Modelling Land Surface in a climate model E. Kowalczyk CSIRO Marine and Atmospheric Research Cape GrimTumbarumba

Role of the land surface schemes (LSS) in climate models Surface energy and water balance in climate model Representation of vegetation processes in CABLE Description of land hydrology and soil temperature - soil moisture & temperature - snow accumulation, melting and properties Examples of use of CABLE coupled to a climate model Outline

Fast biophysical processes Canopy conductance photosynthesis, leaf respiration Carbon transfer, Soil temp. & moisture availibity Slow biogeographical processes Vegetation dynamics & disturbance Land-use and land-cover change Vegetation change Autotrophic and Heterotrophic respiration Allocation Intermediate timescale biogeochemical processes Phenology Turnover Nutrient cycle Solution of SEB; canopy and ground temperatures and fluxes Soil heat and moisture Surface water balance Update LAI, Photosyn- thesis capacity Physical- chemical forcing T,u,Pr,q, Rs,Rl, CO2 Radiation water, heat, & CO2 fluxes daysyears Biogeo- chemical forcing Time scale of biosphere-atmosphere interactions Atmosphere minutes

S net + L net – G = H + λE TfTf TgTg LSS calculates exchanges of moisture, energy, momentum and trace gasses at the land-atmosphere interface. Land surface important characteristics for calculation of SEB: albedo, leaf area index, canopy height, surface moisture. Key task is to calculate Surface Energy Balance: Role of the Land Surface Scheme (LSS) in GCM H λE L S

Surface Water Balance in Climate Model P rec – E vap – R unoff = Δ S now + Δ S oilMoist Land surface important characteristics: soil hydraulic properties & depth vegetation properties; rooting depth leaf area index, max carboxylation rate

Interface to GCM or offline Canopy radiation; sunlit & shaded visible & near infra-red, albedo stomata transp. & photosynthesis Carbon fluxes; GPP, NPP, NEP SEB & fluxes; for soil-vegetation system: E f, H f, E g, H g; evapotranspiration soil moisture snow carbon pools; allocation & flow The general structure of CABLE soil temp.soil respiration Visit the CABLE secured website with your supplied password at Kowalczyk et al., CMAR Research Paper 013,

The main features of CABLE a coupled model of stomatal conductance, photosynthesis and the partitioning of absorbed net radiation into latent and sensible heat fluxes the model differentiates between sunlit and shaded leaves i.e. two-big-leaf sub-models for calculation of photosynthesis, conductance and leaf temperature the radiation submodel calculates the absorption of beam and diffuse radiation in visible and near infrared wavebands, and thermal radiation the vegetation is placed above the ground allowing for full aerodynamic and radiative interaction between vegetation and the ground the plant turbulence model by Raupach et al. (1997) a multilayer soil model is used; Richards equations are solved for soil moisture and heat conduction equation for soil temperature the snow model computes temperature, density and thickness of three snowpack layers. biogeochemical model CASA CNP for carbon, nitrogen and phosphorus including symbiotic nitrogen fixation ( Wang, Houlton and Field,2007).

Role of the land surface schemes (LSS) in climate models Surface energy and water balance in climate model Representation of vegetation processes in CABLE Description of land hydrology and soil temperature - soil moisture & temperature - snow accumulation, melting and properties Examples of use of CABLE coupled to a climate model Outline Representation of vegetation processes in CABLE

Canopy representation CABLE

Coupled model of stomatal conductance and photosynthesis The two-leaf model ( sunlit & shaded ) of Wang & Leuning [1998] is used to calculate 6 variables: Tf - leaf temperature Ds - vapour pressure deficit Cs - CO2 concentration at the leaf surface Ci - intercellular CO2 concentration of the leaf Gs - stomatal conducatnce An - net photosynthesis The set of six equations is used to solve simultaneously for photosynthesis, transpiration, leaf temperature and sensible heat fluxes for a each leaf

Vegetation parameters required for CABLE Geographically explicit data LAI – leaf area index fractional cover C3/C4 - fraction the model calculates: z0 – roughness length α – canopy albedo VEGETATION TYPE 1 broad-leaf evergreeen trees 2 broad-leaf deciduous trees 3 broad-leaf and needle-leaf trees 4 needle-leaf evergreen trees 5 needle-leaf deciduous trees 6 broad-leaf trees with ground cover /short-vegetation/C4 grass (savanna) 7 perennial grasslands 8 broad-leaf shrubs with grassland 9 broad-leaf shrubs with bare soil 10 tundra 11 bare soil and desert 12 agricultural/c3 grassland 13 ice A grouping of species that show close similarities in their response to environmental control have common properties such as: - vegetation height - root distribution - max carboxylation rate - leaf dimension and angle, sheltering factor, - leaf interception capacity

Role of the land surface schemes (LSS) in climate models Surface energy and water balance in climate model Representation of vegetation processes in CABLE Description of land hydrology and soil temperature - soil moisture & temperature - snow accumulation, melting and properties Examples of use of CABLE coupled to a climate model Outline

Multilayer soil model Z 1 =0.02m ZNZN Z N-1 Z1Z1 Z2Z2 Z3Z3 ZNZN ground heat sensible Net Solar + Net Long wave evap Thickness of soil layers (m)

Soil moisture model ZNZN Z1Z1 Z2Z2 Z3Z3 Z N-1 drainage Surface runoff calculated as saturation excess ( + effects of topography if coupled to a climate model) precipitation + snow melt Drainage calculated as excess of soil field capacity or gravitational drainage plant ET surf runoff soil evap Soil moisture is calculated from the solution of Richard’s equation. The assumed form of relationship between the hydraulic conductivity, matric potential and the soil moisture is that of Clapp and Hornberger (1978).

Saturation Saturation: water fills in all available pore space Soil Moisture: some terms and concepts Available Water Available Water: amount of water in the soil between the field capacity and the permanent wilting percentage Field Capacity Field Capacity: water that remains in soil beyond the effects of gravity. Permanent Wilting Permanent Wilting: amount of water after the permanent wilting point is reached Soil moisture Soil moisture : quantity of water in soil, θ = V water / V soil Є ( 0, 0.5 )

Soil parameters required for CABLE Soil types: Coarse sand/Loamy sand Medium clay loam/silty clay loam/silt loam Fine clay Coarse-medium sandy loam/loam Coarse-fine sandy clay Medium-fine silty clay Coarse-medium-fine sandy clay loam Organic peat Permanent ice Soil Properties: - water balance: saturation wilting point field capacity hydraulic cond. at saturation matric potential at saturation - heat storage: albedo, specific heat, thermal conductivity density - soil depth Post, W., and L. Zobler, 2000 Global Soil Types

Variation of hydraulic conductivity with water potential K wet dry

The soil parameters used in the CSIRO climate models. Soil types: (1) Coarse sand/Loamy sand (5) Coarse-fine sandy clay (2) Medium clay loam/silty clay loam/silt loam (6) Medium-fine silty clay (3) Fine clay (7) Coarse-medium-fine sandy clay loam (4) Coarse-medium sandy loam/loam (8) Organic peat (9) Permanent ice SOIL Type 1 Type 2 Type 3 Type 4 Type 5 Type 6 Type 7 Type 8 density soil density kg/m3 sfc field capacity (m3/m3) swilt wilting point (m3/m3) ssat saturation (m3/m3) hyds* hydraulic cond. at saturation (m/s) sucs matric potential at saturation bch b parameter in Clapp-Hornberger relations clay fraction of clay sand fraction of sand silt fraction of silt css soil specific heat (kJ/kg/K) dry soil thermal conductivity is calculated as: sand*0.3 + clay* silt*0.265 [W/m/K] Thickness of soil layers (m)

Role of the land surface schemes (LSS) in climate models Surface energy and water balance in climate model Representation of vegetation processes in CABLE Description of land hydrology and soil temperature - soil moisture & temperature - snow accumulation, melting and properties Examples of use of CABLE coupled to a climate model Outline

Snow modelling

Modelling of snow evolution Snow - properties - high albedo - good thermal insulator - density increases with time -Snow accumulation -Snow albedo -Snow metamorphism and thermal properties -Snow cover interaction with vegetation -Snow melting

Modelling of snow evolution Snow - properties - high albedo - good thermal insulator - density increases with time Snow state variables: - temperature - density - age - mass Snow diagnostic variables: - snow albedo - depth - effective conductivity

Snow-Free Spatially Complete Product January 2002, 0.86µm Overlaying the Snow Albedo Statistics onto the Snow-Free Spatially Complete Albedo Using NISE Snow Extent and Type to Overlay the Snow Albedo Statistics Crystal Schaaf, Boston University)

The Moderate Resolution Imaging Spectroradiometer (MODIS), flying aboard NASA’s Terra and Aqua satellites, measures snow cover over the entire globe every day, cloud cover permitting. The image shows snow cover (white pixels) across North America from February 2-9, 2002.

SNOWMIP I Col de Porte CSIROobservations

Role of the land surface schemes (LSS) in climate models Surface energy and water balance in climate model Representation of vegetation processes in CABLE Description of land hydrology and soil temperature - soil moisture & temperature - snow accumulation, melting and properties Examples of use of CABLE coupled to a climate model for C4MIP phase one study. Outline

Fast biophysical processes Canopy conductance photosynthesis, leaf respiration Carbon transfer, Soil temp. & moisture availibity Slow biogeographical processes Vegetation dynamics & disturbance Land-use and land-cover change Vegetation change Autotrophic and Heterotrophic respiration Allocation Intermediate timescale biogeochemical processes Phenology Turnover Nutrient cycle Solution of SEB; canopy and ground temperatures and fluxes Soil heat and moisture Surface water balance Update LAI, Photosyn- thesis capacity Physical- chemical forcing T,u,Pr,q, Rs,Rl, CO2 Radiation water, heat, & CO2 fluxes daysyears Biogeo- chemical forcing Time scale of biosphere-atmosphere interactions Atmosphere minutes

Negative feedback Neutral Positive feedback Major regulatory mechanisms that lead to either positive or negative feedbacks of C cycle to climate warming PhotosynthesisRespiration Nutrient availability Decomposition Length of growing seasons Drought Warming- or nutrient prone species Stress-tolerant species diminishing acclimation acclimation Luo Annu. Rev. Ecol. Evol Increased evapotranspiration

CSIRO Carbon-climate simulation C4MIP phase I simulation: –Coupled CABLE (CSIRO Atmosphere Biosphere Land Exchange LSS) with CCAM (Cubic Conformal Atmospheric Model). –Used prescribed SST, carbon fluxes from ocean, fossil fuel and land use change from 1900 to –Two atmospheric CO 2 concentrations used: 1) prescribed historical CO2 globally uniform, 2) a result of atmospheric transport of all carbon fluxes including biospheric fluxes. –Two simulations: RUN1: biosphere sees prescribed historical CO2 from 1900 to 2000 RUN2: biosphere sees prescribed historical CO2 from 1900 to 1970, then CO2 is kept constant at 1970 level from 1971 to Law, Kowalczyk & Wang, Tellus, 58B, , 2006.

C-CAM CABLE photosynthesis Fossil Fuel CO 2 emissions fluxes Stem Roots Soil Carbon CO 2 release CO 2 uptake atmospheric transport CO 2 Carbon cycle in C-CAM coupled carbon-climate model CABLE interface to C-CAM Land-use and land-cover change fluxes Ocean Carbon fluxes heterotrophic respiration Phenology hydrology

Conformal-cubic C48 grid used for C4MIP simulations Resolution is about 220 km

Model forcing and modelled climate Sea Surface Temperature: HadISST1.1 dataset, 1x1 o, monthly CO 2 concentration Law Dome (pre 1958), then South Pole and Mauna Loa, smoothed Land air temperature

Carbon fluxes through 20 th century GPP – photosynthesis increases as atmospheric CO 2 increases NPP (photosynthesis minus plant respiration) and soil respiration increase with increasing CO 2 NEE (net exchange with atmosphere) starts ~neutral (tuned) and becomes sink

Map of output locations Red: atmospheric sampling sites, blue: flux tower sites Atmospheric data ‘see’ CO 2 sources/sinks from a larger region than flux towers Mauna Loa Barrow Ulaan Uul Cape Rama South Pole WLEF Tapajos

Seasonal cycle: amplitude and phase Model Observations Peak to peak amplitude – too low in northern mid-latitudes Month of minimum, out by 4-5 months in southern hemisphere Data: GLOBALVIEW- CO2 (2003)

Seasonal cycle: NH sites BarrowUlaan Uul Mauna LoaCape Rama Blue: obs Green: CABLE Red: CASA Data: GLOBALVIEW-CO2 (2003)

Seasonal cycle: southern hemisphere South Pole Blue: obs, green: model, red: CASA Contribution of source from each semi-hemisphere Data: GLOBALVIEW-CO2 (2003)

CO 2 at Mauna Loa CO 2 concentration (ppm) Annual growth of CO 2 (ppm/yr) Model: red, Observed: blue Data: Keeling et al (2005)

CO 2 Growth Rate Components at South Pole Station (ppm/yr) Fossil fuel Total Land use Biosphere Ocean

Future plans - Model simulated GPP,NPP,RP, Rs increased steadily over 20 th century with NEP changing from being slightly positive (source) to being slightly negative (sink) - Tropical rainforest and savanna were main contributors to global NEP variability - CO2 fertilization effect was strongest for tropical forest, savanna and C3 grass/agriculture - Simulated seasonal CO2 cycles were mostly good for Northern hemisphere stations and poor for Southern hemisphere Conclusions - implement new biogeochemical model - improve vegetation phenology - participate in the 2nd C4MIP experiment

Thank you

North-South gradient of CO 2 ( ) 90 o S90 o N N-S gradient too large in model. Assuming transport OK, implies source distribution incorrect – too little sink in northern hemisphere Data: GLOBALVIEW-CO2 (2003)

Zonal mean net ecosystem exchange Source Sink Northern mid-high latitude source, southern low-mid latitude sink indicates problems with biosphere simulation e.g. expect cultivation closer to neutral missing processes e.g. fire suppression, recovery from disturbance, nitrogen fertilisation, tropical deforestation

Modelled climate Land air temperature Screen temperature: DJF Obs Model

Potentially important feedbacks in coupled climate-carbon cycle system. Albedo (α) Increase in α Absorbed S w decrease R n decrease H & EL Cloudiness & Precip. decrease Increase in α Reduction in Soil moisture S w increase R n increase (+)(+) (-)(-) Response of the terrestrial biosphere to : increasing CO2 climate change climate variability Example of a simple albedo feedbacks

Soil moisture precipitation feedback in a climate model from M. Bierkens Univ Utrecht atmospheric moisture soil moisture precipitation evaporation vegetation albedo net radiation turbulence drainage SST variation advection monsoon strength CABLE coupled to three climate models:CCAM, Mk3, UM & TAPM

C4MIP Phase II results Differ by 292 ppm Sink Source Friedlingstein et al, J. Climate, 2006.

The current state of the science u The response of positive carbon-climate feedback is dominated by the terrestrial biosphere, particularly in the tropics. u The gain may be larger if nutrient interactions are included! u The gain in the positive feedback varies from 0.03 to 0.31, or an amplification factor of 1.04 to 1.45 by 2100.

Are those predictions realistic? Most models exaggerated the land biosphere potential to sequester atmospheric CO2!

Carbon–climate feedback (as in ESS) Carbon climate feedback is dominated by the responses of photosynthesis to CO2 ( a negative feedback) and respiration to temperature (a positive feedback) The negative feedback is dominant over the positive feedback in the short-term, but positive feedback will dominate over the negative feedback as  T > 2 o C Most models show a consistent positive feedback in the tropics and less consistent over other regions. Change in nutrient limitation is not considered.

Response to elevated CO2 is reduced when NPP is N-limited de Graaff et al GCB

Landcover effect on climate Brovkin et al., Global Change Biology, 2004

Interannual variability IAV NEE (green) and IAV temperature (red) Fit to long-term trend removed Correlation 0.59, weaker with precip

Amplitude of diurnal cycle WLEF tower, Wisconsin, USA Model levels approximately 37, 179, 456 m above surface Dotted lines – observed median amplitude Data: GLOBALVIEW-CO2 (2003)