Spatial%20and%20Temporal%20Patterns%20of%20the%20Contemporary%20Carbon%20Sources%20and%20Sinks%20in%20the%20Ridge%20and%20Valley%20Ecoregion%20of%20the%20United%20States

1
Spatial and Temporal Patterns of the Contemporary
Carbon Sources and Sinks in the Ridge and Valley
Ecoregion of the United States
Shuguang Liu and Thomas Loveland USGS National
Center for Earth Resources Observation and
Science Sioux Falls, SD 57198
2
Outline

• The US Carbon Trends Project
• Research Questions
• Methodology
• The Ridge and Valley Ecoregion
• Land Cover Change
• Spatial and Temporal Variability of C Stocks and
  Fluxes

3
Overarching Research Questions of the US Carbon
Trends Project

• What is the spatial, temporal, and sectoral
  variability of conterminous U.S. land cover
  change from 1973 to 2000.
• What are the spatial and temporal distributions
  of carbon sources and sinks, and therefore the
  dynamics of carbon storage in the conterminous
  U.S.?
• What are the major driving forces that dictate
  the evolution of US terrestrial carbon storage
  and the CO2 exchange between the land and the
  atmosphere?
• What are the major uncertainties and knowledge
  gaps associated with estimating regional and
  national carbon dynamics?

4
US Land Cover Change

• There is no consistent database available that
  characterizes the contemporary US land cover
  change, because
• Land cover change mapping over large areas is a
  major effort
• Labor intensive
• Money (funding sources)

5
US Land Cover Change Detection

• Probability-based sampling strategy used to
  provide efficient and reliable estimates of land
  cover change over large areas. Goal is to detect
  within one percent of actual change at 85
  confidence level.
• Ecoregions are sampling strata
• Land cover change was detected using Landsat
  images (i.e., 1973, 1980, 1986, 1992, and 2000)

• Sampling units are 20- or 10-km2.
• Samples randomly selected within strata.
• Sample size based on expected spatial variability
  of change in the strata.

6
Spatially Explicit Modeling

• GEMS (General Ensemble Biogeochemical Modeling
  System)
• An advanced modeling systems for spatially
  explicit simulation of biogeochemical cycling
  over large areas
• Developed at USGS National Center for Earth
  Resources Observation and Science
• Deployment of the encapsulated plot-scale model
  in space is based on a Joint Frequency
  Distribution of the major controlling variables
  (e.g., land cover, climate, soil, etc.).
• Included data assimilation algorithms
• It includes a dynamic land cover/use change
  submodel
• Stochastic simulations to incorporate
  uncertainties in input data
• Uncertainty estimate of carbon dynamics
• Major applications (US, Africa, and Central
  America)

7
Spatially Explicit Modeling
National Benchmark Databases
Land Cover USGS Land Cover Trends Soil
STATSGO Climate CRTUS2.0 (1900 2000) N
Deposition National Atmospheric Deposition
Program Crop Information USDA Agricultural
Census Data FIA Forest biomass, NPP, Age
Distribution
GEMS
Carbon dynamics simulated at 60 m x 60 m spatial
resolution within 20 km x 20 km or 10-km by 10-km
sampling blocks
8
Spatially Explicit Modeling
Quantify the spatial and temporal changes of C
stocks, fluxes, and uncertainty at various scales
Block Ecoregion Nation
(10 km)
(60 m resolution)
9
Ridge and Valley Ecoregion
Geographic Location and Samples
The ecoregion spans 8 states. A total of 40
10-km by 10-km sample blocks were randomly
selected for land cover change detection and
subsequent carbon simulations.
10
Ridge and Valley Ecoregion
Land Cover Composition Around 1973
Forest 57.1 Cropland 31.4 Urban 7.9
11
Ridge and Valley Ecoregion
Land Cover Change 1973-2000

• Extensification of forest harvesting activities
• Forest area reduction for2trans gt trans2for
• Ag land reduction ag2for ? for2ag and
  urbanization
• Urbanization (for2urban, ag2urban)
• Annual change rate increases with time

(A) Annual rate of land cover change during four
time periods. (B) the total share percentage of
six major land cover change activities (C through
F) in the total change rate, and (C through F)
the share percentages of the major land cover
change activities.
Land cover compositions ()
12
Forest Inversion
FIA data biomass stock by age class (therefore
biomass accumulation rates implicitly used) and
total standing biomass MODIS annual NPP
2000-2001
13
C Sink vs. C Sequestration
C Sequestration C Sink - C Removal
and C Removal GrainYield WoodHarvested
14
Ridge and Valley Ecoregion
Interannual and Spatial Variability (Blocks)
Data show block-scale annual averages from 1973
to 2000 X axis shows spatial variability across
10-km by 10-km blocks Y axis shows interannual
fluctuations by blocks.
15
Ridge and Valley Ecoregion
Interannual and Spatial Variability (Blocks)
Data show block-scale annual averages from 1973
to 2000 X axis shows spatial variability across
10-km by 10-km blocks Y axis shows interannual
fluctuations by blocks.
Net Primary Productivity (NPP)
Large interannual variability
Total Carbon Stock Change

• C sequestration strength increases from north
  (lower block ID numbers) to south
• Large interannual variability

Soil Organic Carbon Change
Relatively smaller variability
16
Ridge and Valley Ecoregion
C Stock and Land Cover Composition (Blocks)

• C stock at the block scale is
• Positively correlated to forest fraction
• Negatively correlated to cropland fraction
• Not related to other land cover types.

17
Ridge and Valley Ecoregion
C Sequestration and Land Cover Composition
(Blocks)

• C sequestration at the block scale is
• Positively correlated to forest fraction
• Negatively correlated to cropland fraction
• Not related to other land cover types.

18
Ridge and Valley Ecoregion
Carbon Rich Gets Richer (Blocks)
Soil sequestration accounted for about 35 of the
total C sequestration Soil was a C source when
total C sequestration was less than 50 g C m-2 y-1
C change rate in biomass and soils increases with
total C stock
19
Ridge and Valley Ecoregion
Temporal Change of C Stocks (Ecoregion)
20
Ridge and Valley Ecoregion
Temporal Change of C Fluxes (Ecoregion)

• Large inter-annual variability in NPP, C
  sequestration, and total C sink.
• Soil C sink and total C sink is decoupled.
• C sequestration is tightly coupled with C sink
  strength.
• Harvested wood C increased over time because of
  extensification of clearcutting.
• The average C sequestration rate was 96 ? 12 (1?)
  gC m-2 y-1.

21
Ridge and Valley Ecoregion
C Sinks and C Sequestration (Ecoregion)

• C sequestration is tightly coupled with C sink
  strength.

22
Summary

• Land cover change was very dynamic. Major
  changes include urban expansion, reduction in
  cropland area, and extensification of
  clearcutting activities.
• Large spatial and inter-annual variability in
  NPP, C sequestration, and total C sink.
• C change rate in biomass and soils increases with
  total C stock
• Soil C sink and total C sink is decoupled.
• C sequestration is tightly coupled with C sink
  strength.
• Harvested wood C increased over time because of
  extensification of clearcutting.
• The average C sequestration rate was 96 ? 12 (1?)
  gC m-2 y-1.
• Soil sequestration accounted for about 35 of the
  total C sequestration. Soil was a C source when
  total C sequestration was less than 50 g C m-2
  y-1

23
Poster 312

• Soil Organic Carbon Budget as Related to Land Use
  History in the Northwestern Great Plains