1. Building capacity to provide useful climate information: Lessons and insights from agricultural decision-making in Argentina Guillermo Podestá University of Miami, Rosenstiel School

2. Main Points - 1 An effective climate information system (or climate service) will not develop spontaneously An effective climate information system (or climate service) will not develop spontaneously A climate information system has to be informed and supported by an appropriate research program throughout its initial phase A climate information system has to be informed and supported by an appropriate research program throughout its initial phase Research supporting climate information systems must evolve: Research supporting climate information systems must evolve: Exploratory  Pilot  (Semi)operational

3. Main Points - 2 Climate information must include a triad of components Climate information must include a triad of components Historical data and statisticsHistorical data and statistics Recent climate conditionsRecent climate conditions Forecasts of regional climate scenariosForecasts of regional climate scenarios False antinomy between research in the natural and social sciences False antinomy between research in the natural and social sciences

4. Funding Sources Environment and Sustainable Development Human Dimensions of Global Change Regional Integrated Science & Assessments Methods & Models for Integrated Assessment Biocomplexity in the Environment: Dynamics of Coupled Natural & Human Systems Initial Science Program – Phase 2

5. Motivation - 1 Enhanced technological capabilities Enhanced technological capabilities Routine global climate observationsRoutine global climate observations Faster computing capabilitiesFaster computing capabilities Better communications, access to dataBetter communications, access to data Better understanding of climate system Better understanding of climate system Higher awareness of climate influence on some human activities Higher awareness of climate influence on some human activities RESULT: Increased demand for climate information RESULT: Increased demand for climate information

6. Motivation - 2 What we expected … Climate Information Societal Use

7. What we got … Climate Information Societal Use

8. Motivation - 3 A climate information system has to be supported by an appropriate research program throughout its initial phase (i.e., beyond initial design) A climate information system has to be supported by an appropriate research program throughout its initial phase (i.e., beyond initial design)

9. Supporting Research Projects supporting climate services should evolve through various stages Projects supporting climate services should evolve through various stages Exploratory Stage Pilot Stage Operational Stage [Hansen, 2002]

10. Why Argentina? - 1 Pampas are among the most important agricultural regions in the world Pampas are among the most important agricultural regions in the world Agriculture accounts for more than half of exports Agriculture accounts for more than half of exports Argentina-Brazil-Paraguay have increasing weight on soybeans market Argentina-Brazil-Paraguay have increasing weight on soybeans market Production systems similar to those in US Production systems similar to those in US

11. Why Argentina? - 2 Marked climate signals Marked climate signals Interannual variability (ENSO-related)Interannual variability (ENSO-related) Decadal variabilityDecadal variability Buenos Aires harbor flooded by aquatic plants from Parana Basin 2002 flood in City of Santa Fe

12. Research Stages - Exploratory Exploratory Stage Pilot Stage Operational Stage Associations between climate and agriculture? Associations between climate and agriculture? Statistical analyses of historical dataStatistical analyses of historical data Simple modelingSimple modeling Issue scoping: surveys, focus groupsIssue scoping: surveys, focus groups Partners: academic institutions or governmental research agencies Partners: academic institutions or governmental research agencies

13. Historical Maize Yields & ENSO

14. Modeled Maize Yields ENSO Forecast Translation Process Models Distribution of Outcomes

15. ENSO Risk Assessment What is the range of outcomes of an ENSO event on maize production, given current management? What is the range of outcomes of an ENSO event on maize production, given current management?

16. Exploratory Surveys and Focus Groups Great interest in climate information Great interest in climate information Ignorance about local climatology and “normal” variability Ignorance about local climatology and “normal” variability Ignorance about capabilities / limitations of forecasts Ignorance about capabilities / limitations of forecasts 1997/98 El Niño was major turning point in awareness of climate 1997/98 El Niño was major turning point in awareness of climate Importance of context Importance of context

17. Research Stages - Pilot More sophisticated, realistic modeling More sophisticated, realistic modeling Risk management studies Risk management studies How can we react to climate info?How can we react to climate info? Economic, social dimensions Economic, social dimensions Understanding decision-making processUnderstanding decision-making process Value of climate informationValue of climate information Exploratory Stage Pilot Stage Operational Stage

18. Managing Climate Risks - 1 Land Allocation Planting DateFertilization Rate Maize Soybean Early Planting Late Planting High Fertilization Low Fertilization … … … … OGP-CSI funds: Decision maps for maize (in press) Maps for soybean being completed

19. Managing Climate Risks - 2 Climate Scenario A … … Climate Scenario B Preceding Climate conditions Agronomic Decisions Decision Outcomes Decision Context

20. Mapping Realistic Decisions Owned Land 1/3 Maize 1/3 Wheat Soybean 1/3 Soybean Rented Land 1/3 Soybean 1/3 Soybean 1/3 Soybean Land Allocation Decisions Almost 50% of agriculture in the Pampas is done on rented land!!!

21. Mapping Realistic Decisions - 2 OGP Human Dimensions funding (Weber): Exploration of alternative objective functions OGP Human Dimensions funding (Weber): Exploration of alternative objective functions What farmers are really trying to achieve… What farmers are really trying to achieve… Standard economic models often consider only maximization of utility Standard economic models often consider only maximization of utility Wrong assumed objective may imply wrong advice… Wrong assumed objective may imply wrong advice…

22. Mapping Realistic Decisions - 3 Wealth Utility Reference Point Income Gains Losses Utility TheoryProspect Theory

23. Mapping Realistic Decisions - 4 Wheat Soybean Soybean

24. Mapping Realistic Decisions - 5 Increasingly important role for stakeholders (e.g., farmers and their technical advisors) Increasingly important role for stakeholders (e.g., farmers and their technical advisors) Partnerships are first step towards operational status Partnerships are first step towards operational status Difficult to get stakeholders’ “buy-in” Difficult to get stakeholders’ “buy-in” Continuity in interactions (and funding!)Continuity in interactions (and funding!) Commitment to produce usable info/knowledgeCommitment to produce usable info/knowledge

25. Strategic Partnerships AACREA: non-profit farmers’ organization AACREA: non-profit farmers’ organization Groups of 8-12 farmers Groups of 8-12 farmers About 150 groups in Argentina About 150 groups in Argentina “Early adopters” “Early adopters” “De facto” extension functions “De facto” extension functions Large multiplicative effect Large multiplicative effect

26. Research Stages – Operational (?) Research topics: broad spectrum, but highly specific issues Research topics: broad spectrum, but highly specific issues Strategic partnerships crucial: with BOTH operational agencies AND stakeholder groups Strategic partnerships crucial: with BOTH operational agencies AND stakeholder groups Exploratory Stage Pilot Stage Operational Stage

27. Communicating Climate Info New information is interpreted in the context of existing knowledge and beliefs New information is interpreted in the context of existing knowledge and beliefs Before communicating new information, find out what people know… Before communicating new information, find out what people know… What they already knowWhat they already know What they do not knowWhat they do not know What they think they know, but is incorrectWhat they think they know, but is incorrect

28. “Mental Models” of Climate Support from OGP-ESD program Support from OGP-ESD program Open-ended, free-flowing interview (CMU methodology) Open-ended, free-flowing interview (CMU methodology) About 60 farmers About 60 farmers Climatically optimal region (Pergamino), long ag historyClimatically optimal region (Pergamino), long ag history Climatically marginal region (Pilar), recent ag historyClimatically marginal region (Pilar), recent ag history

29. “Mental Models” Results Local climate is different from that of nearby (50 km) areas Local climate is different from that of nearby (50 km) areas Good yields in better soils are confused with better climate conditions Good yields in better soils are confused with better climate conditions “Reverse” association between climate experienced and ENSO phase “Reverse” association between climate experienced and ENSO phase “If it is a dry year, a Niña must be occurring…”“If it is a dry year, a Niña must be occurring…” Opposite ENSO phases occurring simultaneously in different regions Opposite ENSO phases occurring simultaneously in different regions Personal delivery of climate information induces higher trust Personal delivery of climate information induces higher trust

30. Institutional Issues Support from OGP-ESD program Support from OGP-ESD program Role of “boundary organizations” in production/dissemination of climate info Role of “boundary organizations” in production/dissemination of climate info Communication between producers/usersCommunication between producers/users Translation of info (relevance)Translation of info (relevance) Mediation (transparency, legitimacy, credibility)Mediation (transparency, legitimacy, credibility) [Cash et al PNAS 2003]

31. Institutional Issues - 2 Argentine Meteorological Service Argentine Meteorological Service Focused on weather predictionFocused on weather prediction Lack of links with users (no feedback!)Lack of links with users (no feedback!) Unprepared/unfunded to provide climate servicesUnprepared/unfunded to provide climate services Compartmentalized structureCompartmentalized structure Multiple bulletins, data products Multiple bulletins, data products

32. What we are proposing… Strategic partnerships with sectoral boundary organizations Strategic partnerships with sectoral boundary organizations Consolidated climate information products Consolidated climate information products Climatological infoClimatological info Recent conditionsRecent conditions Seasonal forecastsSeasonal forecasts “Pyramidal” organization of info “Pyramidal” organization of info Short summary More detail Detailed supporting info Model output, raw data

33. Climate Information Components Historical data and statistics Seasonal climate forecasts Recent climate conditions Also: “plausible” decadal scenarios, longer scales?

34. Why historical information? Lack of knowledge about local climatology Lack of knowledge about local climatology What is “normal”?What is “normal”? Recent arrivals to agricultureRecent arrivals to agriculture Greater memory of recent eventsGreater memory of recent events How to interpret seasonal forecasts How to interpret seasonal forecasts Boundaries between terciles?Boundaries between terciles?

35. Why diagnostic information? Provides context for decision- making Provides context for decision- making Refine previously- made decisionsRefine previously- made decisions Helps interpret forecastsHelps interpret forecasts Relevant span is sector-dependent Relevant span is sector-dependent

36. Summary - 1 An effective climate information system (or climate service) will not develop spontaneously An effective climate information system (or climate service) will not develop spontaneously A climate information system has to be informed and supported by an appropriate research program throughout its initial phase A climate information system has to be informed and supported by an appropriate research program throughout its initial phase Research supporting climate information systems must evolve: Research supporting climate information systems must evolve: Exploratory  Pilot  (Semi)operational

37. Summary - 2 Climate information must include a triad of components Climate information must include a triad of components Historical data and statisticsHistorical data and statistics Recent climate conditionsRecent climate conditions Forecasts of regional climate scenariosForecasts of regional climate scenarios False antinomy between research in the natural and social sciences False antinomy between research in the natural and social sciences