Testing the Flux Tower Upscaling Hypothesis at a Regional Scale in a Complex Landscape

1
Testing the Flux Tower Upscaling Hypothesis at a
Regional Scale in a Complex Landscape

• P. Bolstad, B. Cook, University of Minnesota
• K. Davis, Penn State University
• F.A. Heinsch, University of Montana
• R. Kolka, M. Kubiske, USFS

2
WLEF Tall Tower
Source M.L. Jensen
3
Flux scaling hypothesis C Flux quantified and
uncertainties reduced with a multi-tiered
measurement and modeling approach Goals Test
upscaling approaches and quantify uncertainties
across time and space
Source Ken Davis, PSU
4
Northern Wisconsin Test of NACP Upscaling Approach
5
Multi-tier Monitoring Concept for the Land
Component of NACP
R.Kolka, USFS
6
Tier 4 Canopy Flux Towers
Lost Creek - shrub fen
Willow Creek 70 year old maple-basswood
Sylvania Wilderness 150 year-old hemlock-birch
old growth
7
Clearcut/young aspen
Mobile flux towers
Young-intermediate aspen
Sedge, ericaceous, sphagnum wetlands
source CAVM
8
Tier 4 Sites

• Photosynthesis Light response, A/Ci, diurnal
  cycles
• Leaf, stem, and branch respiration
• Soil respiration
• Sapflux/transpiration
• LAI, leaf nitrogen, SLA (vertical profiles)
• Litterfall
• Soil chemical and physical properties
• Site biometry height, diameter, 1, 5, and 10
  year growth
• CWD, littermass

• CO2 and H2O Fluxes at 29.6 m
• CO2 and H2O mixing ratio profile (6 levels)
• Above canopy radiation (29.6 m)
• Solar, infrared, and PAR (direct/diffuse)
• Below canopy radiation
• PAR (4 levels), net radiation above the forest
  floor (2 m)
• Air temperature profile (9 levels)
• Wind speed/direction profile (4 levels)
• Soil temperature and moisture profile (6 depths)
• Soil heat flux (7.5 cm)
• Tree stem temperatures (maple, north and south
  facing)
• Leaf wetness indicator (29.6 m)
• Precipitation (rain and snow)

J. Tang
9
Tier 3 inventory plots
FIA plot design, augmented measurement set
9-36 inventory plots km-2, centered on fluxt tower
10
Number of FIA Plots, by County, Forest Condition
(n1208)
Tier 2 FIA Measurement and Analysis
1 plot every 44 sq. km. 1 plot every 11,500 ac.
WLEF
100 km
11
Tier 1 modeling Biome-BGCResultsforPark
Falls,1997-2003
y 1.1007x 0.2361 r2 0.3096
12
Error/Uncertainty SourcesPrimary Contributions

• Poorly formulated model
• Validate MODIS NPP/GPP/R estimates, develop
  wetlands-specific model, evaluate LUE-based
  approach
• Poor model parameterization
• Range of physiological parameters, NPP-age by
  primary vegetation conditions
• Errors in characterizing the environment
• Measured spatial flux variation, improved MODIS
  landcover methods, characterize primary
  vegetation conditions

13
Finis