Developing links with agricultural research centres

1
Developing links with agricultural research
centres
Andrew Challinor and Tim Wheeler
2
The CGIAR network

• In and for developing countries
• 16 centres across the globe
• Autonomous but collaborative
• Broad research base
• Germplasm bank

Con
3
Links with CGIAR centres ...

• to collate information needed for applications
  by agricultural research centres
• potential applications of forecast information
• forecast data when, scale, format,
• any other issues?

4
ICRISAT

• Location India and Niger
• Semi-arid tropics
• Groundnut
• Millet
• Sorghum

5
CIMMYT

• Location Mexico and Zimbabwe
• Maize and wheat
• Visited in Feb. 2002

6
CIAT

• Location Colombia
• Beans and cassava
• Visited Feb. 2002

7
IRRI

• Location The Philippines
• Rice
• Visited March 2002
• Oldest CG centre (F. 1960)

8
Overview of response

• Rainfall seen as the most important variable
• Spatial scale is an issue, especially in
  mountainous regions
• Probabilistic forecasts not always initially seen
  as useful
• Importance of drought / heat stress at critical
  crop stages

9
Country district field
Spatial scale
annual seasonal monthly daily
Leadt i me
Seed production
Risk assessment and crop scheduling
FEWS
Pests and disease control
10
Examples of applications of SF

• Planting of drought tolerant bean varieties
• Forecasting of dry spells near flowering
• Transplant vs. direct seeding for rice

• Rainfall-related cassava diseases
• Maize streak virus
• Bean breeding

11
(No Transcript)
12
Methodology for exploring applications of SF
General Circulation Model
weather forecast
Regional Climate Model
spatial parameters
Application model
model uncertainties
Large area model
crop forecast
output processing
13
Participatory research at CIAT
Oberthur (2002)
14
Combining local and formal knowledge
Local knowledge
10 farmers
8
Relevance
X
?
?
?
Govts.
Formal knowledge
X
World Bank
Accuracy
(Mintzberg)