1. Human Appropriation of NPP (HANPP) An accounting framwork for analysing land use processes in the Earth system
Karlheinz Erb
Institute of Social Ecology, Vienna
in collaboration with: H. Haberl, V. Gaube, S. Gingrich, C. Plutzar, F. Krausmann, W. Lucht, A. Bondeau, et al.
GEOSS support for IPCC assessments
Geneva, Feb. 3, 2011

2. Overview
Background: the integrated land system & the current mainstream state-of-the-art in LULC science
The framework „Human Appropriation of Net Primary Production“: conceptual background & method
Results: Global HANPP 2000
Examples: global production-consumption link, global bioenergy potentials
Conclusions: data requirements, gaps, challenges and opportunites

3. State-of-the-art of LU science
From current mainstram land-use research... +
...towards an integrated understanding of land use
Classification systems creating nominal-scale data
Focus on land cover (biophysical structures, ecological systems)
Focus on forest / non-forest dynamics
Strategy: increasing spatial resolution
Focus on society-nature interactions
Broad range of land uses
Continuous (rational) scales
Explicitly addressing a wide range of spatial scales

4. Bridging disciplinary boundaries: the integrated land system
Outputs - Benefits
Ecosystems
Society
Inputs - Investments
Steffen et al. Science 1998
Matthews et al. 2000
management

5. HANPP – the ‚human appropriation of net primary production‘
Potential NPP
Actual NPP
NPP remaining
after harvest
Outputs - Benefits
Change
induced
through
land
use
natural
ecosystem
managed
Ecosystems
Society
dNPPLC
HANPP
Inputs - Investments
NPPh

6. Data integration NPP0: LPJ-DGVM NPPact NPPh Non-used areas Irrigation
Degradation
NPPact
NPPh
Erb et al., J of Land Use Science, 2007

7. Result: Global HANPP 2000 NPPLC%: Productivity changes due to land
coversions
<< 10% >>
Land use activities
Biomass consumption
HANPP%:
Aggregated effect of
land use and
harvest
<< 24% >>
Source: Haberl et al PNAS 2007
Krausmann et al., 2008

8. Summary of results HANPP 2000
Global HANPP amounts to 24% of NPP0 (aboveground 30%)
Agriculture is the most important driver:
Cropping and grazing contribute 3/4 of global HANPP.
Feeding of livestock consumes 2/3 of the total amount of biomass used by humanity
Considerable regional variation of HANPP, mainly depending on
Consumption level (per capita HANPP in industrialized countries is about twice that of developing countries)
Population density
Technology: yields

9. HANPP data integration: ‚old‘ and ‚new‘ challenges
Ecosystems
Socioeconomic Systems
Area
Land cover
e.g. Modis GLC2000 Globcover
Land use
e.g. Census statistics: agriculture, forestry, grazing, settlements
e.g. national economic data (SNA)
CONSISTENCY
! CONSISTENCY !
CONSISTENCY
Flows
Ecosystem flows
DGVMs: GPP, NPP, Respiration, water, nutrients
Inputs - Outputs
(Census) Statistics: agriculture, forestry, grazing
Socioeconomic models

10. The HANPP framework: Data integration
Consistency
extents and flows: yields [=flow per area and year]
Prioritizing: correspondence of (national) land use census statistics and the (national) spatial extent more important than the accuracy of spatial information. But: how to deal with flawed census data?
Comprehensiveness
all ‘relevant’ land use types, inclusive “non-land-use” areas:
100% of each gridcell

11. Applications

12. Example I Linking ecosystem impacts and socio-economic drivers
HANPP
eHANPP consumption
Source: Erb et al, EE 2009a, Erb et al., 2009b

13. Example I Linking ecosystem impacts and socio-economic drivers
Source: Erb et al,EE 2009
Difference of „production“ and „consumption“ of „embodied HANPP“

14. Example I: Conclusions
A considerable flow: international “transfer” = 1.7 PgC/yr in 2000 [global deforestation: ~1.5 PgC/yr], increasing
Large, densely populated countries, which do not yet participate, will soon do so (e.g. China, India)
Drivers AND consequences of land use are global. No simple causal chains between drivers and associated impacts
Sustainability challenge:
High degree of international interdependence (vulnerability, resilience)
high risk of shifting the environmental burdens to distant locations and withdrawing it from environmental legislation
markets will not minimize burdens, as many ecosystems services have no price
 need for global monitoring and management of biomass demand & supply

15. Example II Global bioenergy potentials

16. A scoping study: Explore the scale and option space on basis of HANPP analyses
Systematic combination of existing (e.g. FAO) assumptions and 2 – 4 modulations on developments until 2050 of:
diets (4)
livestock efficiency (2)
agricultural yields (4)
cropland expansion (2)
 64 combinations (scenarios)

17. Results: Feasibility Analysis: 43 of 64 scenarios “feasible”
Not feasible
Probably feasible
Feasible
Highly feasible
For „feasible“ scenarios: bioenergy potential
on „free“ cropland
on high-quality grazing land
crop residues
Source: Erb et al., 2009c

18. Results Energy crop area [km²] (2.1 – (6.3) – 10.9 mio. km²)
Primary energy supply
Energy crop area [km²] (2.1 – (6.3) – 10.9 mio. km²)
Histogramm: feasible scenarios
Energy crop yield [gC/m²/yr]
Source: Erb et al., forthcoming
Haberl et al., 2010, COSUST
Haberl et al., 2011, Biomass & Bioenergy

19. Example II: Conclusions
Feeding a growing world population is – in principle - possible with ecologically sound agricultural production. Dietary levels will be most important.
Energy crop potentials – ‚conventional wisdom‘ needs to be reconsidered: Sustainability constraints are decisive:
Conservation / biodiversity
Subsistence agriculture, food security, etc.
GHG balance
Climate change impacts are poorly understood but could be strong
Bioenergy and globalization: Largest bioenergy potentials in Subsaharan Africa and Latin America: Caution – problem shifting!
‚Cascade utilization‘ – focus on recycling, re-use and efficiency improvement of biomass flow-chains

20. Conclusions: HANPP studies illustrate
Link land use – land cover is complex: no easy look-up table.
Spatial seggregation between appropriation and consumption: Issues of scale, governance: drivers as well as consequences of land use are global. Important for the construction of causal chains
Future biomass demand-supply: Options/potentials for sustainable biomass utilization are limited – requires integrated perspectives

21. Data challenges...
Land-use assessments require land-cover and additional (‚socio-economic‘) information
Many socio-economic drivers, mechanisms, processes of LU (change) and their impacts are not (yet) well documented. Basic research (still) required. List of EHV not ready yet.
Links to MaB (UNESCO), LTER-LTSER
The spatial and temporal scales of natural and socioeconomic processes are different
Increasing spatial resolution is only a partial solution: the gain in detail allows to better describe LC, but contextual information is required to assess LU; social systems are not organized in grids
Move beyond the S-o-A in LU-LC data:
 consistency and comprehensiveness
 abandon “hybrid”, ambiguous legends
 complement “dominance” classes or “discrete” classification schemes with continuous parameters. Gradients are equally important, for LC and LU
move beyond “agriculture”, “deforestation”, and “urban” land use
land management is key

22. ... and opportunities Data gaps/deficits are ubiquitous:
missing socio-economic data
flawed, incomplete census data
...and RS can contribute
forestry (used vs. unused forests, forest degradation)
grazing (intensity, spatial pattern of grazing, biomass harvest through grazing; effects of grazing)
cropland fallow (where, frequency)
rural infrastructure
soil/vegetation degradation (where? how much land? how intensive?)
()NPP, ()Biomass stocks
 yield the mutual benefits of combining RS data and “ground data”

23. Thank you for your attention!
The End
Thank you for your attention!
Further information/maps/data:
ERC Start Grant LUISE

25. Explore the scale and option space: a NPP perspective
Solid consistent empirical data-bases for 2000
Land use: Consistency between pixels (5 min, 10x10 km) and statistical data at country level (cropland and woodlands according to FAO, FRA und TBFRA). Erb et al J. Land Use Sci. 2,
National biomass balances : Production and consumption of biomass: Feed balances, processing losses, trade, incl. trends Krausmann et al Ecol. Econ. 65,
HANPP: Spatially explicit integration of NPP flows (LPJ-DGVM) and anthropogenic biomass flows (5 min, 10x10 km). Haberl et al., Proc. Natl. Acad. Sci. 104,
NPPact
Harvest

26. Grazing
livestock grazing is the largest fraction of the global biomass harvest (32%), a major driver of the human transformation of terrestrial ecosystems
Statistics comprise only market feed – no information on grazed biomass available. “Grazing Gap” must be modelled as difference between demand & market feed supply
very loose relation of land use and land cover (occurs in almost all ecosystems (hampers application of remote sensing techniques)
Census statistics are of limited practicability, inconsistent, heterogenous definitions (e.g. artificial grasslands vs. natural grasslands)
Grazing Gap
Source: Krausmann et al. Ecological Economics 2008

27. Data Gap: grazing land
‚Result‘ Remaining area = Grazing land

28. Estimates on global grazing lands

29. Grazing land Russ Fed. India Egypt Finnland Norway China Mexico Brazil
Saudi Arabia
Western Sahara
Yemen

30. consequences of land use: Biodiversity Species richness is well correlated with NPPt – indirect support for HANPP/biodiversity hypothesis
 = 0.708
Case study 1: Correlation between NPPt and autotroph species richness (5 taxa) on 38 plots sized 600x600 m, East Austria
Haberl et al., 2004, Agric., Ecosyst. & Envir. 102, p213ff
Case study 2: Correlation between NPPt and breeding bird richness in Austria, 328 randomly chosen 1x1 km squares.
Haberl et al., Agric., Ecosyst. & Envir. 110, p119ff
Case study 3: Correlation between NPPt and vertebrate richness in the Americas, 10,000 randomly chosen 5min gridcells
Haberl et al., forthcoming