1992)
The Mixture approach
The Tile approach
Coniferous forest
Most schemessomewhere inbetween
Averaged surface properties
Deciduous forest
Lowvegetation
Snow
One value each for parameters like LAI, albedo,
emissivity, aerodynamic resistance, per grid
square. One single energy balance.
All individual sub-surfaces have their own set of
parameters as well as separate energy balances.
14
Physiographic information of tiles ECOCLIMAP
(Masson et al. 2001)
In RCA we have two main land tiles forest and
open land. For snow conditions we also have
forest snow and open-land snow.
Leaf Area Index (LAI) is (projected area of leaf
surface)/(surface area)
15
Diagnostic LAI Hagemann et al. (1999)
LAI as a function of deep soil temperature
Tsoil 4th layer in RCA at 65 cm (unaffected by
diurnal variations)
where
where Tmax and Tmin are 293.0 and 273.0 K,
respectively.
16
The surface energy balance componentsof heat
fluxes in the tile approach
Forest canopy(stomata andinterc. water)
ELatent heat
Low vegetation(stomata andinterc. water)
H Sensible heat
Snow in forest
Snow on openland
Bare soil
Forestfloor
17
Parameterisation of energy fluxes
Sensible heat flux (W m-2)
Latent heat flux (W m-2)
Where ? is air density cp is air heat
capacity ? is latent heat of
vaporisation qs is specific humidity
at saturation
the aerodynamic resistance ra is defined as
18
Land surface atmospherefeedback mechanisms
Experiences from one of the PILPS projects
19
Land surface atmospherefeedback mechanisms
Runoff (-) and evaporation (---) for coupled runs
LSS-RCA atmosphere
Runoff (-) and evaporation (---) for LSS forced
by observations
Z0h z0m
Z0h z0m
Z0h z0m
Z0h z0m
20
The surface energy balance componentsof heat
fluxes in the tile approach
Forest canopy(stomata andinterc. water)
ELatent heat
Low vegetation(stomata andinterc. water)
H Sensible heat
Snow in forest
Snow on openland
Bare soil
Forestfloor
21
The Jarvis approach for thecanopy surface
resistance, rsc
Temperature
Vapour pressure def.
near surface air temperature
near surface vapourpressure def.
f5(Ts) is added in RCA to restrictevapotranspirat
ion when soil is frozen
PAR - Photosyntheticactive radiation
Dickinson et al 1991
22
The Jarvis approach for thecanopy surface
resistance, rsc
? volumetric soil moisture (m3 m-3)
Field capacity, ?d
Wilting point, ?w
0.15
0.30
Soil water availability
Combined with soil depth this gives the water
holding capacity.
Shuttleworth 1993
23
The surface energy balance componentsof heat
fluxes in the tile approach
Forest canopy(stomata andinterc. water)
ELatent heat
Low vegetation(stomata andinterc. water)
H Sensible heat
Snow in forest
Snow on openland
Bare soil
Forestfloor
24
The soil surface resistance rsoil forbare ground
evaporation

• Soil (bare ground) evaporation is due to
• Molecular diffusion from the water in the pores
  of the soil matrix up to the interface soil
  atmosphere (z0q)
• Laminar and turbulent diffusion in the air
  between z0q and screen level height
• All methods are sensitive to the water in the
  first few cm of the soil (where the water vapour
  gradient is large). In very dry conditions, water
  vapour inside the soil becomes dominant

added a restriction due to frozen soil
van den Hurk et al. (2000)Viterbo (2004)
25
The surface energy balance componentsof heat
fluxes in the tile approach
Forest canopy(stomata andinterc. water)
ELatent heat
Low vegetation(stomata andinterc. water)
H Sensible heat
Snow in forest
Snow on openland
Bare soil
Forestfloor
26
The forest tile sensible heat flux
Characterized by low tree heat capacity small
rb
Tam
qam
Tforc
rafor
rs, rb
wcfor
Tfora
qfora
rd
rd
Tforsn
rsoilsc
where Tfora is solved from the relationship
qfora
Tfora
are canopy air temperature and
humidity
27
The forest tile aerodynamic resistances rb and rd
The aerodynamic resistance
Choudhury and Monteith (1988)Sellers et al.
(1986)
The aerodynamic resistance
Choudhury and Monteith (1988)Sellers et al.
(1986, 1996)
rb?10 of rd
28
The forest tile latent heat flux
Characterized by low tree heat capacity small
rb
Tam
qam
Tforc
rafor
rs, rb
wcfor
Tfora
qfora
rd
rd
Tforsn
rsoilsc
where qfora is solved for in a similar manner
asfor Tfora using a balance between latent heat
fluxes
qfora
Tfora
are canopy air temperature and
humidity
29
The forest tileresults
30
The forest tileresults
31
Now all the surface fluxes are known
Forest canopy(stomata andinterc. water)
ELatent heat
Low vegetation(stomata andinterc. water)
H Sensible heat
Snow in forest
Snow on openland
Bare soil
Forestfloor
so we can solve for the storages of heat
(temperatures) and water
32
The storage of heat and waterin the tile approach
T_canopy
Interceptedwater
T_low_veg_and_soil
T_sn
Snow water eq.
Interceptedwater
T_snfor
Liquid water
Snow water eq.
T_for_floor
Liquid water
Surface (0-7 cm) anddeep (7-227 cm) soil water
Five layers in the soildown to three
meters(from 1 to 190 cm thick)
33
Interception of rain

• Interception layer represents the water collected
  by interception of precipitation and dew
  deposition on the canopy leaves (and stems)
• Interception (I) is the amount of precipitation
  (P) collected by the interception layer and
  available for direct (potential) evaporation.
  I/P ranges over 0.15-0.30 in the tropics and
  0.25-0.50 in mid-latitudes.
• Two issues
• Size of the reservoir
• Cl, fraction of a gridbox covered by the
  interception layer
• TP-I Throughfall (T) is precipitation minus
  interception

Viterbo (2004)
34
Interception of rainCanopy water budget
Viterbo (2004)
35
Interception of rainCanopy water budget

• Interception layer for rainfall and dew deposition

Viterbo (2004)
36
Interception of rainresults
37
Back to hvfor

• Total evapotranspiration from canopy

Where the Halstead coefficient is (Noilhan and
Planton,1989)
transpiration interception
Allows transpiration also at maximum interception
reservoir, d1!
Viterbo (2004)
38
Forest temperatures
Characterized by low tree heat capacity small
rb
Tam
qam
Tforc
rafor
rs, rb
wcfor
Tfora
where
qfora
rd
rd
Tforsn
rsoilsc
Cforc defined according to Verseghy et al., (1993)
qfora
Tfora
are canopy air temperature and
humidity
39
The soil
Time scale (very dependent onsoil moisture)
Tsn
Tsnc
zT1
1 hour
z?1
Tssn
Tsc
Tsns
Tscsn
1.0 cm
zT2
1 hour 1 day
6.2 cm
zT3
1 day - 1 week
21.0 cm
z?2
zT4
1 week 1 month
72.0 cm
1 month -
zT5
189.0 cm
No-flux boundary condition at 3 m depth
40
The soil energy equation
In the absence of phase changes, heat conduction
in the soil obeys a Fourier law

• Boundary conditions
• Top Net surface heat flux
• Bottom No heat flux OR prescribed climate

Viterbo (2004)
41
Soil water freezing/thawingViterbo et al. (1999)
Soil heat transfer equation
Viterbo (2004)
42
Numerical solution of the soilenergy equations
Gj-1/2
Tj
Dj
Gj1/2
j1
Viterbo (2004)
43
Temperatures in RCA
44
Soil properties

• The soil is a 3-phase system, consisting of
• Solid minerals and organic matter
• Water trapped in the pores
• Moist air trapped in the pores
• The Texture triangle
• the size distribution of soil particles

Hillel 1982
45
Soil properties
Fractions of clay and sand from ECOCLIMAP (Masson
et al. 2001)
46
Soil properties
? volumetric soil moisture (m3 m-3)
Field capacity, ?d
Wilting point, ?w
Soil porosity
0.45
0.15
0.30
Soil water availability
47
Soil properties
The thermal conductivity
where ? is volumetric soil moisture (m3
m-3) ?sat is total porosity (m3 m-3) a is an
empirical parameter ?sat is matric potential at
saturation (m) b is Clapp and Hornberger
parameter
Rosenberg et al 1983
48
Soil water flux
Soil water flux is usually expressed by Richards
equation
where ? is volumetric soil moisture (m3 m-3) ?
hydraulic diffusivity (m2 s-1) ? hydraulic
conductivity (m s-1) S source/sink term
(precipitation, through fall,
snowmelt, evapotranspiration
by root extraction)
49
Soil water flux
Hydraulic diffusivity and conductivity
Mahrt and Pan 1984
50
Soil water flux
In RCA the 2nd term on the rhs is replaced by the
ß formulation
where ? is volumetric soil moisture (m3
m-3) ?wi is wilting point ?fc is field
capacity Sdr(?,z0) is precipitation, through
fall, snowmelt Sdr(?,z1) is root extraction and
drainage Sdr(?,z2) is root extraction and runoff
51
Snow interception
52
Why Include Interception of Snow?

• Intercepted snow feels much less aerodynamic
  resistance than forest floor snow.
• 25-45 of an annual snowfall can evaporate from
  intercepted snow (Pomeroy et al. 1998).
• Affects evaporation/runoff partition.

53
Interception of Snow
Change of intercepted snow

• SOURCES
• Snow interception, SI (m/s)
• Intercepted water friezes, wcfor (m)
• Sublimation of water vapor, E/?w (m/s)

• SINKS
• Evaporation of snow, E/?w rb?10 of rd
• Snow unloading, UL (m/s)

54
Snow Interception Model
The snow interception (SI) and snow unloading
(UL) part of the model is based on Hedstrom and
Pomeroy (1998)
where SNcfor,max f(LAI, 1/?sn(Tc)) 20 mm
k (snow-leaf contact area) /
SNcfor,max PSN snowfall
where U a constant unloading rate coefficient
(SNcfor is put to zero for
Tcgt0ºC)
55
Snow Interception Model
The snow sublimation (E/?w) part of the model is
parameterized as
where ?a air density q specific
humidity rb aerodynamic resistance
ßs evaporative efficiency
(modified from Nakai et al. 1999)
ßs
SNcfor / SNcfor,max
56
RCA simulation
Boundaries ERA-15, Res. 20 km, dt15
min.Accumulated results Sep 1996 - May 1997
Simulated seasonal intercepted snow evaporation
(mm)
Simulated intercepted(snow evaporation)/snow ()
57
RCA simulation and observations
RCA Sodankylänorthern Finland Other studies (observations)
Snow interception of seas. snowfallmax durationduration gt 1 day 7040 days20 events Obs. durations from days up to weeks (Bründl et al. 1997)
Max daily interc. snow evap. 75 W m-2 2.5 mm day-1 1.3 3.9 mm day-1(Lundberg Halldin, 2001)
Mean interc. snow evap.Seasonal 13 W m-20.44 mm day-125 10 50 (LH, 2001)
58
Conclusions about snow interception

• The presented parameterization of snow
  interception gives reasonable results compared to
  many studies but does not perform well according
  to eddy-correlation measurements in Sodankylä.
• As stated by Lundberg and Halldin (2001) the
  evaporation is very sensitive to the aerodynamic
  resistance.
• To improve these preliminary model results we
  need better physiographic description (LAI,
  forest structure) and we also need more
  observations to be able to validate the results.