The Chinese Academy of Agricultural Sciences (CAAS) and the International Food Policy Research Institute (IFPRI) jointly hosted the International Conference on Climate Change and Food Security (ICCCFS) November 6-8, 2011 in Beijing, China. This conference provided a forum for leading international scientists and young researchers to present their latest research findings, exchange their research ideas, and share their experiences in the field of climate change and food security. The event included technical sessions, poster sessions, and social events. The conference results and recommendations were presented at the global climate talks in Durban, South Africa during an official side event on December 1.
Liu Yuan — Crop yields impacted by enso episodes on the north china plain 195...
Peter Verburg — Analysis and modelling of land use change in relation to food security and climate change
1. Analysis and modelling of land use change in
relation to food security and climate change
Peter Verburg
Beijing, 7-8 nov 2011
2. Rationale
Expansion of agricultural area
Intensification of land use
Climate change
All processes happen at same time
systems
depending land use, environmental, socio-
Food/Feed/Fibre/Energy
economic and governance conditions
demand Import from other areas
Change in consumption
pattern
2
3. Rationale
Expansion of agricultural area
Intensification of land use
systems
Food/Feed/Fibre/Energy
demand Import from other areas
Change in consumption
pattern
3
6. Global scenarios of land cover change
Macro-economic models (GTAP/IMPACT) and land
allocation model (IMAGE, LandShift, CLU-Mondo)
Spatial resolution often 50x50 km
One dominant land cover type per pixel
Economic models assume ‘rational’ behaviour
World region economic land demands are downscaled by
simple rules: land suitability, distance to existing land cover
types
Variation in socio-economic and cultural factors disregarded
6
11. Landscapes are mosaics
Composition of landscapes is
important (biodiversity, carbon,
ecosystem services)
Representation by dominant
land cover types is incorrect
at all (feasible) spatial
resolutions
Mosaics should be
represented explicitly
11
12. Regional differences
350
Scenario
300
250
Agricultural area
Global models
200
150
CLUE-Scanner
European scale 100
models
50
0
Trade-off analysis
-50
Africa Asia C&SAmer EU27 NAFTA World
Reference Biofuel, w/o EU Biofuel, with EU
Verburg et al., 2008 Annals of Regional Science
Banse et al., 2010 Biomass and Bioenergy 12
17. Land cover developments
Main trends:
Abandonment of marginal agricultural areas decrease of
agricultural area
Urbanization Loss of most productive agricultural lands
Peri-urban development demand for ecosystem services
besides food production: recreation etc.
Expansion of agriculture in other regions
Intensification of agricultural production on remaining area
17
18. 2000 2050
Ecosystem Service Assessment, CAS 18
19. Land cover developments
Main trends:
Abandonment of marginal agricultural areas decrease of
agricultural area
Urbanization Loss of most productive agricultural lands
Peri-urban development demand for ecosystem services
besides food production: recreation etc.
Expansion of agriculture in other regions
Intensification of agricultural production on remaining area
19
20. Analyze effect of land
use change on
ecosystem services
Kienast et al., 2009
20
21. Land cover developments
Main trends:
Abandonment of marginal agricultural areas decrease of
agricultural area
Urbanization Loss of most productive agricultural lands
Peri-urban development demand for ecosystem services
besides food production: recreation etc.
Adaptation to climate change flood risk and adaptation
measures threaten most productive regions
Expansion of agriculture in other regions
Intensification of agricultural production on remaining area
21
24. Land cover developments
Main trends:
Abandonment of marginal agricultural areas decrease of
agricultural area
Urbanization Loss of most productive agricultural lands
Peri-urban development demand for ecosystem services
besides food production: recreation etc.
Adaptation to climate change flood risk and adaptation
measures threaten most productive regions
Expansion of agriculture in other regions
Intensification of agricultural production on remaining area
24
26. Rationale
Expansion of agricultural area
Intensification of land use
systems
Food/Feed/Fibre/Energy
demand Import from other areas
Change in consumption
pattern
26
27. Challenges
Data on land use intensity are limited
Drivers of intensification largely unknown
• Keys and McConnell, 2005 – meta-analysis of 91 case studies
> Drivers are context specific
> Drivers operate at different spatial/temporal scales/levels
Role of governance unclear
27
28. Challenges
Data on land use intensity are limited
Drivers of intensification largely unknown
• Keys and McConnell, 2005 – meta-analysis of 91 case studies
> Drivers are context specific
> Drivers operate at different spatial/temporal scales/levels
Role of governance unclear
28
31. Challenges
Data on land use intensity are limited
Drivers of intensification largely unknown
• Keys and McConnell, 2005 – meta-analysis of 91 case studies
> Drivers are context specific
> Drivers operate at different spatial/temporal scales/levels
Role of governance unclear
31
32. Drivers of agricultural intensity – Global scale
Crop specific yields,
Actual yield 5 arc-min
[Monfreda et al., 2008]
Frontier yield/ Stochastic frontier
yield gap production function
Reasons for Inefficiency factors /
inefficiency Multiple Regressions
Neumann et al., 2010 Agricultural Systems 32
33. Explaining global distributions of yield gab
Frontier production function
vi = noise
ui = inefficiency
xi = actual productivity
¤i = frontier productivity
Neumann et al., 2010 Agricultural Systems 33
34. Explaining global distributions of yield gab
• Determinants for the frontier yield:
– Temperature, PAR, precipitation, soil fertility constraints
• Determinants for deviation from the frontier yield
(=inefficiency effects):
– Irrigation, market accessibility, market influence,
agricultural population, slope
Neumann et al., 2010 Agricultural Systems 34
36. Accessibility
Labor
Accessibility
Irrigation Market influence
Accessibility Slope
Irrigation
Neumann et al., 2010 Agricultural Systems 36
37. Market influence
Irrigation
Irrigation Accessibility
Market influence Market influence
Market influence
Accessibility
Neumann et al., 2010 Agricultural Systems 37
38. Irrigation Irrigation
Labor
Market strength
Accessibility
Labor
Neumann et al., 2010 Agricultural Systems 38
39. Challenges
Data on land use intensity are limited
Drivers of intensification largely unknown
• Keys and McConnell, 2005 – meta-analysis of 91 case studies
> Drivers are context specific
> Drivers operate at different spatial/temporal scales/levels
Role of governance unclear
39
40. Global distribution of irrigation in farmland
Landscape systems:
Land cover
Land use
Livestock
People
Ecosystem services
Irrigated farmland
Rainfed farmland
Portmann et al., 2010
40
41. Variables at grid cell level
Variable name Description [unit]
Irrigation 1 if irrigation,
0 if rainfed
Slope Slope [%]
Discharge River discharge [mm/yr]
Humidity Humidity, calculated as precipitation [mm] /
potential evapotranspiration (PET) [mm/yr]
[index]
Evap Evaporation [mm/yr]
ET Evapotranspiration
[mm/yr]
Access Travel time to markets [hours]
Population Population density
[persons/km2]
41
42. Variables at country level
Variable name Description [unit]
Water Natural total renewable water resources [m3/yr/ha]
Political stability Likelihood that the government will be destabilized [index]
Control of corruption Control of corruption (the extent to which public power is
exercised for private gain) [index]
Government Quality of public and civil service and the degree of its
effectiveness independence from
political pressures [index]
GDP Gross Domestic Product per capita
[US$]
Democracy Level of institutionalized democracy
[index]
Autocracy Level of autocracy [index]
42
43. Multilevel analysis
Binary logistic model:
• Only grid cell level – no multiple levels
Multi-level Model 1:
• includes all independent biophysical grid cell variables (slope,
discharge, humidity, evaporation and ET)
• Includes country level
Multi-level Model 2:
• includes in addition to these variables the socio-economic
grid cell variables (access and population)
• includes country level variables (water, government
performance and government type)
43
45. Rationale
Expansion of agricultural area
Intensification of land use
systems
Food/Feed/Fibre/Energy
demand Import from other areas
Change in consumption
pattern
45
46. Import and land grabbing
Import: Land Grab:
Macro-economic modelling:
• Partial equilibrium models
• General equilibrium models
Land supply/demand
determines land price
Land supply mostly only
constrained by agro-
ecological suitability
46
47. Conclusion
Land cover and land use change are important drivers of
food security
Socio-economic and governance variables are important
and deserve more attention in global scale assessments
Current studies focus too much attention on biophysical
component of climate change
Local patterns of adaptation need to be accounted for in
global assessments
Knowledge available in the Land Science community may
help analysis of food security and climate change
47
48. The Global Land Project
IHDP and IGBP funding
AIMES
Knowledge, Learning and iHOPE Integrated History
Societal Change (KLSC)
(in preparation) of people on Earth
(led by AIMES). Co-sponsored by
PAGES and IHDP
48