Observatory PhD student George Tsitati is researching locally-led anticipatory action options to counter climate-induced humanitarian crises in East Africa. Here, he reflects on his work so far, drawing on literature as well as discussions with pastoral communities and elders in nortthern Kenya to call formore locally-responsive early warning systems.
The climate crisis
Climate change is no longer a remote peril in East Africa; rather, it has become an undeniable reality for impoverished and marginalised communities in the dryland regions of Kenya.
These communities, whose livelihoods rely on pastoralism and agropastoralism, are at the forefront of this crisis, facing colossal risks. Rising temperatures, compounded by endogenous factors such as locust invasions, intra and inter-communal conflicts and exogenous influences like geopolitical conflicts and soaring food prices, have exacerbated the impacts of droughts and floods, leading to recurring humanitarian disasters.
Early warnings for all
Early warning systems (EWS) are indispensable for community preparedness against such disasters. The Sendai Framework for Disaster Risk Reduction (2015-2030) underscores the global imperative to expand access to multi-hazard early warning systems. Aligning with this objective, the Risk-Informed Early Warning Partnership (REAP) brought together a coalition of government, humanitarian, and development stakeholders to make the case to extend early warning coverage to one billion people in developing nations by 2025. In our region, an Early Warnings For All Action Plan for Africa was launched during the African Climate Summit in September 2023, aiming to ensure that accurate and timely disaster information reaches all segments of African society, with a particular focus on vulnerable populations.
An effective EWS should empower individuals, organisations, and communities to anticipate and prepare for potential disruptions by providing accurate, timely, and actionable information. It must be people-centred and encompass an end-to-end approach, integrating risk knowledge, monitoring, early warning, dissemination, and response capabilities. This involves collecting data on potential hazards, analysing the data to understand the risks, disseminating warnings in accessible and understandable ways, and ensuring that communities can respond effectively.
Early warning disconnects
Jameel Observatory discussions with pastoralist community elders in Isiolo, Kenya, in May 2023 revealed that climate change, urbanisation, globalisation and sociocultural differentiation have disrupted the time-honoured indigenous early warning systems that they use to forecast weather patterns. Further, while modern EWS are emerging as potential solutions within communities, they do not yet provide access to timely, understandable, and actionable information.
Consistently, elders in the Isiolo meeting underscored the significant disconnects between formal early warning systems and the communities they aim to protect.
One elder’s view reflects a widespread sentiment: “Even when early warnings are issued, they often go unnoticed or are misunderstood.”
Elders attributed these gaps to top-down system design, a disregard for indigenous knowledge, the inherent uncertainties associated with the warnings, and the reactive response from the issuers of warning information.
The disconnects between EWS and the needs of communities leave many populations vulnerable to escalating climate-induced shocks such as droughts and floods.
A growing body of research supports these findings (see for example: Farr et al., 2022; Levine et al., 2023). The mechanistic nature of modern EWS often clashes with the complex, socio-spatial decision-making processes embedded within local contexts while the reliance of formal drought response systems on probabilistic rather than deterministic or actionable formats, coupled with a lack of contextualisation and linguistic accessibility, further exacerbate these challenges.
As a result, modern EWS are often divorced from local decision-making processes of communities and elders, who have developed a pervasive distrust of these systems. As one elder pointed out: “We have been let down before by predictions that turned out to be false. It is difficult to trust these warnings now.”
Such experiences create a cyclical pattern where uncertain warnings reinforce existing scepticism, hindering the effective use of climate information and undermining the system’s intended purpose. Similar findings on the impact of uncertainty on the usability of early warning information have been documented within the region.
The potential of indigenous knowledge
Conversations with elders revealed the continued significance of indigenous knowledge in the face of limited access to localised modern early warning information.
In this context, Ake’s (1991) conceptualisation of ‘indigenous knowledge’ is useful as it goes beyond traditional practices to encompass evolving, hybridised systems that reflect contemporary community values, priorities, interests and aspirations.
In northern Kenya, this indigenous knowledge, often communicated in local languages and accessible formats, is rooted in closely observed animal behaviour, vegetation patterns, and subtle environmental changes.
As one elder said: “We predict future droughts by observing animal behaviour and environmental changes.” This knowledge empowers communities to proactively adapt and make informed decisions safeguarding their livelihoods.
Regrettably, this invaluable knowledge is often marginalised: “Our knowledge is valuable, but we are rarely asked to contribute,” lamented another elder.
This dynamic perpetuates a dichotomy in which modern EWS are privileged over indigenous knowledge systems. While indigenous systems offer invaluable, practice-based knowledge, they are frequently marginalised as inferior epistemologies. Research suggests that undervaluing existing indigenous knowledge positions modern EWS as competing rather than complementary systems, diminishing their overall effectiveness and social acceptance.
However, a more nuanced perspective emerges from the discussions with elders. While modern EWS grapple with perceived accuracy, timeliness, usability, and efficacy issues, indigenous EWS also face limitations such as subjectivity. Yet, indigenous systems offer unparalleled localised granularity and specificity, often surpassing the capabilities of modern EWS. Therefore, effectively addressing the needs of northern Kenya requires a balanced approach that acknowledges the strengths and weaknesses of both knowledge systems and explores opportunities for their integration.
Hybrid approaches needed
These discussions reinforce a paradigm shift away from viewing indigenous knowledge and scientific approaches as distinct entities towards collaborative frameworks that leverage their complementary strengths. A hybrid model, rather than replacing one system with another, harnesses the advantages of both.
A growing body of research offers theoretical and practical guidance on such hybrid approaches (e.g., Carter et al., 2019). While full integration may not always be feasible, Dube et al. (2024) advocate capitalising on their complementary nature to use indigenous knowledge to corroborate scientific observations and to enhance community buy-in by incorporating local methods into the early warning process. Doing this is instrumental in building trust in disseminated information.
One elder aptly captured this sentiment: “Warnings need to consider our way of life, our movements, and our economic conditions to be truly useful.”
According to the United Nations Office for Disaster Risk Reduction, such local, ‘bottom-up’, ‘end-mile’, or ‘first-mile’ approaches to early warning, with the active participation of local communities, including marginalised groups, reduces their vulnerabilities while leveraging and strengthening local capacities.
Addressing the escalating climate crisis in Kenya’s drylands requires a fundamental transformation of EWS design and implementation. The prevailing gaps between top-down, scientifically driven systems and the realities of pastoralist communities underscore the imperative for people-centred approaches that prioritise indigenous knowledge and local agency while reducing uncertainties and supporting livelihoods.
Decolonising disaster management systems is essential to achieve effective and equitable EWS, fostering collaboration between scientists, policymakers, and community members.
For this to succeed, we need to critically re-evaluate and reconstruct early warning information, freeing it from cultural bias, to ensure EWS are culturally appropriate.
By investing in local capacity building and bridging the chasms between traditional and modern knowledge, we can create early warning systems that are responsive to local needs and resilient to the challenges posed by climate change.
