Forward and Inverse Forward and Inverse Modeling Modeling of Atmospheric CO of Atmospheric CO 2 2 Scott Denning, Nick Parazoo, Kathy Corbin, Marek Uliasz, Andrew Schuh, Dusanka Zupanski, Ken Davis, and Peter Rayner Acknowledgements: Support by US NOAA, NASA, DoE
26
Embed
Forward and Inverse Modeling of Atmospheric CO 2 Scott Denning, Nick Parazoo, Kathy Corbin, Marek Uliasz, Andrew Schuh, Dusanka Zupanski, Ken Davis, and.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Forward and Inverse Forward and Inverse Modeling Modeling
of Atmospheric COof Atmospheric CO22
Scott Denning, Nick Parazoo, Kathy Corbin, Marek Uliasz, Andrew Schuh, Dusanka Zupanski, Ken Davis, and Peter Rayner
Acknowledgements:Support by US NOAA, NASA, DoE
Signal? Noise? Which is which?Signal? Noise? Which is which?
Cape Grim
Usual approach is to exclude“spikes” as non-“background”
Law et al inverted the “spikes” instead!
Effects of height-time concentration Effects of height-time concentration variation near the groundvariation near the ground
averaged together (diurnal & seasonal cycle removed)
• Some sites show frontal drop in CO2, some show frontal rise … controls?
• Simulated shape and phase similar to observations
• What causes these?
wplsobs
frs
sgp
wkt
hrv
amtlef
ring
Deformational FlowDeformational Flow
• Anomalies organize along cold front
• dC/dx ~ 15ppm/3-5°
Dg
Dt
∂C∂x
+∂C∂y
⎛⎝⎜
⎞⎠⎟=−
∂ug
∂x∂C∂x
+∂vg
∂x∂C∂y
⎛⎝⎜
⎞⎠⎟−
∂ug
∂y∂C∂x
+∂vg
∂y∂C∂y
⎛⎝⎜
⎞⎠⎟
shear deformation- tracer field rotated by shear vorticity
stretching deformation- tracer field deformed by stretching
gradientstrength
ΔC
Δt= u
ΔC
Δx→ C day+1 =
uΔtΔC
Δx=
5ms−1 * 3600s * 24hr *15 ppm
5° *100km= 12 ppm
Lateral Boundary ForcingLateral Boundary Forcing
• Flask sampling shows N-S gradients of 5-10 ppm in [CO2] over Atlantic and Pacific
• Synoptic waves (weather) drive quasi-periodic reversals in meridional (v) wind with ~5 day frequency
• Expect synoptic variations of ~ 5 ppm over North America, unrelated to NEE!
• Regional inversions must specify correct time-varying lateral boundary conditions
• Sensitivity exp: turn off all NEE in Western Hemisphere, analyze CO2(t)
• Run 1: Surfaces fluxes defined everywhere on Earth• Run 2: Surface fluxes set to 0 in Western Hemisphere, including NA• Correlation of the 2 experiments in July (mid-day values only) shows
the importance of lateral flow over NA (R2 = 35-70% in SE!)
Regional Fluxes are Hard!Regional Fluxes are Hard!
• Eddy covariance flux footprint is only a few hundred meters upwind
• Heterogeneity of fluxes too fine-grained to be captured, even by many flux towers– Temporal variations ~ hours to days– Spatial variations in annual mean ~ 1 km
• Some have tried to “paint by numbers,” – measure flux in a few places and then apply
everywhere else using remote sensing
• Annual source/sink isn’t a result of vegetation type or LAI, but rather a complex mix of management history, soils, nutrients, topography not easily seen by RS
A Different StrategyA Different Strategy• Divide carbon balance into “fast” processes that
we know how to model, and “slow” processes that we don’t
• Use coupled model to simulate fluxes and resulting atmospheric CO2
• Measure real CO2 variations• Figure out where the air has been • Use mismatch between simulated and observed
CO2 to “correct” persistent model biases
• GOAL: Time-varying maps of sources/sinks consistent with observed vegetation, fluxes, and CO2 as well as process knowledge
FCO2 (x, y, t) =R(x,y,t)−GPP(x,y,t)
Treatment of Variations for Treatment of Variations for InversionInversion
• Fine-scale variations (hourly, pixel-scale) from weather forcing, NDVI as processed by forward model logic (SiB-RAMS)
• Multiplicative biases (caused by “slow” BGC that’s not in the model) derived by from observed hourly [CO2]
FCO2 (x, y, t) =βR(x,y)R(x,y,t)−βGPP (x,y)GPP(x,y,t)