Agro-meteorological assistance in Tanzania to increase the resilience of farming communities to climate challenges

Tanzania is home to more than 63 million people, a majority of whom depend on the agricultural sector. This sector is dominated by smallholder farms: according to the 2019/2020 national agricultural census, the country has around 7 million smallholder farming households, who together account for close to 75% of national agricultural production (Tanzania National Bureau of Statistics, 2019/20) and produce 95% of the country's food needs.

Many cultivate staple crops such as maize and cassava, which are part of the top three most consumed foods in the country. These crops, which represent the two largest planted areas in the Mara region, are not only consumed domestically but also serve as cash crops for export.

For years, local populations have faced increasingly severe droughts, floods, delays in the rainy season, and increasingly uneven rainfall distribution. Due to climate change, yields of staple crops such as maize and cassava could decline (USAID, 2021). Tanzania is ranked the 26th most vulnerable country in the world to climate risk (USAID, 2021).

Phase 1 (2023-2025)

Weather Impact uses the European Centre for Medium-Range Weather Forecasts (ECMWF) model, whose forecasts are based on ensemble data that make it possible to calculate the probability and uncertainty of a rainfall event. Weather Impact translates this data into tailored voice or text messages. In doing so, the company provides weather forecasts to smallholder farmers, helping them improve their farming strategies and increase their resilience to climate challenges.

The project “Kilimo Thabiti” aims to go further by not only sending weather forecasts and general agronomic advice, but tailoring them specifically to each farmer and each type of crop.

It plans:

• A communication channel by SMS allowing farmers to send and receive agro-meteorological information as well as agronomic advice on cassava and maize, based on real-time weather forecasts and alerts about extreme weather events.
• Training for farmers on good agricultural practices and the interpretation of agro-meteorological information.
• The recruitment of “citizen scientists” to collect data for the validation and calibration of agro-meteorological advisories, and to act as advocates for weather- and climate-smart agriculture within their communities.

Phase 2 (2026-2030)

The project is now moving into a scientific evaluation phase. Wageningen University & Research is joining the consortium to conduct a randomised controlled trial establishing a causal link between the Kilimo Thabiti system and its effects on farmers.

The study will focus on 8,000 smallholder maize-farming households across 240 villages in the Mara region, randomly assigned to three groups:

• a group receiving agro-meteorological SMS messages and online training,
• a second group additionally benefiting from the citizen scientist component,
• and a control group receiving no intervention.

The sample includes both direct beneficiaries (who receive the SMS messages) and indirect beneficiaries (neighbours who may benefit from shared information), with particular attention paid to gender-disaggregated analysis.

The project will also test the viability of two economic models to sustain the service beyond FID funding:

• an integration model with Tanzania's public agricultural extension system, in which Kilimo Thabiti would become an official service offered to farmers by the authorities;
• a bundling model combining the service with complementary offerings, such as agricultural inputs or insurance products, to share costs.

This work will be accompanied by the formalisation of partnerships with the Tanzania Meteorological Agency and the Ministry of Agriculture.

Expected results of phase 2

This impact evaluation aims to document the effects of the Kilimo Thabiti system on maize yields and farming incomes. The study will seek to confirm these results under controlled conditions, while measuring more precisely:

• the specific contribution of the “citizen scientists” component to the quality and relevance of forecasts;
• differences in impact between women and men, and effects on the inclusion of women and youth in agricultural decision-making;
• the extent of spillover effects on farmers not directly enrolled in the service;
• the financial viability of the service, through testing different economic models, with a view to sustaining and scaling it beyond the Mara region — Weather Impact is already active in 10 sub-Saharan African countries and 2 Southeast Asian countries, where the lessons from this evaluation can directly inform the development of similar services.

©Weather Impact