Viewed from the air

1
Lake Superior Region Carbon Cycle

• Viewed from the air

Ankur R Desai Atmospheric Oceanic
Sciences University of Wisconsin-Madison (and the
CyCLeS team)
Lake Superior Biogeochemistry Workshop August 5,
2008
2
Whats in the airwaves?

• Lakes, lands, carbon
• The atmospheric tracer view
• An eddy flux view
• Lake Superior micrometerology

3
Lakes, Land, Carbon
4
The big picture

• Sarmiento and Gruber, 2002, Physics Today

5
Slightly smaller picture

• Cardille et al. (2007)

6
Real Numbers Are Complicated

• Atmos. flux 3-12 Tg yr-1 - 35-140 gC m-2 yr-1

7
An Oceanic Lake

• CyCLeS Cycling of Carbon in Lake Superior
• Adapt the MIT-GCM ocean model to simulate
  physical and biogeochemical environment of Lake
  Superior
• Physical model of temperature, circulation
• Mostly implemented
• Biogeochemical model of trace nutrients and
  air-sea exchange
• In progress

8
Interesting Questions

• How do magnitudes of lake and land flux compare
  and what does it imply for regional carbon
  budgets? (NACP, SOCCR)
• Are interannual variations in lake and land CO2
  surface-atmosphere flux related and if so, due to
  what environmental forcing?
• Can we see and constrain lake (and land) flux
  from regional atmospheric CO2 observations?
• What are impacts on atmospheric forcing
  (temperature, stable layer depth, CO2) on lake
  biogeochemistry?

9
The Atmospheric Tracer View
10
Global CO2

• NOAA/ESRL/GMD/CCGG

11
Global Experiment

• Marland et al., DOE/CDIAC

12
The Inverse Idea
13
The Inverse Idea

• Courtesy S. Denning, CSU

14
The Inverse Idea

• Peters et al (2007) PNAS

15
Inversion and a Very Big Tower

• WLEF-TV (PBS)
• Park Falls, WI
• 447-m tall
• 6 levels CO2
• 11 to 396 m
• 3 levels CO2 flux
• 30,122,396 m
• Mixed landscape
• Representative?
• Running 1995-

16
A 1-point Inversion

• CO2 Air flowing over lake gt CO2 over land

17
Air and Lake CO2 Comparison

• Simple boundary layer budget tracer study
  suggests summer 2007 efflux 4-14 gC m-2 d-1
• extrapolated to 30-140 gC m-2 yr-1
• Analysis requires modeling of stable marine
  boundary layer
• Larger than traditional air-sea pCO2 exchange
  calculation
• Requires significant respiration in water column
• Urban et al. (in press)

18
The Boundary Layer Problem

• Courtesy of S. Spak, UW

19
Getting More Sophisticated

• Courtesy M. Uliasz, CSU
• Tracer transport modeled influence function
  August 2003 at WLEF

entire domain
water
land
20
Great Lakes Influence at WLEF

• Land 85.4
• Lake Superior 9.5
• Lake Michigan 1.8
• Other water 3.1

21
The Potential

• Potential exists for constraining flux and
  interannual var. with local observations of CO2

1996
2003
22
An Eddy Flux View
23
Eddies?

• Tracers in boundary layer primarily transported
  by turbulence
• Ensemble average turbulent equations of motion
  and tracer concentration provide information
  about the effect of random, chaotic turbulence on
  the evolution of mean tracer profiles with time
• In a quasi-steady, homogenous surface layer, we
  can simplify this equation to infer the surface
  flux of a tracer

24
Eddies!
25
The Maths

• Some simplifications made

Storage
Turbulent flux

• Equipment
• 3D sonic anemometer
• Open or closed path gas analyzer
• 5--20 Hz temporal resolution
• Multiple level CO2 profiler

26
The Data
27
The Data Pt. 2
28
The Data Pt. 3
29
Much Data
30
A CHEAS-y Lake
31
Scale This!
32
Some Observations
Desai et al, 2008, Ag For Met
33
The 6x6 km View
34
More Observations
35
Land History
36
Land History

• Have to account for age structure too

37
All The ChEAS Flux Data
38
Magically Scaled
39
The Bottom-Up Flux
40
Evaluation

• Top-down vs Bottom-up

41
Evaluation
42
Land

• 1989-2006 average

43
Lake?
44
Lake and Land
45
Lake Superior Micrometeorology
46
Better Forcing?

• Many observations are sparse

47
Better CO2
48
Coherent Interannual Variability
49
Lake Interannual Variability

• Annual avg. dissolved organic carbon (DOC)

50
More measurements

• CO2 over Lake Superior
• Continuous CO2 eddy covariance on the lake
• Better models of stability over lakes
• Spatial atmospheric met data
• Temp, wind, precip?, shortwave radiation

51
Conclusions

• On annual and decadal timescales, Lake Superior
  is possibly a source of CO2 to the atmosphere
• This source could be on the same order of
  magnitude as the terrestrial regional sink
• Regional carbon budgets have to take lakes into
  account
• We can estimate this flux from a number of
  techniques
• Lake models may need to worry about
  spatiotemporal variability in atmospheric forcing
• Models to tie land carbon flows into lake carbon
  can be useful for Lake Superior
• Model-data fusion/optimization/assimilation
  techniques should be explored

52
Thanks

• Desai lab and friends Ben Sulman, Jonathan Thom,
  Shelley Knuth, Scott Spak
• ChEAS collaborators, esp. Bruce Cook, Paul
  Bolstad, Ken Davis, D. Scott Mackay, Nic
  Saliendra, Sudeep Samanta
• CyCLeS team Galen McKinley, Noel Urban, Chin Wu,
  Nazan Atilla, Val Bennington
• Funding DOE NICCR, NSF, USDA, NSF/NCAR, NASA,
  NOAA, under auspices of the North American Carbon
  Program (NACP)
• Come visit us
• AOSS 1549, desai_at_aos.wisc.edu, 265-9201
• More info
• http//flux.aos.wisc.edu