Using Biophysical Models and Eddy Covariance Measurements to

Using Biophysical Models and Eddy Covariance Measurements to Ask (and Answer) Questions About Biosphere-Atmosphere Interactions Dennis Baldocchi Biometeorology Lab ESPM University of California, Berkeley

What is the State of the Atmosphere? Conservation of Mass, e. g. Solving the Bathtub Problem The ‘Level’ in a Tub depends on the Fluxes IN and OUT of the Tub

Trace Gas and Energy Fluxes between Land the Atmosphere Isotopic exchange Courtesy of Jose Fuentes, UVa

Quantifying Sources and Sinks • Biology: – Leaf area density, a(z) – internal conc, Ci – stomatal resistance, rs • Physics: – Boundary layer resistance, rb – Scalar conc, C(z)

Biogeophysical-Ecohydrological View

Controlling Processes and Linkages: Roles of Time and Space Scales

Sub-Grid Variability: What Errors arise from Averaging?

Eddy Covariance Technique • Oak Savanna • Annual Grassland • Peatland/Pasture • Temperate Deciduous Forest • Boreal Conifer Forest • Crops

FLUXNET: From Sea to Shining Sea 400+ Sites, circa 2007

CANVeg MODEL Physiology Photosynthesis Stomatal Conductance Transpiration Micrometeorology Leaf/Soil Energy Balance Radiative Transfer Lagrangian Turbulent Transfer

Key Attributes of Can. Veg • • • Seasonality – Leaf Area Index – Photosynthetic Capacity (Vcmax) Model parameters based on Site Measurements and Eco. Physiological Rules and Scaling Functions – Stomatal Conductance scales with Photosynthesis – Jmax and Rd scale with Vcmax Multilayer Framework – Computes Fluxes (non-linear functions) on the basis of a leaf’s local environment – Considers • • Sun and Shade Leaf Fraction Leaf Clumping Leaf Inclination Angle Non-local Turbulent Transport and Counter-Gradient Transfer

Models Must Consider Seasonality in Leaf Area Index and Photosynthetic Capacity, Vcmax Wilson et al. 2001 Tree Physiol ESPM 228, Advanced Topics in Micromet and Biomet

Results and Discussion

Can Principles from a Global Network Produce Insights about Global-Scale Fluxes? What is the Upper Bound of GPP and its Variability? Top-Down: GPP Scales with Energy Bottom-Up: Counting Productivity on leaves, plant by plant, species by species ESPM 111 Ecosystem Ecology

Potential and Real Rates of Gross Carbon Uptake by Vegetation: Most Locations Never Reach Upper Potential GPP at 2% efficiency and 365 day Growing Season tropics GPP at 2% efficiency and 182. 5 day Growing Season FLUXNET 2007 Database

Upper-Bound on Global Gross Primary Productivity • Global GPP is ~ 120 * 1015 g. C y-1 • Solar Constant, S* (1366 W m-2) – Ave across disk of Earth S*/4 • • • Transmission of sunlight through the atmosphere (1 -0. 17=0. 83) Conversion of shortwave to visible sunlight (0. 5) Conversion of visible light from energy to photon flux density in moles of quanta (4. 6/106) – Mean photosynthetic photon flux density, Qp Fraction of absorbed Qp (1 -0. 1=0. 9) Photosynthetic efficiency, a (0. 02) Arable Land area (~ 100 * 1012 m 2) Length of daylight (12 hours * 60 minutes * 60 seconds = 43200 s/day) Length of growing season (180 days) Gram of carbon per mole (12) GPP = 1366*0. 83*0. 5*4. 6*0. 9*0. 02*100*1012*43200*180*12/4=108*1015 g. C y-1 ESPM 111 Ecosystem Ecology

Random Sampling Error Reaches Equilibrium with > 60 Sites

Interannual Variability in GPP is small, and not significantly different, across the Global Network

Little Change in Abiotic Drivers-annual Rg, ppt --across Network

Answering Questions with Models • Roles of Structure and Function • • Leaf Angles and Clumping Leaf Area Index Photosynthetic Capacity Phenology • Roles of Microclimate Conditions on Mass and Energy Exchange – Diffuse Light – Humidity – Temperature • Sub-Grid Parameterization, Energy Balance Closure and Scaling – Insights from a 2 -D, ‘Wet’ Daisy. World

CO 2 Flux Model Test: Hourly to Annual Time Scales

Time Scales of Interannual Variability Baldocchi et al. , 2001 Ecological Modeling

Role of Proper Model Abstraction ESPM 111 Ecosystem Ecology

How Long should one Measure Fluxes? : Decadal Power Spectrum of CO 2 and Water Vapor Fluxes

Emergent Processes: Impact of Leaf Clumping on Canopy Light Response Curves

Interaction between Clumping and Leaf Area

How Sensitive are Fluxes to Leaf Inclination Angle Distribution? m-2 a-1) NEE (g. C l. E (MJ m-2 a-1) H (MJ m-2 a-1) clumped -577 1690 1096 random -354 1551 1032 spherical -720 1774 1095 erectophile -1126 2023 1171 ESPM 228 Adv Topics Micromet & Biomet planophile -224 1473 1008

Carbon, Water and Sensible Heat Exchange scale with Photosynthetic Capacity NEE (g. C m-2 a-1) l. E (MJ m-2 a-1) H (MJ m-2 a-1) Vcmax(73) -577 1690 1096 Vcmax(50) -454 1584 1199 % difference -21. 3 -6. 3 9. 3 ESPM 228 Adv Topics Micromet & Biomet

Leaf Size has a Modest Effect on Carbon & Water Exchange, But a Large Effect on Sensible Heat Exchange NEE (g. C m-2 a-1) l. E (MJ m-2 a-1) H (MJ m-2 a-1) 0. 1 m 0. 01 m 0. 001 m -577 1690 1096 -588 1652 1164 -586 1615 1202 ESPM 228 Adv Topics Micromet & Biomet

Are VOCs a Large Source of Carbon?

Net Ecosystem Carbon Exchange scales with Growing Season Length Baldocchi et al, 2001 Ecological Modeling

Soil Temperature: An Objective Indicator of Phenology? ? Baldocchi et al. , 2005 Int J Biomet.

Soil Temperature: An Objective Measure of Phenology, part 2 Baldocchi et al. Int J. Biomet, 2005

Spatialize Phenology with Transformation Using Climate Map Baldocchi, White, Schwartz, unpublished

Flux Based Phenology Patterns with Match well with data from Phenology Network White, Baldocchi and Schwartz, unpublished

How do Sky Conditions Affect Net Carbon Exchange (NEE)? : Data Baldocchi, 1997 PCE Niyogi et al. , GRL 2004

More Diffuse than Direct Light is Intercepted

The ‘Diffuse-Light Enhancement’ is a function of LAI Knohl and Baldocchi, 2008 JGR Biogeosci

There are Trade-Offs between Reducing Light Amount (with Clouds and Aerosol) and Increasing Light Use Efficiency Knohl and Baldocchi, 2008 JGR Biogeosci

Canopy Photosynthesis and Aerosols: Impact on Daily & Annual Time Scales, II reference NEE (g. C m-2 a-1) l. E (MJ m-2 a-1) H (MJ m-2 a-1) -577 1690 1096 direct radiation, 20% -655 1729 1058 % difference 13. 5 2. 3 -3. 5

Simple Model suggests A/T decreases with increasing D or Ci/Ca

Water Use Efficiency: Ci/Ca , Vapor Pressure Deficit and Diffuse Light Fraction But Complex feedbacks among Ci/Ca, humidity and diffuse light need to be considered! Knohl and Baldocchi, unpublished

How Do Changes in vpd and Ci/Ca conspire to affect A/T?

In toto (considering coupled energy balance feedbacks) A/T increases with Ci/Ca

A/T, Stable Isotope Discrimination and Diffuse Light Knohl and Baldocchi, unpublished

Leaf Size and Extinction • Major Extinction at Triassic-Jurassic Boundary during period of Elevated Greenhouse effect – – – 4 fold increase in CO 2 3 to 4 C temperature increase 99% species turnover of megaflora with leaves > 5 cm 10% species turnover of flora with leaves < 0. 5 cm Small Leaves are more effective in transferring heat and experiencing lethal surface temperatures Mc. Elwain et al Science, 1999 ESPM 111 Ecosystem Ecology

Why are Leaves Certain Sizes? Biophysics as an Evolutionary Filter Leaf Temperature and Leaf Morphology

Leaf size, CO 2 and Temperature: Why are oak leaves small in CA and large in TN?

Leaf Temperature and Isotopes? Helliker and Richter 2008 Nature

Sub-Grid Variability: Lessons Derived from Wet Daisy. World Latent Heat Exchange Map

Spatial Variation in sub. Grid T and LE follows Power Law Scaling Baldocchi et al, 2005 Tellus

Sub-Grid Scaling Errors in ET

Conclusions • Biophysical Model aids in understanding the impact of diffuse light on photosynthesis, isoprene emission, water use efficiency and stable isotope discrimination • A cellular automata, energy balance model shows that spatial averaging of energy balance drivers can produce huge errors in grid-scale energy fluxes and can explain lack of energy balance closure

Acknowledgements • Funding – NASA, DOE/TCP, NIGEC/WESTGEC, NSF, Microsoft • Eddy Covariance Measurements – Kell Wilson, Bev Law, Alexander Knohl • FLUXNET – Eva Falge, Lianhong Gu, Deb Agarwal et al • Canopy Modeling and Diffuse Light – Alexander Knohl, Lianhong Gu, Kell Wilson • Isoprene – Peter Harley, Jose Fuentes, Dave Bowling, Russ Monson & Alex Guenther • 13 C Isotopes – Alexander Knohl, Dave Bowling, Russ Monson