Roncoli (2006) reviews ethnographic and participatory research methods, which involve intensive interaction with farmers or other stakeholders, that complement the quantitative economic research and assessment methods described in the preceding sections. This research provides rich insights into the cognitive and cultural landscape in which farmers' understanding of climate and climate information is grounded, and the decision-making process and environment that defines options, constraints, and outcomes.
Taxonomies, seasonal calendars, and ranking matrices of climate events have helped researchers to identify salient attributes that structure people's perceptions and experiences of climate (Orlove 2004; Ziervogel and Calder 2003), and highlight discrepancies between the ways farmers think about climate and the ways forecasts are formulated. For example, the seasonally averaged timeframe and regional scale typical of operational climate forecasts do not match farmers' concern with short-term, localized events (Hansen et al. 2004) or the duration and distribution of rainfall within the growing season (Roncoli et al. 2004). Narrowing this gap will necessitate bringing scientists' own cultural models into the analytical focus to understand how they shape the research agenda.
Ethnographic methods have shown that farmers around the world have a diverse repertoire of shared and specialized forecasting knowledge based on environmental observations and ritual practices. Field data show that farmers do not generally rely on a single indicator, but rather combine signs that arise at different times from various sources (Roncoli et al. 2002; Luseno et al. 2003). Although a dearth of long-term data series for local indicators, such as wild plants or insects, has hampered efforts to assess the validity of indigenous forecasts, innovative cross-disciplinary research, drawing from ethnographic, agronomic and atmospheric data, has established that Andean farmers' rainfall predictions based on the visibility of particular stars have a natural explanation and some skill (Orlove et al. 2002).
Risk communication research indicates that people's grasp of probability is often imperfect, as personal experience and communication practices can lead to cognitive biases. Yet interactive exercises during fieldwork and workshops have shown that farmers' ability to interpret probability and use forecasts in decision-making can improve with interaction and experience (Patt 2001). Field research has also shown that farmers' expectations and assessments of accuracy may differ for traditional and scientific forecasts (Nelson and Finan 2000).
The role of climate forecasts in rural livelihoods hinges on household vulnerability to climate risk. While quantitative methods make it possible to measure and compare levels of vulnerability, qualitative approaches provide valuable insights into subtler dimensions of vulnerability. In-depth interviewing and participant observation has revealed how gender, ethnicity, and caste can limit access and use of climate forecasts among African farmers (Roncoli et al. 2004). By combining participatory methods with quantitative surveys and agent-based modeling, a study among farmers in Southern Africa showed that, while wealthy households realized greater yield gains, climate forecasts benefited poor farmers the most by reducing the likelihood of food shortage (Ziervogel et al. 2005).
Was this article helpful?