Co-designing Indices for Tailored Seasonal Climate Forecasts in Malawi

In central and southern Malawi, climate variability significantly impacts agricultural production and food availability owing to a high dependence on rain-fed maize production. Seasonal climate forecast information has the potential to inform farmers' agricultural planning, thereby improving preparedness to extreme events. In this paper we describe and evaluate an approach to co-designing and testing agro-climatic indices for use in seasonal forecasts that are tailored to farmer-defined decision-making needs in three districts of central and southern Malawi. Specifically, we aim to (a) identify critical maize specific agro-climatic indices by engaging key stakeholders and farmers; (b) compare and triangulate these indices with the historical climate record in study districts; and (c) analyze empirical relationships between seasonal total rainfall and maize specific indices in order to assess the potential for forecasting them at appropriate seasonal timescales. The identified agro-climatic indices include critical temperature/rainfall thresholds that are directly associated with phenological stages of maize growth with direct implications for maize yield and quality. While there are statistically significant relationships between observed wet season rainfall totals and several agro-climatic indices (e.g., heavy rainfall days and dry spell), the forecast skill of the UK Met Office's coupled initialized global seasonal forecasting system (GloSea5) over Malawi is currently low to provide confident predictions of total wet season rainfall and the agro-climatic indices correlated with it. We reflect on some of the opportunities and challenges associated with integrating farmers' information needs into a seasonal forecast process, through the use of agro-climatic indices.

Automated summary

This study investigates the co-design and testing of agro-climatic indices in central and southern Malawi to improve agricultural planning and preparedness. Critical maize-specific indices were identified through stakeholder engagement, historical climate record comparison, and empirical relationship analysis with seasonal rainfall. The study reflects on the challenges and opportunities of integrating farmers' information needs into seasonal forecast processes using agro-climatic indices.