Household modeling for ex-ante
evaluation and targeting of climate
smart agriculture
Mario Herrero
CCAFS Science Meeting, Copenhagen, May 2012
At ILRI: Mariana Rufino, Mark van Wijk, Carlos Quiros
CCAFS theme leaders: Philip Thornton (funding/strategy),
Jim Hansen, Lini Wollenberg, Andy Jarvis
CCAFS Regional Coordinators (funding): James Kinyangi
(EA), Robert Zougmore (WA), Pramod Aggarwal (SA)
CG centres: IWMI, ICRISAT, CIMMYT, ICRAF (for now)
Universities and Research Centres: Wageningen,
Hohenheim, Tasmania, Oregon, Washington State (for now)
This work involves many others
CCAFS engaged heavily in analysing regional and global
impacts on agriculture and exploring future pathways of
agricultural development through scenario analysis
Considerable work on adaptation and mitigation practices at a
local level
Ex-ante assessment and targeting: what might work where
and how this might change depending on the notions of the
future
Household modelling: offers the possibility of helping with
these issues.
A well established area, large community of scientists
Not widely used in climate change activities yet.
Background
W. Africa 1966 – pastoral system 2004 – crop-livestock system
Systems and livelihoods in transition: the target is moving! Can we ensure that the next transition is sustainable, equitable and helps
feed the world?
National
Production
Mixed
rainfed
temperate
Mixed
rainfed
humid
Mixed
rainfed
arid
2030 2050 2030 2050 2030 2050 2030 2050
Burundi 9.1 9.1 14.4 18.1 -1.8 -8.8 - -
Kenya 15.0 17.8 33.3 46.5 -4.6 -9.8 -1.1 -8.4
Rwanda 10.8 14.9 13.4 18.8 5.4 3.6 1.1 2.7
Tanzania -3.1 -8.1 7.5 8.7 -1.6 -6.4 -5.1 -11.1
Uganda -2.2 -8.6 4.9 3.1 -4.6 -12.9 -1.1 -6.3
A game of winners and losers…at all scales Simulated percentage maize production changes to 2030 and 2050, by
country and system
Mean of 4 combinations of GCM and emissions scenario
Thornton et al. (2010)
Winners
Losers
Understanding complex systems
Herrero et al, Science (2010)
J F M A M J J A S O N D
Rainy Dry Dry
Groundnuts
Yams
Sorghum
Grazing Critical Crop
residue
Cut &
Carry
Prot. & Ene.
deficit Food security
Energy
deficit Family’s
nutrition
Feeding calendar
Cropping calendar
Weather calendar
Monthly calendar of different activities of the system
Cash demands Very
high Low High High Low Lo
w
High High
Wa, Upper West, Ghana
Gonzalez-Estrada et al. 2006
0
0.5
1income
food security
GHGwater use
external inputs
mixed
pastoral
There are always trade-offs Different practices…
Different farming systems….
Plot Crop
Homestead Enset, coffee, kale, sweet potato, maize
Plot1 Maize
Plot 2 Sweet potato, wheat
Plot 3 Maize
Plot 4 Sweet potato
Plot 5 Barley
Plot 6 Maize
Plot 7 Unimproved pasture
What’s the likely impact of alternative
interventions?
Site: Sodo, Ethiopia
Profit: Birr 2,381/yr Livestock: 1 cow, 1 ox
Site: Sodo, Ethiopia
Critical Adequate
Current
management
Food security
Cow feeding
Labour/capital
Cash
Soil fertility
Food security
Cow feeding
Labour/capital
Cash
Soil fertility
Site: Sodo, Ethiopia
Critical Adequate
Intervention 1
Food crop Application of fertilizer to
maize plots.
Food security
Cow feeding
Labour/capital
Cash
Soil fertility
Site: Sodo, Ethiopia
Critical Adequate
Intervention 2
Feed crop Replace native grassland
with improved pasture.
Food security
Cow feeding
Labour/capital
Cash
Soil fertility
Site: Sodo, Ethiopia
Critical Adequate
Intervention 3
Food/feed crop Introduction of cowpea in
the system.
Food security
Cow feeding
Labour/capital
Cash
Soil fertility
Site: Sodo, Ethiopia
Critical Adequate
Intervention 4
Food crop Subsidy for barley
production
Adaptation, risk management and
mitigation options will depend largely
on how we shape the world
• Several options exist though largely dependent on our vision of world development and how it plays out in different regions
• essential to link household modelling to scenarios of change
• Different paradigms of agricultural development (industrial vs pro-poor smallholders, large vs family farms)
• Globalisation and trade patterns
• Consumption patterns
• Carbon constraints
• Roles and incentives for technology adoption
• Growth in other sectors
• Power relationships
Global Scenarios
Regional Scenarios
Farmer/village
perspectives Action research
Participatory
scenario building
Global visioning
activities
Global impacts
modelling
Regional impacts
modelling
Household &
community
impacts modelling
Linking research at different levels
Thornton et al 2012
Ex-ante analysis and targeting of
options
• Studying livelihoods transitions
• Targeting the vulnerable (winners and losers)
• Which options could fit in which systems under which conditions?
• How upscalable to broader recommendation domains/regions?
• How robust are options across scenarios and farming systems
• Priority setting for investments (how many farmers, what areas, how much?)
• Mitigation / adaptation synergies
CCAFS activities in relation to
household modelling
Household data collection in the CCAFS sites
• Development of simplified, but robust and standard data collection protocols
• Collecting detailed information from representative farming systems from the CCAFS sites
• 150-200 households per site: approx 3000 surveys
• Data collection during 2012
• Statistical analysis and modelling of adaptation, risk management and mitigation options from 2012 (with input from centres and themes)
• Funded by Theme 4 and the regions
All survey materials and data at ccafs.cgiar.org/resources/baseline-surveys Dataverse at dvn.iq.harvard.edu Adaptation and Mitigation Knowledge Network at amkn.org
CCAFS sites in West Africa, East Africa and South Asia
Data collection + household modeling protocol :
Climate
Family structure/gender
Land management
Livestock management
Labour allocation
Family’s dietary pattern
Farm’s sales and expenses, income
Herrero et al 2007
A review of farm household modelling with a focus on climate change
• A systematic review of 16000 thousand references
• Covered long term adaptation, risk mangement and mitigation
• Covered diverse modelling techniques (LP, agent based models, simulation, others)
• Identification of useful tools for CCAFS work
• Integrated models using more than one modelling technique more suitable for CC questions
• Engagement with other modellers
Workshop: Farm-household Modeling with a focus on Food security, Climate change adaptation, Risk management and Mitigation: a way forward
• Goal: to identify current potential of and weaknesses in farm- and household-level models, and laying out practical pathways to improve these models to address CCFAS systems questions
Amsterdam, The Netherlands 23 to 25 April 2012
Conclusions from the workshop • Modelling approaches are available to address household-level
questions. This needs to include higher levels of integration to capture key drivers.
• It is possible to analyze household-level questions related to climate change in a reasonable short time (6 months to 1 year) span with the existing tools.
• Activities to develop repositories for models and data are urgently needed to increase further development of models and make better use of existing knowledge.
• A team of modellers with different expertise is needed to address questions related to climate change agriculture and food security.
• The research questions must lead to the suite of models and expertise needed. Not much model development is needed.
Next steps
• Continue the data collection in the regions
• Develop suitable databases and repositories for the information for open access by CCAFS and partners
• Data analyses
• Linking with the international household modelling teams to prepare potential tools for analysis
• Extensions to community-level modelling
• Maybe fund some additional development as needed
Conclusions
• Household modelling can play a key role in the ex-ante evaluation and targeting of adaptation and mitigation work of CCAFS
• It can help link work at multiple-scales: for example the scenarios work in the regions to impacts and options in different types of farming systems
• It can lead to robust multi-centre and multi-theme work by exploiting complementary skills to solve complex problems
• ….and provide realistic, sometimes simplified, answers
Thank you