Karlheinz Erb | The HANPP framework | Hamburg | February 10, 2010 | 1 Human Appropriation of NPP (HANPP) An accounting framwork for analysing land use.

Karlheinz Erb | The HANPP framework | Hamburg | February 10, 2010 | 1

Human Appropriation of NPP (HANPP)An accounting framwork for analysing

land use processes in the Earth system

Karlheinz ErbInstitute 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 assessmentsGeneva, Feb. 3, 2011

ERC Start Grant 263522 LUISE


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


State-of-the-art of LU science

– Classification systems creating nominal-scale data– Focus on land cover (biophysical structures, ecological systems)– Focus on forest / non-forest dynamics– Strategy: increasing spatial resolution

From current mainstram land-use research...

+

...towards an integrated understanding of land use

+ Focus on society-nature interactions+ Broad range of land uses+ Continuous (rational) scales+ Explicitly addressing a wide range of spatial scales

SocietyEcosystems


Bridging disciplinary boundaries: the integrated land system

SocietyEcosystems

Outputs - Benefits

Inputs - Investments

management

Long-TermC-Storage

Plant biomass

CO2

Medium-TermC-Storagee.g. Soil C

GPP

NPP

Respiration

new

old

Steffen et al. Science 1998

Input Output

Socio-economic system

Economic Processing

Stocks

Input Output

Socio-economic system

Economic Processing

Stocks

Matthews et al. 2000


HANPP – the ‚human appropriation of net primary production‘

Society

Outputs - Benefits

Inputs - Investments

Potential NPP

Actual NPP

NPP remaining after harvest

dN

PP

LC

NP

Ph

HA

NP

P

Ecosystems

Change

induced

through

landuse

naturalecosystem

managedecosystem


Data integration

NPP0: LPJ-DGVM

Non-used areas

Irrigation Degradation

Erb et al., J of Land Use Science, 2007

NPPact

NPPh


HANPP%:Aggregated effect ofland use and harvest<< 24% >>

Result: Global HANPP 2000NPPLC%: Productivity changesdue to land coversions<< 10% >>

Source: Haberl et al PNAS 2007Krausmann et al., 2008

Forest11%

Cropland49%

Human induced

fires7%

Built-up land4%

Grazing land29%

Food13,7%

Market feed8,4%

Seed0,7%

Other uses11,1%

Industrial Wood6,2%

Fuelwood9,8%

Grazing, fodder50,0%

Land use activities Biomass consumption


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


HANPP data integration: ‚old‘ and ‚new‘ challenges

Ecosystems Socioeconomic Systems

Area

Land covere.g. Modis

GLC2000Globcover

Land usee.g. Census statistics:

agriculture, forestry, grazing, settlements

e.g. national economic data (SNA)

Flows

Ecosystem flows DGVMs: GPP, NPP,

Respiration, water, nutrients

Inputs - Outputs (Census) Statistics:

agriculture, forestry, grazing

Socioeconomic models

CONSISTENCY

! CONSISTENCY !

CONSISTENCY

0

0,5

1

1,5

2

2,5

3

1950 1960 1970 1980 1990 2000 2010 2020

[196

1 =

1]

Area

Yield

Production

0

0,5

1

1,5

2

2,5

3

1950 1960 1970 1980 1990 2000 2010 2020

[19

61

= 1

]

Area

Yield

Production


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


Applications


HANPP

Example ILinking ecosystem impacts and socio-economic drivers

0,0

1,0

2,0

3,0

4,0

5,0

6,0

HANPP NPPh Used extraction Finalconsumption of

biomass

[t D

M/c

ap

/yr]

Ecosystem Impacts Socio-economic drivers

conversionlosses

harvestlosses/damages

productivitylosses

Erb

et

al. 2

009conversion

losses


productivitylosses

conversionlosses


productivitylosses

Erb

et

al. 2

009

eHANPP consumption

Source: Erb et al, EE 2009a, Erb et al., 2009b


Difference of „production“ and „consumption“ of „embodied HANPP“

Source: Erb et al,EE 2009

Example ILinking ecosystem impacts and socio-economic drivers


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


Example IIGlobal bioenergy potentials


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)


Results: Feasibility Analysis: 43 of 64 scenarios “feasible”

Probably feasible

Feasible

Highlyfeasible

Not feasible

Source: Erb et al., 2009c

For „feasible“ scenarios: bioenergy potential1) on „free“ cropland2) on high-quality grazing land3) crop residues


ResultsEnergy crop area [km²] (2.1 – (6.3) – 10.9 mio. km²)

Energy crop yield [gC/m²/yr]

0

2

4

6

8

10

12

14

20-40 40-60 60-80 80-100

100-120

120-140

140-160

160-180

180-200

Bioenergy potential [EJ/yr]

[#]

0

20

40

60

80

100

120

140

160

Primary crops Residues

[EJ/

yr]

Histogramm: feasible scenarios

Source: Erb et al., forthcomingHaberl et al., 2010, COSUST Haberl et al., 2011, Biomass & Bioenergy

Primary energy supply


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


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


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


... 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”


Thank you for your attention!

The End

Further information/maps/data:http://www.uni-klu.ac.at/socec/

ERC Start Grant 263522 LUISE



Explore the scale and option space: a NPP perspective

NPP0

NPPact

Harvest

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. 2007. J. Land Use Sci. 2, 191-224

• National biomass balances : Production and consumption of biomass: Feed balances, processing losses, trade, incl. trends 1960-2000.Krausmann et al. 2008. Ecol. Econ. 65, 471-487.

• HANPP: Spatially explicit integration of NPP flows (LPJ-DGVM) and anthropogenic biomass flows (5 min, 10x10 km).Haberl et al., 2007. Proc. Natl. Acad. Sci. 104, 12942-12947.


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)

-

1

2

3

4

5

6

7

8

Market feed Feedstuff demand

[10^

6 tD

M/y

r]

Crop residues

Fodder Crops

Grazing Gap

Source: Krausmann et al. Ecological Economics 2008


Data Gap: grazing land

‚Result‘ Remaining area = Grazing land


Estimates on global grazing lands

0

10

20

30

40

50

60

70

80R

am

an

kutty

et a

l., 2

00

8

Kle

inG

old

ew

ijk2

00

1

FA

O 2

00

4,

Kle

inG

old

ew

ijk2

00

7

Wh

itta

ker

an

dL

ike

ns

19

73

Ajta

y e

t al

19

79

Erb

et a

l.2

00

7

Wh

ite e

t al

20

00

Ols

on

et a

l1

98

3

Me

yer

an

dT

urn

er

19

92

[mio

km

²]


Grazing land

1E-01

1E+00

1E+01

1E+02

1E+03

1E+04

1E+05

1E+06

1E+07

1E-01 1E+00 1E+01 1E+02 1E+03 1E+04 1E+05 1E+06 1E+07

FAO permanent pasture

Gra

zin

g l

and

(su

btr

acti

ve a

pp

roac

h)

Russ Fed.

India

Egypt Finnland

Norway

China

Mexico

Brazil

Saudi Arabia

Western Sahara

Yemen


consequences of land use: Biodiversity Species richness is well correlated with NPPt – indirect support for HANPP/biodiversity hypothesis

1 2 3 4 5 6 7 8 910 202010

-2

4x10-2

Y = -1.975 +0.485 X R² =0 .549, p < 0.0001

i)

all h

eter

otro

phs

NPPt

0.1 1 10

1

10

100

Y =1.32916+0.69916 X-0.22962 X2

Adj. R2 = 0.69bree

ding

bird

spe

cies

richn

ess

NPPt [MJ/m²*a]

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., 2005. Agric., Ecosyst. & Envir. 110, p119ff

0 200 400 600 800 1000 1200 1400

NPPt [gC/m²/yr]

0

200

400

600

800

1000

Nu

mb

er o

f S

pec

ies

Case study 3: Correlation between NPPt and vertebrate richness in the Americas, 10,000 randomly chosen 5min gridcells

Haberl et al., forthcoming

= 0.708



Karlheinz Erb | The HANPP framework | Hamburg | February 10, 2010 | 1 Human Appropriation of NPP (HANPP) An accounting framwork for analysing land use.

Documents

hanpp framework hamburg

hanpp example

hanpp data integration

global hanpp amounts

erb et

summary of results hanpp

consequences of land

consistency slide