Browsing by Person "Kayamo, Samuel Elias"

Now showing 1 - 1 of 1

Use of seasonal forecasts in smallholder agricultural decision-making in the Central Rift Valley of Ethiopia
(2025) Kayamo, Samuel Elias; Berger, Thomas
Smallholder farmers in Ethiopia’s Central Rift Valley face pronounced risks from climate variability and erratic rainfall, challenges that threaten agricultural productivity, food security, and rural livelihoods. Rising climate hazards have spurred the promotion of seasonal precipitation forecasts as a promising means of supporting adaptation, yet the translation of such information into tangible adaptive action depends on a complex interplay of local agro-ecological conditions, available adaptation strategies, and behavioral responses. This thesis provides a comprehensive, interdisciplinary investigation into the economic value, adoption dynamics, and policy implications of seasonal forecast information for smallholder farmers, integrating agent-based modelling, dynamic risk assessment, crop-growth simulation, and framed field experiments. A principal focus of the research is the evaluation of adaptive management strategies for smallholder farmers enabled by seasonal forecasts. Examined strategies include crop and cultivar selection in response to rainfall outlooks, optimized planting dates, forecast-driven fertilizer management, and flexible in-season adjustments (such as crop switching or tied ridging). Each option is rigorously evaluated using observational, experimental and simulated data. In assessing the practical impacts of integrating seasonal rainfall forecast information into smallholder agricultural decision making, the results of this thesis indicate that forecast-based cultivar selection has the potential to support more effective management strategies for farmers in Ethiopia’s Central Rift Valley. By enabling better alignment of cultivar choices with anticipated seasonal rainfall conditions, farmers can enhance the adaptive capacity of their management practices in the face of climate variability. While the observed financial gains under realistic forecast accuracy are modest, these findings highlight that forecast-based cultivar selection can serve as a valuable decision-support tool. However, realizing the full potential of this approach depends not only on improvements in forecast skill, but also on the availability of reliable evidence regarding cultivar performance under diverse weather conditions and on substantial changes to seed breeding and distribution systems. Only when forecast-matching cultivars are made available to farmers promptly can the benefits of high-accuracy seasonal rainfall forecasts be more fully achieved. In the subsequent analysis, this thesis applies a state-contingent embedded risk framework to systematically explore how the timing of smallholder management decisions—specifically crop choice, sowing date, tied-ridging, relay cropping, and fertilization—can be optimized in light of seasonal rainfall forecast information. Using multi-stage discrete stochastic programming, the study evaluates adaptive strategies at the whole-farm level by simulating crop yield responses to management choices across 2,400 possible weather trajectories. The results show that forecast-informed management decisions can improve farmer income, but the extent and consistency of these benefits vary across seasons. The findings further reveal that opportunities for in-season adjustment—rather than choices made solely at the start of the season—are especially critical for achieving positive results in response to forecast information. By evaluating the long-term impacts of forecast-based decision making at the whole-farm level in the Central Rift Valley, this study emphasizes the need for more tailored and effective communication and advisory services of seasonal rainfall forecasts. In addition, the analysis highlights the inherent unpredictability of agricultural outcomes under climate uncertainty and demonstrates the continuing importance of building empirical understanding of how management actions and varying weather conditions together shape farm performance. These insights suggest that policy interventions aimed at strengthening real-time advisory systems and supporting farmers’ capacity for flexible, adaptive management are essential for fully realizing the benefits of seasonal rainfall forecasting in smallholder agriculture. The third component of the thesis explores how smallholder farmers receive, interpret, and act upon seasonal precipitation forecasts, drawing on evidence from framed field experiments conducted in Ethiopia’s Central Rift Valley. The analysis demonstrates that neither improvements in forecast accuracy nor dissemination of information alone are sufficient to induce significant behavioral change among farmers. Adoption is most likely when seasonal precipitation forecasts are communicated repeatedly, presented in clear and actionable formats, and tailored to local realities through trusted channels. The results further indicate that factors such as farmers’ education levels, prior experience with seasonal forecasts, and regular engagement with extension services play a central role in facilitating effective use of such information. The findings highlight the potential of digital innovations, such as smartphone-based advisories and AI-supported tools, to improve the reach and personalization of seasonal precipitation forecasts, provided these solutions are developed through participatory and user-centered approaches. Overall, the study underscores the importance of aligning advisory services with both the informational and contextual needs of smallholder farmers in order to foster more effective and inclusive adaptation to climate variability. Overall, the results of this thesis emphasize that the benefits of seasonal rainfall forecasts can only be fully realized through an integrated approach. This requires the combination of advances in forecast technology, adaptive input systems, effective communication, and supportive policy environments. Comprehensive and locally tailored adaptation packages—linking seasonal rainfall forecast information to improved access to seed and inputs, credit, training, and extension services—emerge as the most effective strategy for strengthening resilience. Ultimately, by connecting quantitative modeling, empirical experimentation, and policy analysis, this thesis provides a robust foundation for scaling up inclusive, impactful advisory systems based on seasonal rainfall forecasts to better equip smallholder farmers for managing risks associated with increasing rainfall variability.