**4.1. CASE STUDY 1: Using probabilistic seasonal forecasting to improve farmers' decision in Kaffrine, Senegal (Ousmane Ndiaye, Robert Zougmoré, Jim Hansen, Aida Diongue, El Hadji Seck)**

Although agriculture and pastoralism occupy 80 per cent of the population in the Sahel, climate information is not yet widely integrated into farm management decision systems. However, many efforts have been made in the region to produce climate information such as the yearly climate outlook forum preceeding the incoming rainy season57. Yet, this hasn't benefitted the user community, particularly the most vulnerable to climate variability and change. This paper documents one ongoing demonstration project in Kaffrine, Senegal, within the peanut growing basin, where rural communities, policy-makers and relevant institutions are testing the use of probabilistic seasonal forecasts for managing climate risk. The process, from training the farming community to evaluating the use of the forecast information, is outlined.

#### *4.1.1. Background*

496 Risk Management – Current Issues and Challenges

taken, and the rules are negotiated and made55,56.

predict.

social vulnerability indicators with physical variables across timescales, (2) embracing risk communication as an interactive social process and, (3) supporting governance of a collaborative framework for early warning across spatial scales46. Forecasts need not be perfect to make early warning useful. For longer-term EWS, it is also important to note that although a trend in the drought-based indicators may serve as a warning, the actual point of transition or threshold (e.g. dune mobilization) to increased severity remains difficult to

Traditional warnings, with justification, remains an important source of climate information in many rural communities. At the community level, farmers in Zimbabwe and Malawi have identified local language radio programs as credible and accessible mechanisms to deliver forecasts if they occur with follow up meetings with extension agents or other intermediaries52. Internet based tools, such as Google maps, and graphical tools are already being used for participatory, large-scale information development. However, these tools are inherently limited in communicating the relevant local context and the consequences (positive and negative of information use). For most locations, the governance context in which EWSs are embedded is also key. The links between the community-based approach and the national and global EWSs are weak at present53. Improving the complementarity and legitimacy of both approaches is a new challenge to address especially in developing the institutional foundations for global climate early warning information systems envisioned by the Global Framework on Climate Services (see section 1. Introduction).

There is a critical need to approach and support early warning through DRR and Climate Change Adaptation (CCA)54,22, and the overarching processes involved in CRM. This requires a framework that uses climate change scenarios not above but within risk and vulnerability profiles, thereby capturing the nature of capabilities and decision-making networks. These form the basis for effective EWS design and implementation. The cases above, and other efforts, have demonstrated that social protection and early warning information interventions can provide DRR while helping to meet the goals of adaptation to changes in extreme events. Furthermore, sustainable development prospects are very dependent on the effectiveness of the many networks of EWS'57. In these networks, subtle rules of interaction emerge that shape the context in which resource-related decisions are

To ensure that DRR and CRM are integrated utilising appropriate systems, information and tools, some transversal capacities need to be established between the scientific community studying and analyzing the climate information (at timescales relevant to both DRR and CRM), and the decision-makers who are required to consider the full spectrum of the impacts of climate variability and change. Decision makers across all facets of society also need to be aware of the changes, risks and impacts threatening their societies and find appropriate ways to adapt to and protect these from the most damaging changes. They should also consider climate as a resource, with beneficial aspects that can be exploited, through application of timely and appropriate climate information, tools and products.

Rainfall in the Sahelian region of West Africa experiences strong variability over time-scales ranging from intra-seasonal (including long dry spells and false onset) to inter-annual and decadal. At the longest time scale, climate change is shifting the desert boundary and altering the landscape. This strong variability has an impact on many sectors, including health, agriculture and water management. The major impacts of climate variability in this region make CRM an imperative for the livelihoods of Sahelian communities. Each time scale of variability requires a specific climate risk plan.

As is the case in most French colonized countries in Africa, ANACIM, the national weather service of Senegal, is in charge of providing meteorological services to the country. ANACIM, in partnership with the CGIAR research program on Climate Change, Agriculture and Food Security (CCAFS), initiated a pilot project in 2011 to test the usefulness of probabilistic seasonal forecast information to peanut farmers in Kaffrine. In addition to ANACIM, the key stakeholders participating in the project include local government technical services, local farmers and NGOs. A big challenge in the training was to go through many key and important steps in achieving good CRM, including producing

skillful seasonal forecasts at district level, presentation of probabilistic seasonal forecasts using easy-to-understand terminology, training farmers with the new information for them and translate this into decisions. This study concludes with lesson learned during the initial year of the project.

Improving Climate Risk Management at Local Level –

Techniques, Case Studies, Good Practices and Guidelines for World Meteorological Organization Members 499

When the wind changes direction to fetch the rainfall

Birds crying as if it calls men to go to field and woman to stay at

Early flowering of many trees species: Néré, dimb, tamarinier, sone

The shooting star direction indicates which zone will receive excess

When the rain is settled in June the 24th and we start the millet

Apparition of stars shaped as elephant

Butterflies and libellees are numerous Some persons feel heavy in their body

When dark clouds become white **Good rainy season** When snakes and frogs are more numerous than usual

Net appearance of seven stars in the sky

When we observed dew in the morning

Another challenge was how to explain to farmers, in easy-to-understand terms, how climate forecasts work. Many farmers knew about weather forecasts, communicated through the weather bulletin on national TV. But the real challenges were to convince them that the rain could be forcast one to two months ahead, and to help them to understand the probabilistic nature of forecasts at this lead-time. The basis of seasonal forecasting was explained to the farmers by calling upon their intuition. When asked, "When it is hot, why do people go to the beach?" they responded that the sea breeze brings fresher air. They were then asked,

around the 14th of July we can expect good harvest

their elders were wrong and only the scientists were right. The strategy was to listen to them and understand the aspects of their traditional knowledge that might be climate related. The farmers were welcomed as guardians of knowledge passed from generation to generation, and invited as experts to share their indigenous climate knowledge. According to their tradition, being elder means possessing wisdom. The farmers were asked to specify whether each indicator was for immediate weather, or for climate conditions for the upcoming season. Some of the indicators were clearly just coincidental events with no apparent link to the climate, but many were very much related to climate and specifically with the high humidity and high temperature associated with the monsoon system (Table 2). After carefully listening, the scientists acknowledged the farmer's memories and explanations.

**Climate variability Indicators**

Hot night time **Major rainy event** When wind is shifting direction

rain this year

When the sky is high

**Table 2.** Quotes from farmers on their perceived climate variability indicators

**End of the season** When frogs start chanting

*4.1.5. Explaining the basis of seasonal forecasting* 

home

**Onset of the rainy** 

**Good cropping** 

**season** 

**season** 
