Kenya's national drought early warning system has a new capability: the ability to forecast outbreaks of acute child malnutrition weeks and months before conditions deteriorate on the ground.
A machine learning tool developed by researchers at the University of California, Berkeley, in collaboration with Cornell University, has been formally integrated into the drought monitoring processes of Kenya's National Drought Management Authority. The announcement was made in Lyon, France, at the One Health Summit, a high-level international gathering convened by French President Emmanuel Macron.
The tool was expanded and tested in Kenya with support from the Jameel Observatory for Food Security Early Action, an international partnership led by the University of Edinburgh with the International Livestock Research Institute, Save the Children, the Abdul Latif Jameel Poverty Action Lab, and Community Jameel.
From Raw Data to Forward-Looking Forecasts
Kenya's NDMA collects monthly measurements of mid-upper arm circumference for children under five across the country—a routine indicator used to estimate acute malnutrition prevalence. That data existed but had never been used to generate forecasts. The machine learning approach developed by Susana Constenla-Villoslada, a doctoral researcher at UC Berkeley, addresses that gap directly.
Historical malnutrition rates are combined with data on weather patterns, conflict, and food prices to train models that learn to recognize how nutritional risk builds over time. Because malnutrition changes gradually and its underlying drivers tend to persist, past measurements are strongly predictive of future ones—enabling reliable forecasts at one, three, and six-month timeframes.
Rather than predicting drought directly, the tool anticipates nutritional risk based on observed trends in child measurements, giving government agencies earlier sight of where wasting is likely to rise before conditions worsen. Findings published in the Proceedings of the National Academy of Sciences confirm that the methodology substantially outperforms previous approaches, which struggled to predict malnutrition even in the present day.
A Dashboard Built into Government Systems
Forecasts generated by the tool feed into a dashboard that has been integrated into the NDMA's formal early warning processes and is used during meetings of the Kenya Food Security Steering Group—a body that brings together government departments, United Nations agencies, donors, and non-governmental organizations to coordinate food security analysis and response.
Technology transfer has enabled the NDMA to operate, update, and maintain the tool independently. The model updates on a monthly basis and produces outputs at both national and sub-national levels. Work is now underway to explore how the forecasts could support trigger mechanisms for anticipatory action, with piloting planned across three counties: Isiolo, Tana River, and Marsabit.
The Organization Behind the Observatory
The Jameel Observatory was established in 2021 in Nairobi, Kenya, as part of a broader network of research institutions supported by Community Jameel. Founded by Mohammed Abdul Latif Jameel, KBE, Community Jameel is a global organization that advances science and learning to help communities address the compounding pressures of climate change, food insecurity, and disease.
Mohammed Jameel's philanthropic focus has long centered on building durable scientific institutions rather than short-term interventions—a philosophy evident in the Observatory's community of practice, which brought together scientists, humanitarian professionals, pastoralists, and international organizations to adapt and refine the machine learning model before its operational deployment in Kenya.
George Richards, director of Community Jameel, said the tool represents the kind of evidence-based action the Observatory was designed to enable: harnessing machine learning to forecast when children face acute malnutrition risk and turning that data into action that governments can use.
The Scale of the Problem
Kenya is home to more than 6 million children under the age of five. Recent national food and nutrition analysis estimated that more than 740,000 children aged 6 to 59 months required treatment for acute malnutrition between April 2025 and March 2026—concentrated largely in arid and semi-arid areas where drought, conflict, and food price shocks compound one another.
Globally, the 2024 SEASNUT report produced by the Jameel Observatory in collaboration with UNICEF estimated that 45 million children under five were wasted worldwide, with wasting linked to roughly 13% of deaths among children in that age group each year.
The Kenya deployment is closely aligned with commitments made at the 2025 Nutrition for Growth Summit in Paris, where Community Jameel served on the International Advisory Group. The summit's organizers have pointed to data-driven forecasting tools as central to closing the nutrition gap and ensuring that investment in nutrition delivers measurable results for children in need.
© 2026 ScienceTimes.com All rights reserved. Do not reproduce without permission. The window to the world of Science Times.
