1. CITRIS Seminar: Measuring and Interpreting Fluxes of Trace Gases Across Local and Global Networks A Nexus between Biometeorology and Engineering
Dennis Baldocchi
Professor of Biometeorology
Environmental Science, Policy and Mgt
University of California, Berkeley
April 25, 2012

2. Smart Dust Offers the Possibility of Spreading 100s and 1000s of
Sensors across the Environment

3. Wireless Networks Offer Possibility of Distributing Many Nodes
of Sensors Across a Landscape without Tethering to Wires

4. Measuring the State of the Environment has a Long History
Rudolf Oskar Robert Williams Geiger (August 24, 1894 in Erlangen, Germany – January 22, 1981 in Munich, Germany) was a German meteorologist and climatologist.

5. Challenges/Objectives
It’s en vogue to measure environmental state variables
Sensors are cheap, easy to replicate
Sensors Must have Good and Representative Exposure and Implementation and Known and Stable Calibrations
Measuring State Variables Should Not be the Means to an End
Changes in State Variables reflect Fluxes
Eddy Covariance
Flux measurements requires a system, 3-D Sonic Anemometer and Trace Gas Sensor
Sensor performance is stringent—Sensors must respond up to 10 Hz
Flux Gradient
Requires Resolving Very Small Gradients in Well-Mixed Fluids Above Vegetation
Flux-Gradient Theory is Error Prone in Shear Zones, like Canopy-Atmosphere Interface

6. ++ Thermometer is Shielded and Aspirated
Sensors Must Be Positioned Not to cause Artifacts
++ Thermometer is Shielded and Aspirated
-- Anemometer in Lee of Rain Gauge
-- Pyranometer Shaded by Trees, affecting light Quality and Quantity
--Rain Gauge May Under Count due to its height and wind
New Weather Station on West Circle

7. Conceptual Diagram of Changes in Atmospheric State Variables
in the Planetary Boundary Layer

8. ESPM 228 Adv Topics Micromet & Biomet
Mixed Layer Budget Eq.
Flux in from the top
Time rate
of change
of Cm
Growth – subsidence of PBL, h
Flux in the
bottom
ESPM 228 Adv Topics Micromet & Biomet

9. Question 1
Is Wireless Network Technology Ready for Flux-Gradient Measurements?
Yes, if Fine Vertical Gradients in Temperature, Wind and Humidity can be measured
Yes, if Flux-Gradient Theory is Applied Correctly

10. Restrictions for Application
K-Theory infers Fluxes; it does not measure them directly.
Gradients are Measured within the Constant Flux Layer, but above Vegetation
Sensors are Best placed Outside the Vegetation Canopy and Roughness Sub-layers
Extensive, Upwind Fetch must Exist
Steady-State Conditions Should Hold
ESPM 228 Adv Topic Micromet & Biomet

11. ESPM 228 Adv Topic Micromet & Biomet
Flux-Gradient Theory
Momentum Flux
Latent Heat Flux
(kg m-2 s-1)
(J m-2 s-1)
Sensible Heat Flux
Trace Gas Flux
(J m-2 s-1)
(mole m-2 s-1)
ESPM 228 Adv Topic Micromet & Biomet

12. ESPM 228 Adv Topic Micromet & Biomet
K Theory
Aerodynamic Technique: Assessed with Wind Velocity Profiles
ESPM 228 Adv Topic Micromet & Biomet 12

13. Measure Atmospheric State at Several Levels Above the Canopy
Comments/Recommendations
Measure Atmospheric State at Several Levels Above the Canopy
Measurements at One point Above the Canopy and Several within the Canopy Is Useless!!—though commonly observed
Better Precision is Obtained with Differential Measurements
Thermocouples resolve 1/40th C per millivolt and 1/80th C per millivolt if wired differentially

14. Flux Density: mol m-2 s-1 or J m-2 s-1
Eddy Covariance,
Flux Density: mol m-2 s-1 or J m-2 s-1
ESPM 228 Adv Topics Micromet & Biomet

15. ESPM 228 Adv Topics Micromet & Biomet
Eddy Covariance
Direct Measure of the Trace Gas Flux Density between the atmosphere and biosphere, mole m-2 s-1
In situ
Quasi-continuous
Integrative of a Broad Area, 100s m2
Introduces No artifacts, like chambers
ESPM 228 Adv Topics Micromet & Biomet

16. Flux Covariance is Determined by Resolving the Contribution across
a Spectrum of Slow to Fast Fluctuations in Wind Velocity and Scalar Mixing Ratio
CoSpectral Density
Wavenumber

17. ESPM 228 Adv Topic Micromet & Biomet
Real-time Sampling
Sample instruments at 10 to 20 Hz;
fsample = 2 times fcutoff (f=nz/U)
Store real-time data on hard disk, 10-30Mb/day
Calibrate Periodically
Process and Compute Means, Variances, Covariances, Skewness and Kurtosis.
Merge turbulence and meteorological data
Apply calibration coefficients and gas law corrections to compute unit-correct flux densities and statistics
Compute power spectra and co-spectra; examine instrument response and interference effects
ESPM 228 Adv Topic Micromet & Biomet

18. Question 2 Is Wireless Network Technology Ready for Eddy Covariance?
Yes, if Systems can Ingest High Input Information (10 Hz) from Flux Covariance Systems, operating off Solar Panels
Yes, if Low Power, 3 Dimensional Sonic Anemometer can be Coupled to Low Power Infrared Spectrometer.
Yes, if ample data storage (10-30 Mb/day) is available

19. New Low Power Methane and CO2/H2O Flux System
Power Consumption:
<24 W or 2 amp at 12vDC
CH4: 8 W
H2O/CO2: 10 W
Data logger, Met and Sonic Anemometer:
< 1 W

20. Past and Current Biomet Study Sites
Crops: soybeans, alfalfa, wheat, corn, rice
Grassland and peatland pastures
Boreal Conifer Forest
Deciduous Forest
Savanna Woodland

21. Annual Time Series of Trace Gas Exchange
Xu and Baldocchi, AgForMet, 2004

22. What is the State of the Flux Footprint?

23. ESPM Adv Topics Micromet & Biomet
2D-Footprint Model of Detto-Hsieh
ESPM Adv Topics Micromet & Biomet

24. Methane Flux Footprint of a Peatland Pasture
Detto et al AgForMet

25. ESPM 228 Adv Topic Micromet & Biomet
Representative Sampling of Energy Fluxes is Needed to Attain Better Energy Balance Closure
ESPM 228 Adv Topic Micromet & Biomet

26. FLUXNET: From Sea to Shining Sea 500+ Sites, circa 2011

27. Attributes of an Effective Network--People
You Need Boots on the Ground – e.g. Technicians, Students/Postdocs, Scientists--to Turn Measurements into Great Data
Integrated Data-Base is Needed to Process, Share, Distribute, Archive and Query Data
Scientific Community Needs to Access the Data to turn it into Information and Knowledge

28. Networks Widen the Number and Span of Climates, Biomes and Treatments (e.g. Biophysical Attributes and Traits, Disturbance, Land Use, and Management)
Baldocchi, Austral J Botany 2008

29. Net Ecosystem Carbon Exchange Scales with Length of Growing Season
Baldocchi, Austral J Botany, 2008

30. Question 3
How Do We Measure the Environmental Conditions of a Site in a Representative Way, that are Used to Interpret Fluxes?

31. Temperature/Humidity Sensor
ESPM 129 Biometeorology

32. Temperature Microclimate
Canopy Ht
3m Sample:
100 W m-2/(1.2 * 1005 * 0.4 * 3 * .3) = 0.23 C/m
20 m Sample:
3m canopy: 100 W m-2/(1.2 * 1005 * 0.4 * 20 * .3) = C/m

33. Temperature and Humidity Sensors Must be
Shielded from the Sun and Aspirated!—Hence, Power is Required
At 20 C, the T difference is C

34. Biases between Aspirated and Unaspirated Thermometers are too Large,
Compared to Gradients and Treatment Differences One Seeks to Quantify in Nature

35. Transects of PAR under a forest canopy
Radiation Field is Bi-Modal, Non-Gaussian and Possesses a High Coefficient of Variation (> 50%)
ESPM 129 Biometeorology

36. A statistical estimate of the number of sensors needed to define the light environment
CV, coefficient of variation (per cent)
n, number of samples (within 10% of the population mean)
n, number of samples (within 5% of the population mean)
150
609
865
100
270
382 50 68 96 25 17 24 10 3 4
ESPM 129 Biometeorology

37. Tram with radiation sensors traversing in the understory of a savanna woodland
ESPM 129 Biometeorology

38. Radiation Tram System

39. Quantum Sensors are Not Ideal Substitutes for Pyranometers
If Energy Flux Density (J m-2 s-1) is preferred to Quantum Flux Density (moles of quanta m-2 s-1)

40. Vegetation Selectively Filters Light Energy in Visible,
And Enhances Scattering in NIR

41. Circuit and Power Supply
LEDs as Sensors for Measuring Reflected Light in Narrow Wavebands
LED components
Incoming Light
Data Logger
Circuit and Power Supply
LED Casing
Figure by Mirabel Jaquez, Future UCB Engineering Graduate Student

42. Multi-Band LED Reflectance Sensor to Assess Canopy Structure and Function

43. Multi-Band LED Reflectance Sensor

44. Links between NDVI and Canopy Photosynthesis

45. Upward Looking Digital Cameras Assess Canopy Structure and Phenology
Promise for Cheap Networks of CCD/Cameras

46. Digital Cameras Produce Continuous Measure of Canopy Structure
Ryu et al.

47. CO2 Microclimate

48. Probes for Studying CO2 in the Soil

49. Don’t‘ Trust Manufacturers Specs and Calibrate;
Calibrate with Trusted Standards tied to NOAA ESRL

50. Each Sensor Has Unique Calibration,
which can cause +/-100 ppm differences in Readings Among Sensors

51. Soil CO2 Gradients and Strong and Amenable for Flux Computations

52. Tests Prove Adequate

53. Continuous Measurements of Soil Respiration with CO2 Gradient System is Possible

54. Time Domain Reflectometers are Expensive to Replicate,
Soil Moisture
Time Domain Reflectometers are Expensive to Replicate,
Good for Measuring Time Series
ESPM 129 Biometeorology

55. Wireless Soil Moisture Network
To Measure Volumetric Water Content
+/- 2% with 95% C.I.
4-19 Sensors for Area < 900 m2
44-80 sensors for Area: m2
Robinson et al 2008 Vadose Zone Journal

56. Soil Moisture Fields with ElectroMagnetic Inductance, EMI
Complex Spatial Fields Exist and Merit Scrutiny
Analysis by Trenton Franz, Arizona

57. ET and Soil Water Deficits: Root-Weighted Soil Moisture
Baldocchi et al., 2004 AgForMet

58. Concluding Points
Fluxes Diagnose the State of the Atmosphere; State Variables Don’t Infer Fluxes Well
Make sure you Measure the State of the Environment, Not the State of the Sensor
Calibrate, Calibrate, Calibrate
New Challenges are Needed to Develop Flux Systems that can be Networked Over Space to Measure Advective Fluxes and Spatial Heterogeneity
Wireless Network of LED sensors, Digital Cameras, Soil probes for moisture,
Revolution in New Trace Gas Sensors with Tunable Diode Laser Spectroscopy